IAEA TECDOC 1721

IAEA TECDOC SERIES

TECDOC No. 1721

Optimizing Productivity of
Food Crop Genotypes in
Low Nutrient Soils

Prepared by the Joint FAO/IAEA Division of

Nuclear Techniques in Food and Agriculture

@
OPTIMIZING PRODUCTIVITY OF
FOOD CROP GENOTYPES IN
LOW NUTRIENT SOILS
 IAEA-TECDOC-1721

OPTIMIZING PRODUCTIVITY OF
FOOD CROP GENOTYPES IN
LOW NUTRIENT SOILS

PREPARED BY THE
JOINT FAO/IAEA DIVISION OF NUCLEAR TECHNIQUES IN FOOD AND AGRICULTURE

INTERNATIONAL ATOMIC ENERGY AGENCY

VIENNA, 2013
 COPYRIGHT NOTICE

All IAEA scientific and technical publications are protected by the terms of the Universal Copyright
Convention as adopted in 1952 (Berne) and as revised in 1972 (Paris). The copyright has since been
extended by the World Intellectual Property Organization (Geneva) to include electronic and virtual
intellectual property. Permission to use whole or parts of texts contained in IAEA publications in
printed or electronic form must be obtained and is usually subject to royalty agreements. Proposals for
non-commercial reproductions and translations are welcomed and considered on a case-by-case basis.
Enquiries should be addressed to the IAEA Publishing Section at:

Marketing and Sales Unit, Publishing Section

International Atomic Energy Agency
Vienna International Centre
PO Box 100
1400 Vienna, Austria
fax: +43 1 2600 29302
tel.: +43 1 2600 22417
email: [email protected]
http://www.iaea.org/books

For further information on this publication, please contact:

Soil and Water Management and Crop Nutrition Section

International Atomic Energy Agency
Vienna International Centre
PO Box 100
1400 Vienna, Austria
Email: [email protected]

© IAEA, 2013
Printed by the IAEA in Austria
November 2013

IAEA Library Cataloguing in Publication Data

Optimizing productivity of food crop genotypes in low nutrient

soils. – Vienna : International Atomic Energy Agency, 2013.
p. ; 30 cm. – (IAEA-TECDOC series, ISSN 1011-4289
; no. 1721)
ISBN 978-92-0-113113-3
Includes bibliographical references.

1. Food crops – Yields – Research. 2. Genotype-environment

interaction. 3. Agriculturally marginal lands. 4. Food crops – Soils.
I. International Atomic Energy Agency. II. Series.

IAEAL 13-00843
 FOREWORD

Global climate change is likely to exacerbate plant abiotic stress in the coming
decades by increasing water stress and by accelerating soil fertility degradation. To respond to
this set of challenges, there is a need to develop agricultural systems with significantly greater
productivity and resilience that at the same time use limited natural resources more
efficiently. Low phosphorus (N) and nitrogen (P) availabilities are primary limitations to
productivity in low input agriculture, and fertilizers are primary resource inputs in intensive
agriculture. A critical feature of future agricultural systems will be new crop varieties with
improved conversion of soil resources to yields. These new cultivars would have improved
productivity in low input systems and decreased input requirements in high input systems.

Many scientists are currently turning their attention to roots, the hidden half of the
plant, as central to their efforts to produce crops with better yields without causing
environmental damage. Several root traits are known to be associated with P and N
acquisition efficiency in low N and P soils. These root traits include root hairs, root length,
root branching and root density. The identification of root traits for enhanced P and N
acquisition is enabling crop breeders to develop new genotypes with better yields in low
fertility soils of Africa, Asia and Latin America. However, in order to use a trait as a selection
criterion for crop improvement, either direct phenotypic selection or through marker assisted
selection, it is necessary to develop protocols to measure accurately the root traits that
enhance N and P acquisition in the glasshouse and in the field, which can provide robust and
rapid evaluation of many root systems’ architectural traits in targeted production
environments using different crops.

The objective of the Coordinated Research Project on Optimizing Productivity of

Food Crop Genotypes in Low Nutrient Soils was to develop integrated crop, soil and nutrient
management practices that help increase crop production in marginal lands by identifying and
promoting the development of food crop genotypes (cereals and legumes) with enhanced
N and P usage efficiency.

This CRP was implemented following the recommendations of a consultants meeting

of international experts. The research network included ten contract holders from Brazil,
Burkina Faso, Cameroon, China, Cuba, Ghana, Malaysia, Mexico, Mozambique and the
United States of America and six agreement holders from Australia, Benin, France, Germany,
Kenya and Nigeria.

The CRP was conducted in collaboration with national agricultural research systems
in Africa, Asia and Latin America, and with three centres of the Consultative Groups on
International Agricultural Research (CGIAR): The Africa Rice Center (WARDA), the
International Institute of Tropical Agriculture (IITA) and International Center for Tropical
Agriculture (CIAT). The CRP was supported by in-house research and the provision of 15N/
14
N isotope ratio analysis of 15N enriched plant samples at the FAO/IAEA Agriculture &
Biotechnology Laboratories, Seibersdorf, Austria. Upstream research on 15N and 32P
methodologies, protocols for evaluation of plant root traits that enhance N and P acquisition
and utilization efficiencies were carried out at the IAEA prior to the commencement of the
CRP and through an individual research contract.

The IAEA wishes to acknowledge P.M. Chalk and all of the CRP participants for their
valuable contributions. The IAEA officer responsible for this publication were
J.J. Adu-Gyamfi and L.K. Heng of the Joint FAO/IAEA Division of Nuclear Techniques in
Food and Agriculture.
 EDITORIAL NOTE

The papers in these Proceedings have been prepared from the original material as submitted by the authors and have not
been edited by the editorial staff of the IAEA. The views expressed do not necessarily reflect those of the IAEA, the
governments of the nominating Member States or the nominating organizations.
This publication does not address questions of responsibility, legal or otherwise, for acts or omissions on the part of any
person.
Although great care has been taken to maintain the accuracy of information contained in this publication, neither the IAEA
nor its Member States assume any responsibility for consequences which may arise from its use.
The use of particular designations of countries or territories does not imply any judgement by the publisher, the IAEA, as to
the legal status of such countries or territories, of their authorities and institutions or of the delimitation of their boundaries.
The mention of names of specific companies or products (whether or not indicated as registered) does not imply any intention
to infringe proprietary rights, nor should it be construed as an endorsement or recommendation on the part of the IAEA.
The authors are responsible for having obtained the necessary permission for the IAEA to reproduce, translate or use
material from sources already protected by copyrights.
Material prepared by authors who are in contractual relation with governments is copyrighted by the IAEA, as publisher,
only to the extent permitted by the appropriate national regulations.
The IAEA has no responsibility for the persistence or accuracy of URLs for external or third party Internet web sites referred
to in this report and does not guarantee that any content on such web sites is, or will remain, accurate or appropriate.
 CONTENTS

SUMMARY ...........................................................................................................................1
ROOT TRAITS FOR BETTER PHOSPHORUS ACQUSITION AND USE IN LOW-P
SOILS ....................................................................................................................................9
Differential ability of maize and soybean to acquire and utilize phosphorus from sparingly
soluble forms in low-and medium-P soils using 32P ..............................................................11
J.J. Adu-Gyamfi, M. Aigner, S. Linic, D. Gludovacz
Contribution of root traits to phosphorus acqusition efficiency by maize landraces in acid soils
of the highlands in Central Mexico .......................................................................................25
J.S. Bayuelo-Jiménez, N. Hernández-Bravo, M.L. Magdaleno-Armas, V.A. Pérez-Decelis ...
M. Gallardo-Valdéz,L. C. Paredes-Gutiérrez, J.P. Lynch
Selection and evaluation of maize genotypes tolerance to low phosphorus soils ....................55
J.C. Yang, H.M. Jiang, J.F. Zhang, L.L. Li, G.H. Li
Phosphorus use efficiency by Brazilian upland rice genotypes evaluated by the 32P dilution
technique ..............................................................................................................................79
V.I. Franzini, F.L. Mendes, T. Muraoka, E.C. Da Silva, J.J. Adu-Gyamfi
Evaluation and selection of common bean (Phaseolus vulgaris L.) genotypes for root traits
associated with phosphorus (P) acquisition efficiency and the use of 32 P isotope in studies on
P uptake by root hairs ...........................................................................................................93
M.A. Miguel, C. Jochua, J.P. Lynch
Phosphorus use efficiency by Brazilian common bean genotypes assessed by the 32P dilution
technique ............................................................................................................................ 103
V.I. Franzini, T. Muraoka, J.J Adu-Gyamfi, J.P. Lynch
Selection of common bean lines, recombinant inbred lines and commercial genotypes tolerant
to low phosphorus availability in an Acrisol soil on the basis of root traits and grain yield .. 119
A. Garcia, L.A. Gomez, A. Morales, M. Mosquera, J. Pascual, M. Hernandez, N.Chávez,
J.M. Dantin, G. Herrero S. Curbelo, J.L. Reyes, J.A. Fustes, E.E. Rodriguez, S.González,
R. Ramos J.J. Drevon
Phosphorus use efficiency for symbiotic nitrogen fixation in voandzou (Vigna subterranea)
using isotopic exchange method in Rhizotron ..................................................................... 135
A. Andriamananjara, L. Rabeharisoa, D. Masse, L. Amenc, C. Pernot, J.J. Drevon
C. Morel
Evaluation of soybean and cowpea genotypes for phosphorus use efficiency ...................... 151
F.K. Kumaga, K. Ofori, S.K. Adiku, Y.O. Kugblenu, W. Asante, H. Seidu, J.J. Adu-Gyamfi
Genotypic variation in phosphorus use efficiency for symbiotic nitrogen fixation in Voandzou
(Vigna subterranea)............................................................................................................ 173
A. Andriamananjara , L. Rabeharisoa, M. Malam, Abdou D. Masse, L. Amenc, C. Pernot,
J.J. Drevon
Genotypic variation in phosphorus use efficiency for symbiotic nitrogen fixation in cowpea
(Vigna unguiculata L) ........................................................................................................ 187
A. Andriamananjara,M. Malam Abdou, C. Pernot, J.J. Drevon
Selection of green manure species for efficient absorbtion of poorly-available forms of soil
phosphorus ......................................................................................................................... 201
V.I. Franzini, F.L. Mendes, T. Muraoka, E.C. Da Silva, J.J. Adu-Gyamfi

ROOT TRAITS FOR BETTER NITROGEN ACQUSITION AND USE IN LOW-N SOILS
........................................................................................................................................... 213
Role of translocted signals in regulating root development and nutrient uptake in
legumes .............................................................................................................................. 215
C.A. Atkins
Assessment of root morphological traits of 16 tropical and four temperate maize cultivars for
nitrogen efficiency in short-term nutrient solution experiments with the cigar roll and growth
pouch methods ................................................................................................................... 223
S. Saifu, G. Schulte auf’m Erley, W.J. Horst
Evaluation and selection of maize (Zea mays L.) genotypes tolerant to low N soil .............. 251
C. The, M.L. Ngonkeu, C. Zonkeng, H.M. Apala
Productivity of upland rice genotypes under different nitrogen doses .................................. 265
K. Traore, O. Traore, V.B. Bado
Using upland rice root traits to identify N use efficient genotypes for limited soil nutrient
conditions ........................................................................................................................... 275
K. Traore, O. Traore, V.B. Bado
Evaluation of upland rice genotypes for efficient uptake of nitrogen and phosphorus .......... 287
A.R. Zaharah, M.M. Hanafi
Biological nitrogen fixation efficiency in Brazilian common bean genotypes as measured by
15
N methodology ................................................................................................................ 299
V.I. Franzini, F.L. Mendes, T. Muraoka, A.R. Trevisam, J.J. Adu-Gyamfi
Genotypic variation of early maturing soybean genotypes for phosphorus utilization
efficiency under field grown conditions .............................................................................. 311
R.C. Abaidoo, A. Opoku, S. Boahen, M.O. Dare
IAEA PUBLICATIONS ON SOIL AND WATER MANAGEMENT AND CROP
NUTRITION ...................................................................................................................... 321
ABBREVIATIONS AND ACRONYMS……………………………………………….......323
LIST OF PARTICIPANTS ................................................................................................. 325
 SUMMARY

This 5-year Coordinated Research Project (CRP) entitled ‘Selection and evaluation of
food (cereal and legume) crop genotypes tolerant to low nitrogen and phosphorus soils
through the use of isotopic and nuclear-related techniques’ established a research network and
supported the efforts of teams of scientists in sixteen Member States (Australia, Benin,
Burkina Faso, Brazil, Cameroon, China, Cuba, France, Germany, Ghana, Kenya Malaysia,
Mexico, Mozambique, Nigeria and the United States of America The aim of this CRP was
the development of integrated crop, soil and nutrient management practices to increase crop
production in marginal lands by identifying and promoting the development of food crop
genotypes (cereal and legume) with enhanced nitrogen (N) and phosphorus (P) use efficiency
and greater productivity in marginal lands. The research teams adopted an integrated approach
to crop, soil, and nutrient management practices in predominant cropping systems during the
project implementation. Studies were conducted along four main areas of investigation to (1)
Develop and validate screening protocols for plant traits that enhance N and P acquisition and
utilization in major food cereal and legume crops grown in low fertility soils (2) Employ
validated screening protocols to identify genotypes with superior N and P acquisition and / or
utilization. (3) Identify mechanisms for adaptation and high productivity of selected legumes
and cereals to low N and P soils using of isotopic techniques (stable 15N and radioactive 32P /
33
P) and (4) Assess the selected genotypes with traits for enhanced nutrient acquisition and/or
utilization in selected cropping systems, including yield and productivity.

The studies conducted within this CRP concern two major food security cereal crops
namely upland rice (Oryza sativa L.) and maize (Zea mays L.) and three legumes namely
common bean (Phaseolus vulgaris L.), soybean (Glycine max L.) and cowpea (Vigna
unguiculata L.). Studies were conducted across a wide geographical area in both the northern
and southern hemispheres under a wide range of environmental and edaphic conditions.
Experiments were conducted in the laboratory or in the glasshouse for rapid screening at the
early seeding stage using the paper-roll cigar method, while the final evaluation and selection
of the genotypes were carried out under field conditions. The main specific research results
and recommendations arising from this CRP are summarized based on the following four
outputs below:

1. PROTOCOLS FOR EVALUATION OF ROOT TRAITS (ARCHITECTURE AND

MORPHOLOGY) CONTRIBUTING TO ENHANCED N AND P ACQUISITION

Root architectural phenes influence P and water acquisition from the soil. Crop
genotypes with shallow roots, many basal root whorls, adventitious roots and basal roots have
advantages in acquiring P from low P soils, while genotypes with deeper basal roots and
longer primary roots will acquire water from deeper soil horizons. Developing protocols to
accurately measure root traits that enhances N and P acquisition in the glasshouse and in the
field, which can provide robust and rapid evaluation of many RSA traits in targeted
production environments using different crops is vital. A simple visual method to evaluate
root phenes of crops at the early seedling stage using the paper-roll cigar method or at the late
growth stage in the field phenotyping using shovelomics were developed and validated. This
methodology is available in 3 different languages at a website http://roots.psu.edu or
http://www.naweb.iaea.org/nafa/swmn/news-swmcn.hmtl. The field phenotyping using
shovelomics should have utility for evaluating food crop genotypes for low P and drought
tolerance in developing countries of Africa, Asia and Latin America. During the 5-year period
of the CRP, a new version of the SIMROOT model, capable of simulating a large diversity of
root systems, was created. In addition, protocols for fractionation of soil P using 32P to

1
elucidate mechanisms of P acquisition from different soil P pools were developed and fine-
tuned in the Seibersdorf Laboratories to support the CRP.

2. VALIDATION OF SCREENING PROTOCOLS TO SELECT GENOTYPES WITH

SUPERIOR N AND P ACQUISITION AND / OR UTILIZATION

Germplasm of maize, upland rice, common bean, cowpea and soybean were
exchanged among the participants or were acquired from four CGIAR Institutes (soybean and
cowpea lines from IITA, maize from CIMMYT, common bean from CIAT, upland rice from
The African Rice Center (formerly WARDA) and other advanced research institutes. One
hundred and fifty to two hundred genotypes of maize, common bean, soybean, rice and
cowpea were rapidly screened at the early seedling stage for enhanced N and P acquisition
using the paper-roll cigar method ( Fig. 1a) or PVC tubes of length 15.0 cm and 3.4 cm inner
diameter (Fig. 1c). Twenty five genotypes with different abilities to grow under low P and N
conditions selected from the different crops at the seedling stages were further evaluated
under field conditions (Fig. 1b) at two or more sites under diverse agro ecological
environments. Data on environmental variables (latitude, longitude, altitude, rainfall, and
temperature), soil classification, soil physico-chemical characteristics and systems studied
were recorded. Root characteristics evaluated included Basal Root Whorls Number (BRWN),
Root Hairs Length Density (RHLD), Basal Root Growth Angle (BRGA), Root Length (RL),
Root Length Density( RLD), Root Angle (RA), Root Branching (RB), Adventitious Root
Length (ARL), Adventitious Roots Number (ARN) Adventitious Root Branching (ARB),
Basal Root Length (BRL), Basal Root Number (BRN), Basal Root Branching (BRB), Basal
Root Depth (BRD), Primary Root Depth (PRD), Primary Root Branching (PRB), Seminal
Root Length (SRL), Lateral Root Length (LRL), Lateral Root Number (LRN), Seminal Root
Elongation (SRE), Root Arbuscular Mycorrhiza Colonization (RAMC) and Root Biomass
(RB). In addition, data on Shoot Biomass (SHB), Grain Yield (GY), Stem Diameter (STDIA),
Leaf Area Index (LAI), Plant Height (PLHT), Leaf Chlorophyll (LCHL) and Nodule Weight
(NODWT) for legumes were recorded (see Fig. 2).

The results from the 16 countries for the five crops (upland rice, maize, soybean,
common bean, soybean and cowpea showed that (i) Branching angle interval and seminal root
length were identified as suitable root selection parameters for soil N use efficiency, while
adventitious rooting and root hair formation were identified as suitable plant parameters for
selecting P use efficiency (ii) P efficiency strongly correlated with genotypic differences in
root hair length, root hair plasticity, lateral root number (iii) Genotypes with more root
cortical aerenchyma (RCA) had deeper roots and produced 2 times shoot biomass in low N
soils than genotypes with less (RCA). For beans, RHLD, BRGA, BRWN, ARL and BRN
were identified as the most suitable traits (Fig. 3 and Table 1). We concluded that (1) seedling
screening tools demonstrated significant genotypic variation for root traits, including root
length, angle, number of axial roots and branching as well as root hair parameters (length and
density), (2) Cultivars identified with some of these traits proved superior for uptake of P and
N under conditions of nutrient stress, and (3) Cultivars with superior growth, nutrient
acquisition and use efficiency obtained good yields of grain under conditions of nutrient stress
(Fig. 3).

2
 (a) (b) (c)

FIG. 1. (a) A growth chamber cigar roll method (b) field sampling and (c) transparent glass tube for
evaluating genotypic variations in roots (root architecture and morphology) and plant growth traits
associated with N and P acquisition efficiency in maize at the vegetative growth stage.

Root Architecture Basal Root Whorls Number (BRWN)

Root Hairs Basal Root Growth Angle (BRGA)

FIG. 2. Root phenes associated with genotypic differences in adaptation to low nitrogen and
phosphorus.

3
 4
203
NER ER
133

2 34 227 230
206 237
124 234 185
239 219 214
138 135 27 216 6 119
165 118 205
5 215 150
Shoot dry weight (standardized value of shoot

208 209
240 225 241 235 233232
120 143 20 180 198 189
0 238 62
115127 181 182 199 210 41 184
57 197
96 63 111 75 125 161 195
117
85 102 78 162
83 109 113 140 172
99 38
65 53 114 171
-2 55 186

NENR ENR
ENR
-4
-1.5 -1.0 -0.5 0.0 0.5 1.0 1.5

4
dry weight at high P)

185

2 6 241
118
240 237
234
91 115 235 144
138 106
124 20 182
238 119
0 239
227 215 216
85 198
140 75
62 117
109 199
83 78 127
99 113
-2

-4 ENR
-1.5 -1.0 -0.5 0.0 0.5 1.0 1.5

The scores of phosphorus efficiency index (PEI)

FIG. 3. Division of charts of 242 and 50 common accessions according to P efficiency and
standardized value of shoot dry weight under high P conditions. Phosphorus efficiency is expressed as
PEI, which is an assessment index calculated from principal component analysis. Standardized values
of shoot dry weight are estimated as the following function: = ( − )⁄ . Categories
represented by efficient and responsive (ER), non-efficient and responsive (NER), non-efficient and
non-responsive (NENR), and efficient and non-responsive (ENR). Accession numbers are indicated
[1].

3. ASSESS THE EFFECT OF SELECTED GENOTYPES ON CROPPING SYSTEMS

PERFORMANCE

The assessment and selection of selected crop genotypes with enhanced different root
characteristics that explore nutrients from different soil depths under field conditions is
relevant for enhancing food security and long-term sustainability of soil fertility. Five to ten
genotypes were further accessed in regard to their performance for improved productivity in
low-input systems. Five rice and five maize genotypes were selected that had the highest N
use efficiency (66˗80%) and the highest P use efficiency (6˗8%), and also provided a 15–30%
yield increase over the other genotypes with low N and P use efficiency. For common bean,

4
soybean and cowpea, genotypes with deep root systems that produced 20–40% better yield,
45% increase in BNF and 40% less soil erosion in low P soil were identified. P–efficient
legumes contributed to soil fertility by enhanced BNF, which is quite sensitive to P supply.
Economically, the greater productivity of N efficient genotypes would permit third world
farmers greater flexibility in soil management options, purchasing fertility inputs, etc., in
addition to greater food security and household income.

TABLE 1. PHENOTYPIC CORRELATIONS AMONG NODAL ROOT TRAITS AND SHOOT

BIOMASS, NUMBER OF NODAL ROOTS (NODAL_NO), NODAL ROOT LENGTH
(NODAL_RL), NODAL BRANCHING (NODAL_BR), NODAL ROOT ANGLE (NODAL_RA),
SHOOT DRY WEIGHT (SHDW), AND GRAIN YIELD (GY)

HP/LP a Nodal_No Nodal_RL Nodal_Br Nodal_Ra ShDW Gy

Experiment 1 (n = 242)
Nodal_No 0.49*** 0.25*** 0.39*** 0.37*** 0.38*** 0.10
Nodal_RL 0.42*** 0.39*** 0.50*** 0.48*** –0.09 0.24**
Nodal_Br 0.55*** 0.51*** 0.49*** 0.51*** 0.10 0.21**
Nodal_Ra 0.33*** 0.31*** 0.46*** 0.32*** –0.02 0.07
ShDW 0.49*** –0.03 0.21** –0.02 0.50*** 0.09
Gy 0.18* 0.07 0.28** 0.10 0.16 0.68***
Experiments 1 + 2 (n = 50)
Nodal_No 0.32** 0.36** 0.32** 0.13 0.41** 0.09
Nodal_RL 0.19 0.39*** 0.52*** 0.31** 0.00 0.21*
Nodal_Br 0.25* 0.54*** 0.26* 0.24 –0.03 0.18
Nodal_Ra 0.05 0.21 0.45** 0.32** 0.04 0.25*
ShDW 0.48*** –0.01 –0.01 –0.24* 0.36** 0.05
Gy 0.01 0.38*** 0.35** –0.06 0.02 0.54***
a
For each environment, values below the diagonal represent correlations within the low P treatment; values
above the diagonal represent correlations within the high P treatment; values on the diagonal (italic) correspond
to across-P treatment correlations.
*, **, ***
denote significance at P<0.05, P<0.01, P<0.001, respectively.

4. IDENTIFY MECHANISMS FOR ADAPTATION AND HIGH PRODUCTIVITY

TO LOW N AND P SOILS USING OF ISOTOPIC TECHNIQUES

Nuclear, isotopic and related conventional techniques were employed to obtain

quantitative estimates on optimization of N and P uptake and utilization from fertilizers and
soils. For instance, stable 15N and radioactive 32P) techniques were employed to obtain
quantitative estimates for identifying N– and P–efficient crop genotypes in low N and P
environments. In order to understand the mechanisms of the genotypic tolerance to low–P
soil to utilize P from the sparingly soluble P forms, 5 maize genotypes selected out of 116
inbred lines, were used as the criteria in a 32P isotope tracer experiment to follow the
recovery of 32P in soil P fractions. The L–value and P availability of soil was also assessed.
After the addition of 32P–Pi to the soil with no P fertilizer applied for 25 d, 29.0% of 32P was
quickly transformed into Ca2–P (rapidly available P), and 66.1% of 32P was transformed into
Al–P, Fe–P and Ca8–P (slowly available P). Only 5.0% of 32P was transformed into O–P and
Ca10–P (plant–unavailable P). Moreover, in the soil with P fertilizer applied, 32P
transformation into Ca2–P increased, and the transformation into Ca8–P + Fe–P + Al–P and
O–P, Ca10–P significantly decreased compared to the soil with no P fertilizer applied (P<
0.05). This result suggested a higher rate for water–soluble P transformation to slowly
available and plant–unavailable P in P deficient soil than in soil with sufficient P. Low–P
tolerant cultivar DSY–32 regulated soil P–use efficiency and plant P content according to
exogenous P fertilizer application. However, another low–P tolerant cultivar, DSY–2, used

5
soil P more efficiently, regardless of the application of exogenous P [2]. It was therefore
concluded that the 32P tracer technique proved to be a valuable tool that sought physiological
explanations for superior genotype performance (Table 2).

TABLE 2. L-VALUES OF MAIZE GENOTYPES WITHOUT EXTERNAL P FERTILIZER

Variety no. Specific activity of L-value

Plant (Bq µg-1) Soil (Bq g-1 soil) (µg P g-1 soil)
DSY–30 22.4 ± 2.5† 4216 ± 144 190 ± 28
DSY–2 20.0 ± 0.1 4484 ± 154 225 ± 7
DSY–32 7.8 ± 1.0 3965 ± 136 513 ± 45
DSY–79 11.4 ± 0.8 3572 ± 122 315 ± 32
DSY–48 64.9 ± 2.9 3623 ± 124 56 ± 1
†Values following means are ± standard errors

A similar result was reported in a low–P soil for maize that was more efficient than
soybean in taking up soil P [3]. The available P (Bray II) and the Ca–P were the fractions
most depleted by plants followed by the Fe–P fractions. For common bean, efficient
genotypes with long root hairs had lower specific activity values compared to inefficient
genotypes, since these were able to take up P from two different pools with a greater total P
accumulation (Fig. 4). For upland rice, 15N and 32P were employed to obtain quantitative
estimates for optimization of plant N and P by N– and P–efficient crop genotypes in low N
and P environments. The 15N enrichment in plants ranged between 0.629 to 0.753 atom %
15
N excess, while the 15N enrichment in upland rice seeds ranged from 0.553 to 0.757 atom %
15
N excess. Variety Merah showed the highest N use efficiency in upland rice with 80.7%
and the lowestN use efficiency in upland rice was obtained from variety Tenom with 40.7%.
N use efficiency by these upland rice genotypes showed that high N is utilized (40–80% of
applied N), with good grain yield (Table 3) and P use efficiency is similar to other crops (2.4–
8%).
Specific Radioactivity , (dpm/mgP
x1000)

Short Root Hairs Long Root Hairs

FIG. 4. Specific radioactivity in plant tissue of 4 genotypes contrasting for root hairs:
Genotypes with short root hairs (open bars) and genotypes with long root hairs (full bars).
Genotypes GxG 41 and DxG 53 have long root hairs, while genotypes GxG 23 and DxG 11
have short root hairs. Genotype DxG 53 (a RIL from DOR364xG19833).

6
TABLE 3. 15N ENRICHMENT, AMOUNT OF NITROGEN DERIVED FROM FERTILIZER AND
PERCENTAGE NITROGEN USE EFFICIENCY OF UPLAND RICE
15
Variety N (atom % excess) N derived from fertilizer (kg ha-1) % N use
Plant Grain Plant Grain Total efficiency
Nabawan 0.671 a 0.553 a 40.2 de 23.0 bc 63.2 bc 42.1 bc
Tenom 0.721 a 0.614 a 31.3 e 29.7 abc 61.0 c 40.7 c
WRDA 20 0.629 a 0.684 a 27.7 e 41.5 ab 69.2 bc 46.2 bc
WRDA 99 0.731 a 0.696 a 48.3 cde 51.8 a 100.1 abc 66.7 abc
Sintok 0.711 a 0.654 a 64.0 bcd 38.8 ab 102.7 ab 68.5 ab
Pulut Petai 0.770 a 0.758 a 70.6 bc 43.0 ab 113.6 a 75.7 a
Merah 0.729 a 0.714 a 109.3 a 11.7 c 121.0 a 80.7 a
Kuku Belang 0.753 a 0.685 a 78.6 b 21.1 bc 99.7 abc 66.4 abc
Data within a column followed by the same lower case letter are not significantly different (P<0.05)

5. CONCLUSIONS

The main conclusions from the CRP are summarized as (i) Seedling screening tools
demonstrated significant genotypic variation for root traits. These included root length, angle,
number of axial roots and branching as well as root hair parameters (length and density) (ii)
Cultivars identified with some of these traits proved superior for uptake of P and N under
conditions of nutrient stress (iii) Cultivars with superior growth, nutrient acquisition and
efficiency obtained good yields of grain under conditions of nutrient stress (iv) In some cases
positive agro-ecological outcomes were identified that are related to the performance of
cultivars selected for favourable root traits (vi) Nuclear tools, specifically the use of 15N and
32
P as tracers proved valuable in studies that sought physiological explanations for superior
genotype performance (v) The genotypes of rice, common bean, maize, soybean and cowpea
identified in a number of cases provide valuable resources for plant breeding programmes
aimed at enhancing P and N use efficiency in crops.

The CRP created a database on how cereals and legumes can acquire N and P in low
nutrient soils, and this database could be further expanded and interpreted using multivariate
analysis on how cereal and legume crops can acquire N and P in low nutrient soils. The
studies carried out within the framework of the CRP fall within the major FAO / IAEA
programme on crop improvement in harsh environments, and clearly show the value and
strength of an interdisciplinary research approach. The expertise in agronomy and soil science
of the majority of participants was complemented by expertise in plant breeding and genetics
contributed by other participants and IAEA staff at Seibersdorf and Headquarters in Vienna.

REFERENCES

[1] BAYUELO-JIMÉNEZ, J.S., et al., Genetic variation for root traits of maize (Zea mays
L.) from P'urhépecha Plateau, under contrasting phosphorus availability, Field Crops
Res. 121 (2011) 350–362.
[2] YANG, J.C., et al., Inorganic phosphorus fraction and its translocation dynamics in a
low-P soil, J. Environ. Radioactiv. 112 (2012) 64–69.
[3] ADU-GYAMFI, J.J., AIGNER, M., GLUDOVACZ, D., Variations in phosphorus
acquisition from sparingly soluble forms by maize and soybean in low-P and medium-
P soil using 32P. Proc. XVI Int. Plant Nutr. Colloquium, UC Davis, USA (2009).
http://repositories.cdlib.org/ipnc/xvi/1317

7
ROOT TRAITS FOR BETTER PHOSPHORUS ACQUSITION AND
USE IN LOW–P SOILS
 DIFFERENTIAL ABILITY OF MAIZE AND SOYBEAN TO ACQUIRE AND
UTILIZE PHOSPHORUS FROM SPARINGLY SOLUBLE FORMS IN LOW-
AND MEDIUM–P SOILS USING 32P

J.J. ADU-GYAMFI, M. AIGNER, S. LINIC

Soil and Water Management & Crop Nutrition Laboratory,

D. GLUDOVACZ
Nuclear Material Laboratory, Safeguard Analytical Services,

International Atomic Energy Agency,

Seibersdorf,
Austria

Abstract

A glasshouse pot experiment was conducted to evaluate the differential ability of maize (Zea
mays) and soybean (Glycine max) to utilize soil phosphorus (P) for plant growth from total–P,
available–P and inorganic (Ca–P, Al–P and Fe–P) soil P pools using a carrier-free 32P solution. A
maize variety (DK 315) and a soybean variety (TGX 1910–4F) were grown in pots containing 1 kg of
a low available P soil (Hungarian) or a medium available P (Waldviertel) soil labelled with 32P for 42
days or without 32P (unlabelled) for 42 and 60 days. The shoot and root biomass of maize and soybean
were significantly greater when grown on the Waldviertel than on the Hungarian soils. The shoot P
concentrations were higher for soybean (1.7–2.2 g kg-1) than for maize (1.1–1.4 g kg-1). The total
radioactivity (dpm × 106) was higher in plants grown in Waldviertel than in Hungarian soil and the
values reflected plant P uptake and shoot biomass of soybean and maize. The L–values (µg P g soil-1)
of maize and soybean were higher in Waldviertel (72–78) than in Hungarian (9.6–20) soil. No
significant differences in L–values were observed for maize and soybean grown on the Waldviertel
soil, but for the Hungarian soil, the L–values were higher for maize (20.0) than for soybean (9.6)
suggesting that in this low–P soil, maize was more efficient than soybean in taking up soil P. The
available P (Bray II) and the Ca–P were the fractions most depleted by plants followed by the Fe–P
fractions in the two soils, but differences between the crops were not significant. When soil P is
limited, maize and soybean are able to access P mainly from the available P (Bray II), Fe– and Ca–P
sparingly soluble fractions and not Al–P from the soil.

1. INTRODUCTION

Soils characterized by poor phosphorus (P) availability are widespread globally [1]
and for these soils to be agriculturally productive, they require regular application of water-
soluble superphosphate or ammonium phosphate fertilizers to either maintain the soil P status
of fertile soils or increase that of soils with inherently low P fertility. Soluble phosphate
applied to P deficient soils is retained by iron (Fe), aluminum (Al) and calcium (Ca) ions and
are virtually unavailable to most plant species. It is reported that P fertilizer recovery
efficiencies commonly range from 5 to 20% (sometimes are as low as 1% in high P–fixing
soils) for the first crop and thereafter 1–5% for subsequent crops, indicating there is scope for
improving P use by crops [2].

Plants differ greatly in their ability to grow on low P soils because they have
developed specific physico-chemical mechanisms/processes to utilize P compounds in these
low P fertility soils [3]. These mechanisms include (i) alterations (morphological and
physiological) to root systems, i.e. mycorhizal plants have better water uptake and Al
tolerance in acid soils [4] (ii) secretion of low-molecular-weight organic compounds

11
(exudates production), i.e. malonic, oxalic, citric, malic and piscidic acids secreted by roots of
pigeon pea help to release low-soluble P compounds in soils [5] (iii) secretion of enzymatic
compounds, i.e. phosphatases, and (iv) molecular changes such as enhanced expression of P
transporters.

Studies using crop species having more efficient acquisition of P is a strategy to

evaluate and identify crop plants with genotypic variation in their ability to access and utilize
sparingly soluble forms of soil P (Ca-P, Al-P and Fe-P). This idea has been proposed as a
possible means for overcoming P deficiency stress in soils and as a means to optimize P
fertiliser use in cropping systems where P is poorly available [6]. Intra-specific variations in a
crop’s ability to use sparingly soluble forms (P associated with Al, Fe and Ca) in low-P
available soils have been well documented for lupin [6], pigeonpea [7], soybean, cowpea and
maize [8].

The amount of plant available P in a soil can be defined as the amount of

orthophosphate that leaves the solid phase of the soil and arrives in the soil solution at a time
when the plant is growing and can take it up. Radio-isotopic P techniques, using the principle
of isotopic exchange, allow measurement of the amount of orthophosphate that can be
transferred from the soil solid to the solution over a given time [9], and can thus provide a
powerful alternative means for characterizing soil P availability and the sources of P, with
minimum modifications of soil P forms compared with conventional extraction methods. The
technique has been used to measure the quantity of available P in soils in determining the E-
value or exchangeable P [10], the L-value or labile P [11] using plants grown in a soil labelled
with carrier-free 32P or 33P-orthophosphate, and the A-value or available P [12]. Studies in
soils labelled with 32P to distinguish between different forms of P accessed by plants have
shown that white lupin can access up to six times more P compared with a range of crop
plants [13]. Cluster roots may represent more than 60% of the total root mass. Since P
isotopes provide the unique advantages of being able to identify P sources and quantify their
impacts [14], greater use of 32P and 33P isotopes in P flux and balance measurements offers
considerable advantages over conventional techniques.

The study aimed to evaluate the differential ability of maize and soybean to access and
utilize P from different soil pools using two soils, one low- and one medium-P. The
experiment aimed to test the hypothesis that the efficient P uptake from sparingly soluble P
forms (Al-, Fe- and Ca-P) by different crops can be used as criteria to evaluate crop plants
tolerant to low available P soils.

2. MATERIALS AND METHODS

2.1. Plant growth conditions

Two experiments were simultaneously set up to include two treatments consisting of

non-labelled (without 32P) and labelled (with 32P) imposed on soils with low and medium
available P with maize (cereal) and soybean (legume) as test crops in a factorial design with
four replications. In all there were 20 pots for the radioisotope (including four pots without
plants as control) and 36 pots (including four pots without plants and two sampling periods)
for the unlabelled treatment.

A low-P soil from Hungary (total P, 302; available P (Bray P2), 21; Olsen P, 13.3 mg
kg-1 ; pHKCl, 5.6; classified as Dystric Eutrocrepts), and a medium-P soil (Waldviertel) from
Austria (total P, 502; available P (Bray P2), 44; Olsen P, 12.8 mg kg -1; pHKCl, 5.6) were used.
More detailed physical and chemical characteristics of the two soils are given in Table 1.

12
TABLE 1. PHYSICAL AND CHEMICAL PROPERTIES OF THE SOILS

Properties† Hungarian soil Waldviertel soil

Sand (g kg–1) 830 273
Silt (g kg–1) 88 582
Clay (g kg–1) 82 145
Bulk density (g cm–3) Not determined 1.29
Particle density (g cm–3) Not determined 2.65
Pore volume (%) Not determined 0.51
Saturated water content (%) Not determined 47
pH (H20/KCl) 5.5/4.6 6.5/6.0
Total P (mg kg–1) 302 502
Available P (Bray/Olsen) (mg kg–1) 21/13.3 44/12.8
Inorganic P (Ca/Al/Fe) (mg kg–1) 36/85/65 56/144/68
EC (25oC) (µS cm–1) Not determined 166
Total N (g kg–1) 0.83 1.21
Organic C (g kg–1) 7.91 20
Ca (cobalthexamine) (cmol k g–1) 1.82 13.69
Mg (cobalthexamine) (cmol kg–1) 0.61 3.13
K (cobalthexamine) (cmol kg–1) 0.09 0.15
Na (cobalthexamine) (cmol kg–1) 0.04 0.06
CEC (cobalthexamine) (cmol kg–1) 2.66 23.59
†EC, electrical conductivity; CEC, cation exchange capacity

A maize variety (DK 315) from Austria and a soybean variety (TGX 1910–4F) from
IITA, Nigeria, were grown in plastic pots (1 plant per pot) containing 1 kg of soil in a
naturally lit glasshouse with a temperature regime of 34/21oC day/night and relative humidity
of 40–70%. Each pot received basal fertilizer equivalent to 200 kg−N ha-1 as ammonium
sulphate and 50 kg K ha-1 as potassium chloride. Prior to planting, the weight and P
concentration of soybean and maize seeds used for the experiment were determined. The
amount of P in seed was 0.98 mg P kg-1 soil (0.35 g with 2.8 mg kg-1 P) for maize and 3.48
mg P kg-1 soil (0.59 g with 5.9 mg kg-1 P) for soybean. Phosphorus–32 labelled K2H32PO4
(specific activity of 40.7 GBq mmol-1) was applied to the pots.

To ensure uniform labelling of the soil five portions of dilute 32P solution (50 ml for
Waldviertel and 30 ml for Hungarian soils were applied to a layer of 200 g soil) to achieve
70% of the water holding capacity of each soil. A total of 250 ml for the Waldviertel and 150
ml for the Hungarian containing 12.4 MBq (335 µCi) of a K2H32PO4 solution was applied to
each of the 20 pots containing 1 kg soil. Pre-germinated maize and soybean seeds were sown
at one per pot immediately after the addition of 32P. Twenty ml of inoculum (Bradyrhizobium
japonicum) mixture was added to all the soils. On 20th August (28 days after sowing),
nitrogen was applied at 100 kg−N ha-1 as ammonium sulphate to the maize to ensure adequate
N for the plants. The plants were watered on a pot weight basis and were supplied everyday
with an amount of water equal to the evapotranspiration loss, maintaining as much as possible
constant soil water content in the pots.

2.2. Plant and soil sampling and analyses

The first plant sampling for the labelled and the non-labelled treatments was done at
42 days after sowing (DAS) whereas the non-labelled treatments were allowed to grow till 60
DAS. Soil samples (10–12 g) was taken with a special soil auger (inner diam. 8 mm, outside
diam. 10 mm and length 25 cm) at 0, 1, 5, 42 and 60 DAS, oven-dried at 70 oC for 18 h,
milled and a portion used for analysis. Plants were harvested and separated into shoots

13
(radioisotope-labelled) and shoot and roots (non-radioisotope), chopped into small pieces,
oven dried, weighed, and ground. The root dry weight of the radioactive plants was estimated
using the root/shoot ratio of the non-radioisotope plants, in order to prevent the complications
of root sampling and washing for the 32P-labelled pots. Total P in soils was determined using
the colorimetric method [15] after acid digestion, and available P (Bray P2 and Olsen)
determined by the colorimetric method after extraction.

The inorganic soil P fractions were measured according to a fractionation scheme

based on the method described by Sekiya [16]. Briefly the fractionation involved a sequential
extraction of Ca-P (300 mg of soil extracted with acetic acid), Al-P (extracted with
ammonium fluoride after the extraction of Ca-P) and Fe-P (soil after extraction of Al-P was
washed twice with saturated sodium chloride and discarded and then extracted with and
sodium hydroxide) and the P in extracts determined by a colorimetric method. The 32P
radioactivity in all the fractions (total-P, available-P, Ca-P, Al-P and Fe-P) was measured by
liquid scintillation spectrometry (Packard 2000) using 1ml solution and 9 ml of Aquasol-2
(NEN research product). The ground plant materials were wet digested in 4 ml H2SO4 and 3
ml H2O2 for 2 min until the digest was colourless, total P was measured on diluted aliquots
[17] and 32P was determined by liquid scintillation counting. P in the maize and soybean
seeds was determined after five seed samples, each of 100 mg, were ground and acid digested.
The total radioactivity in each pot at sowing was 744 × 106 dpm.

3. RESULTS

3.1. Physico-chemical characteristics of the soils used

The Waldviertel (silty loam) was slightly acidic, medium total P, available P, and C
content whereas the Hungarian (loamy sand) was acid, comparatively low total P, available P
and carbon content (Table 1). The soil P fractions of the original soil indicated higher Ca-P
and Al-P and exchangeable bases (Ca, Mg, K, and Na) in the Waldviertel than the Hungarian
soil.

3.2. Plant growth and P uptake

The shoot and root biomass of both maize and soybean were significantly greater
when grown in the Waldviertel than in the Hungarian soil and there was a significant increase
in shoot dry weight per plant from 42 to 60 DAS (Fig. 1). Shoot weight of maize increased
from 8.1 g at 42 DAS to 19.1 g plant-1 at 60 DAS in Waldviertel, and from 2.2 g at 42 DAS to
5.7 g plant-1 in the Hungarian soil. For soybean, there was an increase from 4.1 g at 42 DAS
to 9.8 g plant-1 for the Waldviertel and from 1.0 g at 42 DAS to 2.3 g plant -1 in the Hungarian
soil. Shoot weight of maize and soybean decreased by 75–80% when grown in Hungarian
compared to Waldviertel soil irrespective of the two sampling dates (Fig. 1).

Soybean plants grown in Waldviertel soil were well nodulated (0.56 g plant -1) on the
roots whereas plants grown in the Hungarian soil produced very few nodules. There was no
significant difference in shoot and root weight of labelled and unlabelled treatments
confirming that the addition of carrier-free P–32 solution did not result in any change in soil
chemical properties. The root/shoot ratio was generally higher at 42 (0.6–0.7) than at 60 (0.2–
0.45) DAS for maize and soybean and higher for Hungarian than for Waldviertel soil except
at 60 DAS.

The shoot P concentrations were higher for soybean (1.8–2.2 mg g -1) than for maize
(1.1–1.4 mg g-1) and decreased with plant age for maize but not significantly for soybean (Fig.

14
2). Plant shoot P reflected the total and available P concentrations in the two soils with shoot
P of plants grown on Waldviertel soil higher than those grown on the Hungarian soil (Fig. 2).
Although the plant P concentrations were high in soybean compared to maize, the total P
amount (mg P) was significantly higher in maize than in soybean and the values for maize
increased in Waldviertel from 17.6 mg P at 42 DAS to 27.3 mg P at 60 DAS, while for
soybean values decreased from 13 mg P at 42 DAS to 22 mg P at 60 DAS (Fig. 2). There was
80% or more reduction in total P amount when plants were grown in Hungarian compared to
Waldviertel soil.

FIG. 1. Dry weight of shoot (a) and root (b) for maize and soybean grown in Waldviertel and
Hungarian soils labelled (L) and unlabelled (UL) with 32P.

In the Hungarian soil, there was no significant change in Ca-P but a slight decrease in
Fe-P, but a substantial increase in Al-P (85–98 mg kg -1) from 0 to 60 DAS (Fig. 3). In the
Waldviertel soil, Ca-P decreased (from 56 to 45 mg kg-1), Al-P (from 144 to 133 mg kg-1) and
Fe-P (from 68 to 59.5 mg kg-1) at 0 to 45 DAS (Fig. 3).

The percentage distribution of the inorganic and the available P pools are shown in
Fig. 4. The Bray2 and the Ca-P were the fractions depleted most by plants followed by the Fe-
P fractions in the two soils, and differences observed between the crops were not significant
All the soil P fractions (except Al-P that increased slightly) decreased from 0 to 60 DAS in
the two soils.

3.3. Isotopic exchange parameters

3.3.1. Total radioactivity and specific activities (SA) in soils

Total radioactivity in total P decreased from 1 to 42 DAS whereas all the other
fractions (available P and the inorganic fractions) increased in soils. For total P the total
radioactivity (dpm × 106) decreased from 746 at 1 DAS to 568 (Waldviertel) and 688
(Hungarian) at 42 DAS. The total radioactivity (dpm × 106) in the available P decreased from
82.1 (Waldviertel) in to 388.5 (Waldviertel) and 539 (Hungarian). A similar increase was

15
observed for Ca-P, Al-P and Ca-P. The variation in specific activity (kBq mg -1 P) in the
different soil P pools of the two soils at 1 and 42 DAS is shown in Fig. 5. In the Waldviertel
soil there was a sharp increase (6 times) in the SA in the Bray-P fraction and an increase (2
times) in the Fe-P fraction in the soybean. The SA of the soil fractions from maize and
soybean treatments was higher in the Hungarian than in the Waldviertel soil but the trend was
similar to that of the Waldviertel soil.

FIG. 2. Phosphorus concentration of shoot (a) and root (b) and plant P amount (c) for maize and
soybean grown in two soils. For legend see Fig. 1.

16
FIG. 3. Inorganic P pools (Fe-P, Al-P and Ca-P) and available-P extracted from the two soils at 0, 1,
5, 42 and 60 DAS.

FIG. 4. Percentage distribution of inorganic P pools (Fe-P, Al-P and Ca-P) and available-P (Bray
P2) extracted from the two soils at 0, 1, 5, 42 and 60 DAS.

3.3.2. Radioactivity in plants and L-values

The total radioactivity (dpm × 106) in shoot of maize was higher (108.0) than that in
soybean (76.6) irrespective of the soil used. The total radioactivity was higher in plants
grown in Waldviertel than in Hungarian soil and the values reflected the plant P uptake and
shoot biomass of soybean and maize (Fig. 6). However, the specific radioactivity (dpm × 103
mg-1 P or kBq × 103 mg-1 P) showed reversed trends to the values of total radioactivity, with
soybean recording higher values than maize. Percentage recovery of radioactivity in shoot
was higher for maize (14.5%) than for soybean (10.3%) and was four times higher in
Waldviertel (mean recovery 12.4%) than in the Hungarian soil (mean recovery 2.5%) (Fig. 6).
In the roots the recovery for maize was 7.2% (Waldviertel) and 4.6% (Hungarian), and for
soybean 5.0% (Waldviertel) and 1.0% (Hungarian), respectively.

17
FIG. 5. Total radioactivity and specific activities of maize and soybean grown on low- (Hungarian)
and medium-P (Waldviertel ) soils.

FIG. 6. Variations in specific activities in the different soil pools from 1 to 42 DAS in the two soils.

18
 To assess the amount of isotopically exchangeable P, the L-value was estimated using
the following equation:

L= (31P shoot – 31P seed) × 32P added to soil / (32P shoot)

where L is the L-value (µg P g -1 soil), the initial applied dose of 32P (Bq kg-1 soil), 32P shoot is
the activity of shoot mass (Bq g-1 DM), 31P shoot is the total amount of P in shoot biomass
(mg P g-1 DM).

The L-values (µg P g-1 soil) in maize and soybean were higher in the Waldviertel
(>70.0) than in Hungarian soil (<20.0). No significant differences in L-values were observed
for maize and soybean grown on the Walviertel soil, but for the Hungarian soil, the L-values
were higher for maize (20.0) than in soybean (9.6) suggesting that in this soil, maize was
more efficient to take up P than soybean (Fig. 7).

FIG. 7. Percentage P recovery in plant shoots and the L-values calculated for maize and soybean
grown on two soils.

19
4. DISCUSSION

4.1. Plant growth and P uptake

Maize and soybean grown in the Hungarian soil had low shoot P concentrations (<1
-1
mg g P) suggesting a strong P limitation. Plants showed symptoms of P deficiency 42 DAS
and produced three time less biomass than when grown in the Walviertel soil (Fig. 1). The P
concentration in shoot tissue of maize (1.1–1.5 mg g-1 P) was low compared to that in soybean
(1.8–2.2 mg g-1 P) and this is attributed to a dilution effect as biomass increased (see Fig. 2
and Fig. 3). However, in the low available P soil, total P accumulation in shoot was higher in
maize that in soybean suggesting that maize has a greater ability to take more P from low
available P soils than soybean. The Hungarian soil used had 21 mg kg -1 (Bray II) and 13.5
mg kg-1 Olsen) P, and for 1 kg soil, more than 12 mg P is expected to be available to the plant.
This suggests that the available P extraction using Bray II may contain other P that is not
easily available to maize and soybean. It is reported that Resin-P and NaHCO3-Pi are labile
soil fractions that are considered the most available for plant growth [18]. Other factors such
low pH (4.6) may have contributed to P fixation, and the low organic C (7 mg kg-1) may have
also soil. In addition, the fact that no nodules were observed on soybean grown in the
Hungarian soil despite the Rhizobium inoculation, while soybean grown on the Waldviertel
was well nodulated (0.6 mg DM plant-1) suggesting that nodule formation in the soybean was
severely impaired by P deficiency [18] no doubt contributing to the poor growth. Under low
P conditions, maize and soybean distributed more dry matter and P in roots than in shoots as
shown by the higher root/shoot ratio in the Hungarian than in the Waldviertel soil indicating
that, when P is limiting, the sink capacity for DM and P in roots is high compared to the
shoot.

4.2. Dynamics of P fractions

The available P measured by Bray P2 was double that in the Waldviertel compared to
the Hungarian soil, although no difference in Olsen P was observed. The fact that plants
(maize and soybean) showed severe deficiency when grown on the Hungarian soil, and the P
concentration in plant was below 1 mg g-1, suggests that Bray P2 overestimated the
available/labile P fractions in the soil. The relatively strong acidic reagents used in Bray P2
could dissolve a substantial amount of inorganic P during extraction of the soil [19].
Aluminium-P and Fe-P were the highest inorganic fractions extracted using the Sekiya
method, and in the Waldviertel soil the Al-P fraction was more than double the amount in the
Ca-P and Fe-P fractions at the beginning of the experiment. The Bray P2 and the Ca-P were
the fractions depleted by plants followed by the Fe-P fractions in the two soils, and
differences observed between the crops were not significant. Our results suggest that more P
was released in the soil solution (labile) from the available and Ca-P fractions for plant uptake
than from the Fe- and Al-P fractions. It is reported that the amount of Ca-bound P in soil is
more important for crop production; if Ca-bound P is present plants may be able to take up
more P to acidify the rhizosphere [20].

The principal aim of the direct chemical extraction method is to extract the labile
inorganic-P (Pi) in the soil (i.e. Pi fraction that can move into soil solution). However, the
available P extracted with a relatively strong acid (0.1N HCl + 0.03N NH4F) may extract more
than labile Pi and include non-labile or stable soil P, or may fail to extract labile Pi in certain
soil types [21]. For example, an acid extractant such as Bray-1 cannot be used on calcareous
soils because of neutralization. On the other hand, Bray-1 and Mehlich-3 extractants are
designed to extract P from non-calcareous soils or soils with a pH lower than 7.4 [22, 23]
while the Olsen test is preferred for calcareous soils. Available P, defined as the quantity of P

20
which will come into solution for plant uptake during its life cycle, is both time- and plant-
specific. Therefore, there is a high probability that the depletion of P from the Bray P2-
fraction may also include soil Ca- and Fe-P fractions. However, neither maize nor soybean
depleted a significant amount of P from the Al-P fractions (Fig. 5), although it has been
reported that some crops such as rice, groundnut and pigeon pea have the ability to take up P
from Fe- and Al-P fractions, possibly because of the release of carboxylic anions from roots
[8]. Al-P was the inorganic fraction least taken by cowpea genotypes [9].

Among the inorganic pools, Al-P, Fe-P and Ca-P are the three major fractions in the
soil. Our results imply that when P is not supplied, maize and soybean are able to access
mainly Fe- and Ca-P but not Al-P from the soil. Alternatively, the plant P uptake resulted in a
considerable decrease on Bray P2 with time, which would enhance dissolution of the
sparingly soluble P. This is supported by the fact that the changes over time in available P
were more sensitive than changes in total P [24].

4.3. Isotopically exchangeable parameters and P uptake by maize and soybean in low-
and medium-P soils

The high total radioactivity in maize compared to soybean suggests that plant P uptake
from soil was greater by maize than by soybean irrespective of the soil used. In addition,
whereas there was a 1.4 times increase in radioactivity in maize over soybean in the
Waldviertel soil, there was 4.3 times increase by maize over soybean. These data suggest that
maize could take up more P at low-available P conditions than soybean. This is supported by
the fact that the L-value (with seed P uptake correction factor) of maize was double that of
soybean in the low-P Hungarian soil, even though others [25] observed that determination of
E and L-values is not precise enough to identify plant species or cultivars able to take up P
from slowly or non-exchangeable P pools, or to quantify precisely the rate at which P in the
soil organic matter is mineralized.

In the low-P soil (Hungarian) both crops depleted more P from Bray P2 followed by
Fe- and Ca-P fractions. Addition of Fe-P resulted in a significant increase in shoot dry matter
yield (SDMY) whereas the addition of Al-P did not lead to an increased SDMY in cowpea
[11]. Wang et al. [26] showed the differential ability of cotton, wheat and white lupin to
utilize P fractions in the rhizosphere soil, where P utilized by cotton was mainly from
NaHCO3-Pi, whereas wheat and white lupin markedly depleted the HCl-Pi pool under P
deficient conditions [26]. Similar results have been reported for common bean and wheat
grown in mono-cropping and intercropping systems [7].

Maize and soybean grown in medium-P (Waldviertel) soil had lower specific
radioactivity (dpm × 103 mg-1 P or kBq x 103 mg-1 P ) in shoot than those grown on the low-P
Hungarian soil, and the values were lower in maize than in soybean. Low specific
radioactivity indicates that plants were using otherwise unavailable P sources. These results
therefore suggest that maize was more efficient in taking up P from otherwise sparingly
soluble inorganic-P sources than soybean. A high specific activity in soybean compared to
maize suggests that soybean is not efficient in taking up P from the sparingly soluble
inorganic pools. Furthermore, the percentage P recovery of radioactivity in shoots was higher
for maize (14.5%) than for soybean (10.3%), and was four times higher in Waldviertel (mean
recovery 12.4%) than in the Hungarian (mean recovery 2.5%) soil. In the roots the recovery
for maize was 7.2% (Waldviertel) and 4.6% (Hungarian), and for soybean 5.0% (Waldviertel)
and 1.0% (Hungarian), respectively. This confirms the results described above i.e. that maize
is more efficient in taking up P from the soil than soybean.

21
 Contrary to our expectation, the results indicated that plants could take up P more
efficiently from the other sparingly soluble P fractions in the Waldviertel (medium-P) than in
the Hungarian (low-P) soil. The specific radioactivities of Bray P2 and the Fe-P fractions
increased significantly whereas that of the Ca- and Al-P fractions decreased from 0 to 42
DAS. However, this could not explain precisely the uptake of P from slowly or non-
exchangeable P pools. Other factors such as the low N and C concentrations coupled with the
low pH (4.6) could have hampered the uptake of P (especially by soybean which is
susceptible to low pH) in the Hungarian soil. Other factors such as P concentration in the soil
solution (P intensity) may have been the limiting factor for plant growth. Rhizosphere
processes such as organic acid exudation may have caused sparingly soluble soil P to become
exchangeable, but without significantly renewing the P concentration in the soil solution
which is vital to raise P uptake. Thus adjusting the soil pH and raising the intensity through
the application of small quantities of P fertilizer could be essential to obtain adequate yields of
maize and soybean on the soils like the Hungarian example used here.

5. CONCLUSIONS

The main findings from this study is that maize was more efficient in taking up P from
otherwise sparingly soluble inorganic-P sources than soybean in the medium-P (Waldviertel)
soil; as indicated by the low specific radioactivity (dpm × 103 mg-1 P or kBq × 103 mg-1 P) in
shoots. The L-values (with seed P uptake correction factor) of maize was double that of
soybean in the low-P (Hungarian) soil suggesting the superiority of maize to access sparingly
soluble P from soils compared with soybean. When P is not supplied, maize and soybean are
able to access P mainly from the available P (Bray P2), Fe- and Ca-P sparingly soluble
fractions and not Al-P from the soil. Maize and soybean showed severe deficiency when
grown on the Hungarian soil and the P concentration in plant was below 1 mg g-1 suggesting
that Bray P2 overestimated the available/labile P fractions in the soil. Contrary to our
expectation, the results indicated that plants could take up P more efficiently from the other
sparingly soluble P fractions in the Waldviertel (medium-P) than in the Hungarian (low-P)
soil.

REFERENCES

[1] RAGHOTHAMA, K.G., KARTHIKEYAN, A.S., Phosphate acquisition, Plant Soil

274 (2005) 37–49.
[2] WICHMANN, W., World Fertilizer Use Manual, International Fertilizer Industry
Association (IFA), Paris (1992) 632 p.
[3] HIRADATE, S., et al., Strategies of plants to adapt to mineral stress in problem soils,
Adv. Agron. 96 (2007) 66–133.
[4] MIYASAKA, C., HABTE, M., Plant mechanisms and mycorrhizal symbioses to
increase phosphorus uptake efficiency, Commun. Soil Sci. Plant Anal. 32 (2001)
1101–1147.
[5] ISHIKAWA, S., Genotypic variability in phosphorus solubilizing activity of root
exudates by pigeonpea grown in low-nutrient environments, Plant Soil 245 (2002) 71–
81.
[6] NARUZZAMAN, M., Distribution of carboxylates and acid phosphatase and
depletion of different phosphorus fractions in the rhizosphere of a cereal and three
grain legumes, Plant Soil 281(2006) 109–120.
[7] AE, N., et al., Phosphorus uptake by pigeonpea and its role in cropping systems of the
Indian subcontinent, Science 248 (1990) 447–480.

22
[8] NWOKE, O.C., et al., Utilization of phosphorus from different sources by genotypes
of promiscuous soybean and cowpea in a low-phosphorus savanna soil, Afr. J. Agric.
Res. 2 (2007) 150–158.
[9] BUHLER, S., et al., Isotope methods for assessing plant available phosphorus in acid
tropical soils, Eur. J. Soil Sci. 54 (2003) 605–616.
[10] MCAULIFFE, C.D., Exchange reactions between phosphates and soils: Hydroxylic
surfaces of soil minerals, Proc. Soil Sci. Soc. Am. 12 (1947) 119–123.
[11] LARSEN, S., The use of P in studies on the uptake of phosphorus by plants, Plant Soil
4 (1952) 1–10.
[12] FRIED, M., DEAN, L.A., A concept concerning the measurement of available soil
nutrients, Soil Sci. 73 (1952) 263–271.
[13] HOCKING, P.J., et al., “A comparison of the ability of different crop species to access
poorly-available soil phosphorus”, Plant Nutrition for Sustainable Food Production
and Environment, (ANDO, P., et al., Eds), Kluwer, Tokyo (1998) 305–308.
[14] NACHIMUTHU, G., Isotopic tracing of phosphorus uptake in corn from 33P labelled
legume residues and 32P labelled fertilizers applied to a sandy loam soil, Plant Soil 314
(2009) 303–310.
[15] MURPHY, J., RILEY, J.P., A modified single solution method for the determination
of phosphorus in natural waters, Anal. Chim. Acta 27 (1962) 31–36.
[16] SEKIYA, K., “Phosphorus”, Methods of Soil Analysis, (BUNSEKIHOU, D.Y., Ed.),
Ministry of Agriculture, Forestry and Fisheries (1983) 225–257.
[17] TIESSEN, H., MOIR, J.O., “Characterisation of available P by sequential extraction”,
Soil Sampling and Methods of Analysis, (CARTER, M.R., Ed.), Lewis Publishers,
London (1993) 75–86.
[18] ADU-GYAMFI, J.J., et al., Phosphorus fractions in relation to growth in pigeon pea
(Cajanus cajan (L) Millsp.) at various levels of P supply, Soil Sci. Plant Nutr. 36
(1990) 531–543.
[19] CHIEN, S.H., “Soil testing for phosphate rock application”, Use of Phosphate Rocks
for Sustainable Agriculture, (ZAPATA, F., ROY, R.N., Eds), FAO Fertilizer and Plant
Nutrition Bulletin 13, FAO, Rome (2004) 148 p.
[20] VAN RAY, B., VAN DIEST, A., Utilization of phosphate from different sources by
six plant species, Plant Soil 51(1979) 577–589.
[21] SHARPLEY, A.N., Soil phosphorus extracted by iron-aluminium-oxide-
impregnated filter paper, Soil Sci. Soc. Am. J. 55 (1991) 1038–1041.
[22] BRAY, R.M., KURTZ, L., The determination of total, organic and available forms of
phosphorus in soils, Soil Sci. 59 (1945) 39–45.
[23] MEHLICH, A., Mehlich 3 soil test extractant: a modification of Mehlich 2
extractant, Commun. Soil Sci. Plant Anal. 15 (1984) 1409–1416.
[24] SONG, C., et al., Changes in phosphorus fractions, sorption and release in Udic
Mollisols under different ecosystems, Biol. Fertil. Soils 44 (2007) 37–47.
[25] FROSSARD, E., Soil isotopically exchangeable phosphorus. A comparison between E
and L-values, Soil Sci. Soc. Am. J. 58 (1994) 846–851.
[26] WANG, X., et al., Phosphorus acquisition characteristics of cotton (Gossypium
hirsutum L.), wheat (Triticum aestivum L.) and white lupin (Lupinus albus L.) under P
deficient conditions, Plant Soil 312 (2008) 117–128.

23
 CONTRIBUTION OF ROOT TRAITS TO PHOSPHORUS ACQUSITION
EFFICIENCY BY MAIZE LANDRACES IN ACID SOILS OF THE
HIGHLANDS IN CENTRAL MEXICO

J.S. BAYUELO-JIMÉNEZ, N. HERNÁNDEZ-BRAVO, M.L. MAGDALENO-

ARMAS, V.A. PÉREZ-DECELIS
Instituto Nacional de Investigaciones Agropecuarias y Forestales,
Universidad Michoacana de San Nicolás de Hidalgo,
Tarímbaro, Michoacan

M. GALLARDO-VALDÉZ
Instituto Nacional de Investigaciones Forestales, Agrícolas y Pecuarias,.
Campo Experimental Uruapan,

México

I. OCHOA
Unipalma S.A.
Bogotá D.C.,
Colombia

L. C. PAREDES-GUTIÉRREZ
Centro Nuclear Dr. Nabor Carrillo Flores,
Instituto Nacional de Investigaciones Nucleares,
Municipio de Ocoyoacac, Salazar,
México

J.P. LYNCH
Department of Horticulture,
The Pennsylvania State University,
University Park, PA, USA

Abstract

Plants have a wide range of mechanisms and morphological features that increase availability
and acquisition of orthophosphate from soil. Root growth, root branching, and root hair morphology
are important for the efficient acquisition of phosphorus (P). The series of studies reported here was
based on the hypothesis that Mexican maize landraces, which have developed mostly in environments
with low P availability and have a well-developed root system, could be a source of variation for the
improvement of phosphorus acquisition. Several studies were conducted to evaluate genotypic
variation in both root (root architecture and morphology, including root hairs) and plant growth traits
associated with P acquisition efficiency (PAE) and/or P utilization efficiency (PUE) of maize
landraces in a P-deficient Andisol in the Central Mexican Highlands, and to identify genotypic
differences, among both efficient and inefficient in P acquisition and responsive and non-responsive
maize landraces to applied P. The results showed that accessions differed greatly in plant growth,
grain yield, root morphology, total uptake of P, PAE, PUE, and P efficiency defined as growth with
suboptimal P availability. Phosphorus-efficient accessions had not only greater biomass per unit of
absorbed P, but also larger root systems, greater P uptake per unit root weight, more nodal roots, nodal
root laterals, and greater root hair density of nodal root main axes and first-order laterals than did P-
inefficient accessions under P deficiency. Root biomass allocation, as quantified by the allometric
partitioning coefficient (K) was not altered by P availability in the efficient accessions, but inefficient
accessions had less biomass partitioning to roots (i.e. a lower K) under low P conditions. Accessions
with enhanced nodal rooting and laterals had greater P uptake and growth under low P. Dense root

25
hairs on nodal root main axes and first-order laterals conferred a marked benefit under low P, as
evidenced by increased plant biomass and grain yield. Late maturity accessions had improved growth
and yield under low P. Accessions DP x Tromba, HV313 x DEM, Macho III-04, and CIMMYT-1,
were categorized as the most P efficient under low P and as the most responsive to increased P
availability on a P-deficient acidic soil of this region. P-efficient accessions such as DP x Tromba
accessed soil P not available to P-inefficient accessions. These results indicate that landraces of the
Central Mexican highlands exhibit variation for several root traits that may be useful for genetic
improvement of P acquisition efficiency in maize.

1. INTRODUCTION

Low soil fertility, especially suboptimal phosphorus availability is a principal,

pervasive constraint to maize production in the Purhepecha Region [1]. Application of lime
and P-containing fertilizers is usually the recommended treatment for enhancing soil P
availability and stimulating crop yields. However, in this region, few maize producers can
afford intensive chemical inputs. Genetic improvement of phosphorus efficiency, defined as
yield ability at low phosphorus supply and better response to phosphorus inputs, is therefore
an attractive prospect for the productivity of low-input systems [2]. Several genetic traits have
been identified with potential for breeding phosphorus efficient crops, including root
exudates, root-hair traits, cortical aerenchyma, topsoil foraging through basal and adventitious
rooting [3].

Local maize germplasm have already been identified and evaluated for root traits
conferring phosphorus acquisition efficiency, and specific traits responsible for P uptake
efficiency among local landraces under low P environments have been characterized. Several
root traits with potential in breeding phosphorus efficient maize include root hair traits and
topsoil foraging through seminal, nodal, and lateral rooting. Therefore, it seems possible to
screen a large number of maize accessions under field conditions for those root traits adapted
to low P efficiency at different plant growth stages.

The present sequence of studies aimed to (a) assess the genetic variability of local
genotypes for root traits that enhance phosphorus acquisition in a low P soil, (b) identify and
validate phenotypic traits conferring P acquisition efficiency in maize, and (c) employ
validated screening protocols for use of 32P isotope techniques to identify genotypes selected
for superior phosphorus acquisition and /or utilization.

2. ASSESS THE GENETIC VARIABILITY OF LOCAL GENOTYPES FOR ROOT

TRAITS THAT ENHANCE PHOSPHORUS ACQUISITION IN A LOW P SOIL 1

2.1. Introduction

Suboptimal soil phosphorus (P) availability is a principal yield-limiting factor for

maize in many areas of the developing world, affecting at least eight million hectares of maize
production globally [4]. Few unfertilized soils have adequate phosphorus availability for
intensive crop production [5]. Low response to P fertilization and low P fertilizer recovery are
the main obstacles to increased P efficiency in cereals, especially on P-adsorbing, acid soils,
where applied P fertilizers often have poor efficacy [6].

1
Full details of this study have been published as BAYUELO-JIMÉNEZ, J.S., et al., Genetic variation for root
traits of maize (Zea mays L.) from Purhépecha Plateau, under contrasting phosphorus availability, Field Crops
Res. 121 (2011) 350–362.

26
 Plants display a variety of adaptations to low phosphorus availability, including
changes in root morphology and architecture [3, 7, 8], increased production and secretion of
P-mobilizing root exudates [9], increased proliferation and elongation of root hairs [10],
modification of carbon metabolism and alternative respiratory pathways and enhanced
expression of Pi transporters [11]. Phosphorus is relatively immobile in soil and its availability
is typically greater in topsoil and declines substantially with depth [3]. Root architectural
traits that enhance topsoil foraging such as shallower growth of basal roots, adventitious
rooting and greater dispersion of lateral roots [3, 12, 13], and root hair formation [10, 14] may
therefore enhance phosphorus acquisition in low phosphorus environments. Genotypic
variation in adventitious root formation has been observed in several crops including common
bean (Phaseolus vulgaris L.) [12] and maize (Zea mays L.) [15]. In common bean phosphorus
availability regulates adventitious rooting [12] and two major quantitative trait loci (QTL)
accounted for 19 to 61% of the total phenotypic variation for adventitious root traits under
low phosphorus conditions [16].

Root hairs are particularly important for phosphorus acquisition. Considerable genetic
variation in root hair length is correlated with phosphorus acquisition among genotypes of
common bean (P. vulgaris L.) lentil (Lens culinaris L.) [17], wheat (Triticum aestivum L.),
and barley (Hordeum vulgare L.) [18].

The deployment of root architectural traits in plant breeding programs has great
potential to alleviate P deficiency, a primary constraint to crop production in world agriculture
[2]. Evaluation of P efficient germplasm among existing Mexican landraces is of interest in
this regard since Mexico is the global centre of maize genetic diversity [19]. In particular,
landraces of Michoacan State are well adapted to low P environments and may possess traits
not common in elite germplasm. One of the most important traditional maize growing areas in
this region is the Purhepecha Plateau [1]. In this area, it is common to find both synthetic
hybrids and hybrids that have gone through creolization, a process by which improved
varieties are exposed to farmer management, seed selection, and hybridization with landraces
[20]. Over 60 percent of the total arable land is phosphorus-deficient in this region [1]. The
objective of this study was to evaluate landraces of the Purhepecha Plateau for adaptation to
low P soils, and their expression of root traits that could be important in P acquisition.

2.2. Materials and methods

2.2.1. Cigar roll culture system

A growth chamber screening- cigar roll method was conducted to identify phenotypic
traits conferring phosphorus efficiency in 108 maize genotypes at the emergence stage (VE)
in 2007. Maize seeds were surface sterilized for 1 min in 0.5% solution of NaOCl and then
washed in deionized H2O before germination. In each replicate, four typical seeds from the
seed stock for accession were selected and wrapped in brown germination paper (Anchor
Paper, St. Paul, MN, USA) as a cigar roll. Two batches of 108 cigar rolls were soaked
vertically in two plastic containers filled with 5 l nutrient solution with low or high
phosphorus. The low and high phosphorus treatments were amended with KH2PO4, at 1 µM
(LP) and 1 mM (HP), respectively, in nutrient solution. Seedlings were germinated in
darkness at 28 ± 1 oC in a growth chamber for 3 days, then grown under a photoperiod of 14 /
10 h at 28 / 22 oC (light/darkness) with photosynthetically active radiation (PAR) of 200 µmol
photons m-2 s-1 at the soil level. The relative humidity was 65%. The nutrient solution pH was
adjusted to 6.0 daily. Seedlings were grown in a growth chamber for healthy root system
growth for up to 10 d. Plant seedlings were harvested 12 days after germination and the roots
were preserved in 30% ethanol. At the time of the harvest, the root system consisted of a

27
primary root with emerged lateral roots, and 0–6 seminal roots. Seminal roots were dissected
from seedlings. Seminal root length, seminal root number and root hairs were visually
evaluated [21]. The experiment was a randomized complete block design with a split-plot
arrangement of treatments. The main plots were low- and high-P; subplots were 108
accessions (3 maturity types and 4 breeding groups). There were four replicates staggered in
time. For each replicate, there were four plants.

2.2.2. Field studies

Two experiments were conducted under low and high P fertilization and rain-fed
conditions, in farmers’ fields in Pontzomaran and Bonilla, in the central highlands of
Michoacan, Mexico. Ponzomaran is located at 19o 24´ N, 101o 38´ W, 2400 masl; 800–1000
mm precipitation during the cropping season. Bonilla is located 19o 39´ N, 101o 01´ W, 2400
masl; 900–1100 mm precipitation. Soils of the study sites were vitrands (Table 1).

Experiments were arranged in a randomized complete block design with four

replications in a split plot arrangement of treatments where P level was the main plot and
accessions the sub-plots. Each experimental unit (plot) consisted of five 5-m long rows for
each accession. Seeds were sown at 6 cm depth, 25 cm spacing between plants, and 60 cm
spacing between rows. Experiments were conducted in P-deprived soils, because during the
four previous maize growing seasons no fertilizers were applied. The low (LP) and high (HP)
phosphorus treatments consisted of 23 kg P2O5 ha-1 and 97 kg P2O5 ha-1, applied as calcium
super phosphate, at seeding. All plots were additionally supplemented with 60 kg−N ha-1as
urea at seeding and 60 kg−N ha-1 at silking. Maize accessions were seeded within the optimum
sowing dates (around April 17–23, after the beginning of 2007 and 2008 rainy season). Weeds
were controlled with bromoxynil (3, 5-dibromo-4-hydroxy-benzonitrile) at 1 l ha -1 applied 20
days after emergence and also by manual cultivation.

2.2.3. Plant material

Two hundred and forty two local maize accessions were grown in 2007 (Table 2). All
maize accessions are originally from the Purhepecha region, which have been recently used
by the Instituto Nacional de Investigaciones Forestales, Agrícolas y Pecuarias (INIFAP)
maize breeding programme, and they differed in both yield, P responses to fertilization, stalk
lodging susceptibility and grain type (floury to flint corn). The maize accessions were
represented by three maturity types according to the number of days to silking: Early (E) 75–
85 d, Intermediate (I) 85–95 d, and Late (L) 95–105 d, and four breeding groups within each
maturity type: Landraces (C), Advanced landraces (AC), Hybrids x landraces (HxC), and
Synthetic hybrids (S). Landraces in advanced generation are advanced breeding lines of
landrace types collected in low P soils from Purhepecha Plateau and crossed with advanced
generation of commercial hybrids. Contrasting accessions (P-efficient and P inefficient ones)
were selected from the 242 maize accessions and planted in the succeeding crop cycle,
Experiment 1 in Ponzomaran (2007) (n = 50 accessions) and Experiment 2 in Bonilla (2008)
(n = 50), which allowed a comparison across locations (Table 2).

2.2.4. Plant measurements

Root crowns were excavated at the same growth stage by removing a cylinder of 30 to
40 cm soil depth and 50 cm from the shoot base (4 replicates) between 60 and 69 days after
planting (DAP), which corresponded to the late vegetative growth stage (V8). The excavated
root crowns were then shaken briefly to remove large fractions of the soil adhering to the root
crown. The root crowns were immersed in water with P-free soap for about 8 min in order to

28
facilitate soil removal. After washing, the clean root crowns were stored in 25% ethanol at 4o
C. Root architectural measurements were taken at the laboratory. The root system was
characterized by four root types: the embryonic primary root, seminal roots emerging from
the scutellar node, shoot-borne crown (nodal) and brace roots [22].

TABLE 1. PROPERTIES OF THE TOPSOIL (0–20 CM) OF THE ANDISOL IN PONZOMARAN,

SAN JUAN TUMBIO, BONILLA, CHARAHUEN, AND LADERAS LOCATIONS, MICHOACAN,
MEXICO

Parameter Pontzomaran Juan Bonilla Charahuén Laderas

Tumbio
Environmental variables
Latitude (N) 19o 24´ 19o 31´ 19o 30´ 19o 39´ 19o 30´
Longitude (W) 101o 38´ 101o 36´ 101o 41´ 101o 41´ 101o 41´
Altitude (m.a.s.l.) 2280 2140 2240 2340 2240
Rainfall (mm) 800–1000 700–1400 900–1100 800–1100 900–1100
Temperature (oC) 5.4–24.1 14–20 14–20 14–22 15–20
Climate Temperate sub-humid Cw Cw Cw Cw Cw

Soil classification
Soil Survey Staff Vitric Vitric Vitric Vitric Vitric
Andisol Andisol Andisol Andisol Andisol
FAO-ISRIC-SICS Eutric Eutric Eutric Eutric Eutric
Haplaund Haplaund Haplaund Haplaund Haplaund

Soil characteristics
Sand (%) 38.6 48.1 55.1 53.7 59.7
Clay (%) 38.5 18.9 11.6 13.6 7.6
Silt (%) 22.9 36.0 33.3 32.7 32.7
Apparent density (g cm-3) 0.86 1.08 0.89 1.01 0.85
pH 5.5 6.1 6.1 6.1 6.0
Organic matter (%) 7.9 6.2 4.2 3.7 4.1
Cation exchange capacity 18.6 14.7 15.3 16.2 16.2
(cmol kg-1)
Exchangeable aluminium 0.09 0.04 0.03 0.12 0.04
(cmol kg-1)
K (mg kg-1) 131 444.7 454 644 469
Ca (mg kg-1) 1225 1450 2254 1583 1860
Mg (mg kg-1) 84 248 394 375 324
Fe (mg kg-1) 40.8 33.5 37.6 57.1 37.1
Mn (mg kg-1) 26.8 23.2 20.7 65.8 9.56
Inorganic N (mg kg-1) 35 27 17.4 18.9 57.1
Available phosphorus Bray 1 1.20 4.75 2.74 3.48 6.04
(mg kg-1)
Low P 1.07 3.06 1.39 3.20 –
High P 3.84 8.69 2.24 9.7 –

System studied Zea mays L.

Production cycle (Spring- Rain-fed Rain-fed Rain-fed Rain-fed Pot
Summer)/system
Fertilization: Nitrogen (Urea)† 120 120 120 120 60
Fertilization: 23 23 23 23 23
Phosphorus (P2 O5)† 97 97 97 97 50
†Unit is kg ha-1

29
TABLE 2. LANDRACES FROM THE PURHEPECHA PLATEAU, MICHOACAN, MEXICO
(EXPERIMENTS 2007, 2008, 2009, 2010)

Maize types (Breeding groups, BG) Maturity #

Early 78 Intermediate 102 Late 62
Landrace (C) † 32 ( 9) (2) (1) c,d,e 39 (19) (9) (3) 36 (12) (6) (3)
Advanced Landrace (AC) ‡ 21 (10) (4) (2) 46 (11) (4) (1) 17 (9) (7) (3)
Hybrids x Landrace (H x C) 18 (8) (4) (2) 10 (7) 6 (2) (4) (1)
Synthetics (S) (checks) 7 (6) (6) (2) 7 (3) (3) (1) 3 (2) (1)
Racial classification Conico, Purhepecha Purhepecha
Chalqueño Chalqueño Chalqueño
Purhepecha
Kernel White, Purple, White, Purple, White, Yellow,
Yellow, Red Yellow, Red Red
Days to flowering 75–85 85–95 95–105
Collected altitude (m.a.s.l.) 2500–2600 2200–2400 2200–2400
Highland Western region Eastern region
†Native population (landrace).
‡Advanced generation of landrace.
#Numbers in parenthesis indicate the common accessions included in Experiment 1 (Cigar rolls), Experiment 2
(2008) and Experiments 1, 2, and 3 (2009 and 2010), respectively.

Eight root traits were visually scored: seminal root length and number, seminal root
branching and angle; nodal root length and number, nodal root branching and angle. A rating
scale of 1–4 was used to rank crown root architecture, particularly for root branching and root
angle where 1 = first order lateral branching and 4 = multiple lateral branches with up to 4
orders of branching. For nodal root angle, one indicates shallow root angles (1 = 0–22.5o); 2 =
22.5–45o; 3 = 45–67.5o, and four indicates steep root angles (4 = 67.5–90o).

Before root hair analyses, nodal roots were immersed in an ultrasound bath for 5–10
min in order to remove remaining soil particles without damaging root hairs. For root hair
evaluation, root fragments were dyed in 0.05% trypan blue, and placed in a Petri dish with
deionized water. Root hairs were visually evaluated using a rating scale of 1–9 to rank the
density/length as follow: 1 = no root hairs; 3 = low root hair density/length; 5 = intermediate
root hair density/length, 7 = between 5 and 9 rating scale; 9 = abundant root hairs [21]. Eight
characters were measured for root hairs of nodal roots; root hair density and root hair length
from the basal and middle region of nodal root main axis and root hair density and length
from the basal and middle region of nodal root first order laterals, using a dissecting
microscope (Zeiss STEMI DV4, Göttingen, Germany) at 30x magnification.

One sample per replication per maize accession were harvested between 60 and 69
DAP, which correspond to the vegetative growth stage V8. Shoot and root biomass was
determined after drying at 60 oC to constant weight. Grain yield, adjusted to 10% moisture,
was determined from the total harvest area. Relative biomass allocation to roots was
determined by the allometric partitioning coefficient (K) derived from a series of paired
measurements of root dry weight and shoot dry weight by linear regression of the form: K =
(log R – log b)/ log S where R is root dry weight (g), S is shoot dry weight (g), b is a constant,
and K is the allometric coefficient [23].

2.2.5. Statistical analyses

Before analyses of variance, data of mean values of each genotype for each variable
were subjected to test for heterogeneous error variances using Bartlett´s Test [24]. Statistical

30
differences among maturity groups were ascertained from the SAS Generalized Linear Model
Procedure [24]. A Protected Least Significant Difference (PLSD) was constructed when the
F-tests indicated statistically significant differences among treatment effects (P ≤ 0.05).

Phosphorus efficiency of maize accessions was determined by phosphorus efficiency

index (PEI) [25] and assessed using principal component analysis (PCA) of standardized
values of plant growth and grain yield parameters at low P, and relative values at low P to
those obtained under high P supply. PCA was computed by means of SAS FACTOR
procedure with the PRIN option and VARIMAX method for orthogonal rotation [24] on all
242 maize accessions. The relative weight of each principal component was weighted by the
corresponding contribution rate accounting for variation of all growth traits. Consequently,
PEI values of different accessions were calculated according to the retained principal
component and their relative weigh, namely PEI = 242∑i=1 PCi x RWi. The criterion used for
classification of maize accessions was determined by the method of cluster analysis and maize
accessions were divided into 3 categories according to the P efficiency index (Fig. 1), and 4
categories according to P efficiency index in combination with growth potentials (shoot dry
weight or grain yield at high P) (Fig. 2). A distance matrix was produced with the
standardized data with the dissimilarity Euclidian distance coefficient and a Ward´s minimum
variance clustering method was performed, rendering a dendrogram through the SYSTAT
software [26].

2.3. Results and discussion

2.3.1. Seedling growth stage (VE): Phosphorus efficiency and correlation with root traits

Phosphorus efficiency was assessed using principal component analysis of 4 growth

parameters of the 108 accessions. The 4 parameters at low P along with 4 indexes at low P
relative to high P differed significantly among accessions. Three principal components (PCs)
of each accession with eigen values greater than one were retained whose cumulative
contribution was 97%. The relative weight of each principal component was weighted by the
corresponding contribution rate accounting for variation of all growth traits. All accessions
were grouped into 3 clusters: cluster 1 (PEI < –0.85) with 6 inefficient accessions, cluster 2 (–
0.85 < PEI < –0.10) with 53 moderately efficient accessions and cluster 3 (PEI > 0.22) with
49 efficient accessions (data not shown). The responses of the parameters of the 108
accessions into three phosphorus efficient groups indicates that the root and shoot dry weights
and root to shoot ratio increased with increasing P efficiency at low P (76%, 33%, 32%,
respectively). The efficiency groups also differed in the pattern of seminal roots and root hair
formation. Significant variation existed in all three phosphorus efficiency groups, though
variation was least in the lowest efficiency group. Accessions of the phosphorus efficient
group increased seminal root length under LP environments (35%), whereas root hair density
(45%) and length (34%) of seminal roots were significantly greater in LP than in HP. These
results suggest an adaptive benefit of increased seminal root length for the uptake of immobile
nutrients such as P by increasing soil exploration [8].

The results of correlation analysis between PEI and root traits (Table 3) indicate that P
efficiency was positively correlated with dry weights of roots and shoots at LP and HP,
suggesting that screening for P efficiency should be conducted under P deficiency or under P
deficiency along with a P-adequate control. Moreover, shoot dry weight should be used as an
ideal parameter at the seedling stage for breeding P-efficient accessions of maize. PEI was
also correlated with root architectural traits including seminal root length and tap root length
at low P (Table 3). In addition, PEI was correlated positively with basal root hair density and

31
length of seminal main root and tap root axes at low P (Table 3). Therefore, it was indicated
that the P-efficient accessions facilitated biomass accumulation, root growth, and P uptake
under P deficiency, generally consisted with previous reports in maize [29].

TABLE 3. PEARSON CORRELATION COEFFICIENT AMONG PHOSPHORUS EFFICIENCY

INDEX (PEI) AND ROOT TRAITS OF MAIZE ACCESSIONS GROWN IN A P-DEFICIENT
SOLUTION WITH LOW P (1 µM, LP) OR WITH HIGH P ADDITION (1 MM, HP) AT THE
SEEDLING GROWTH STAGE

Variables P Correlation coefficient (r)

Level Dry weight (mg plant-1)
Root Shoot
HP LP HP LP
PEI 0.23 0.88*** 0.33* 0.55***
Tap root length (cm) HP 0.35** 0.57*** 0.56*** 0.59*** 0.51***
LP 0.43** 0.49*** 0.55*** 0.43*** 0.49***
Seminal root length HP 0.30* 0.62*** 0.55*** 0.52*** 0.49***
(cm) LP 0.39** 0.49*** 0.53*** 0.43*** 0.50***
Seminal root hair HP –0.14 0.06 –0.06 –0.01 –0.13
density LP 0.19 0.27* 0.26* 0.24* 0.34**
Tap root hair density HP 0.34** 0.49** 0.51*** 0.48*** 0.42**
LP 0.27* 0.31* 0.32* 0.31** 0.41**
Seminal root hair length HP 0.26* 0.38** 0.40** 0.32** 0.25*
LP 0.24* 0.21 0.26* 0.16 0.32**
* , **, ***, Significant at P<0.05, P<0.01, P<0.001.

2.3.2. Late vegetative growth stage (V8): Phosphorus efficiency and P responsiveness

We observed substantial variation among maize landraces from the Central Mexican
highlands for growth in low P soil. Accessions could be grouped into 3 categories of P
efficiency based on growth and grain yield parameters at low P and their relative values to
those at high P (Fig. 1) and 4 categories according to P efficiency (PEI) in combination with P
responsiveness (shoot dry weight or grain yield at high P) (Fig. 2). This study indicated that
41 accessions of Exp. 1 (n = 242) and 14 accessions across locations (n = 50) had the lowest
growth and yield reduction and the highest levels of P efficiency (PEI > 0.69 and PEI> 0.56,
respectively) under low P. Among common accessions, ZR-SGC (75), CB-CMP (78),
Coangantzio (119), DP X Tromba (127), Cruz Gorda (140), San Gregorio (144), Macho III-
05 (180), Macho III-05 (182), Cimmyt-1 (185), CBVA-AS-1 (198), CCHEDE (199), Macho-
II 04 (237), Macho IV-03 (239) and Macho IV-05 (241) had consistently higher PEI in LP
(Table 4).

When the combination of PEI with P responsiveness at high P is considered,

accessions CB-GMP (57), CB-RSL (63), Santa Clara (115), Camémbaro (118), HV-313 X
DE (135), Macho III-04 (181), Macho III-05 (182), Macho I-03 (233), Macho-II 04 (237) and
Macho IV-05 (241) were the best accessions for P-deficient soil of this region (Fig. 2). These
accessions were categorized as the most P efficient and as the most responsive to increased P
availability. Applications of 97 kg P2O5 ha-1 (equivalent to 41.7 kg P ha-1) increased shoot
biomass and grain yield of the accessions. However, the increase was nearly equal to the
difference between accessions in low P soil (23 kg P2O5 ha-1 equivalent to 9.9 kg P ha-1) (Fig.
3). Thus, in low P soil, the P-efficient accessions seem to produce the same amount of dry

32
matter and grain yield as accessions with 97 kg P2O5 ha-1 less P fertilizer applications. These
results demonstrate the added advantage of selection and breeding for P efficiency for
maintaining productivity in low P soils.

Cluster I (5-240), 98 accession

Cluster II (3-241), 103 accessions
Cluster III (17-239), 41 accessions

40

Sum of Squares, E
30
20
10
I II III
II

0
5 1 91 54 59 6 13 3 31 17 160 18 56 17 16 17
1 2 93 55 60 7 5 12 32 0 163 0 74 14 5 2
9 8 97 57 62 9 24 14 39 17 181 18 75 8 17 20
2 15 101 58 64 34 0 20 14 3 182 8 89 15 8 5
3 16 102 61 65 40 10 24 3 17 203 22 82 0 19 20
2 18 104 63 79 42 8 27 14 6 204 1 11 18 8 6
5 26 106 6 83 45 10 35 5 17 207 22 2 3 20 20
4 28 130 68 85 49 9 44 18 7 211 2 11 18 2 8
6 29 152 71 86 51 11 47 7 17 212 22 3 4 21 20
5 33 154 73 87 53 0 76 19 9 231 4 11 18 4 9
2 50 156 77 98 105 11 95 0 19 242 22 4 5 21 21
6 138 157 80 99 131 1 13 19 1 48 5 11 18 5 0
6 147 174 84 10 139 11 3 2 19 72 22 5 6 21 23
7 238 217 10 0 144 6 13 20 6 78 6 69 18 9 0
0 7 10 12 4 0 19 158 22 11 9 22 38
8 3 0 13 20 7 167 7 9 19 3 41
6 1 21 175 22 12
14 23 3 8 6
FIG. 1. Clustering of 242 accessions for P efficiency by the Ward cluster method according to PEI of
each accession, where PEI is the parameter for assessing P efficiency obtained from principal
component analysis. Cluster I-III represents high P efficiency, moderate P efficiency and low P
efficiency, respectively. Underlined entries represent the 50 accessions selected from the 242
accessions and planted in the subsequent crop cycle.

33
 4
203
NER ER
133

2 34 227 230
206 237
124 234 185
239 219 214
138 135 27 216 6 119
165 118 205
5 215 150
Shoot dry weight (standardized value of shoot
208 209
240 225 241 235 233232
120 143 20 180 198 189
238 62
0
115127 181 182 199 210 41 184
57 197
96 63 111 75 125 161 195
117
85 102 78 162
83 109 113 140 172
99 38
65 53 114 171
-2 55 186

NENR ENR
-4
-1.5 -1.0 -0.5 0.0 0.5 1.0 1.5

4
185

2 6 241
118
240 237
234
91 115 235 144
138 106
124 20 182
238 119
0 239
227 215 216
85 198
140 75
117
109 62 83 78 127
199
99 113
-2

-4
-1.5 -1.0 -0.5 0.0 0.5 1.0 1.5

The scores of phosphorus efficiency

FIG. 2. Division of charts of 242 and 50 common accessions according to P efficiency and
standardized value of shoot dry weight under high P conditions. Phosphorus efficiency is expressed as
PEI, which is an assessment index calculated from principal component analysis. Standardized values
of shoot dry weight are estimated as the following function: = ( − )⁄ . Categories
represented by efficient and responsive (ER), non-efficient and responsive (NER), non-efficient and
non-responsive (NENR), and efficient and non-responsive (ENR). Accession numbers are indicated.

A common finding in many screens for P efficiency is that traditional varieties or

landraces have greater P-use efficiency than modern cultivars [6, 27]. This is often attributed
to the impact of breeding programs being performed under nutrient replete conditions [28].
Moreover, it is considered that while such programs have produced cultivars with large
harvest indexes, which are also relatively internally P-efficient, further gains in P-limited
yield are most likely to be achieved by improving the external P efficiency [27]. A higher P
efficiency in plants can be achieved by improving P acquisition and/or internal P utilization
[28]. Late P-efficient accessions ZR-SGC (75), Coangantzio (119), DP X Tromba (127),
Macho IV-03 (239), and Macho IV-05 (241) had an excellent relationship between grain yield
and shoot biomass (r2 = 0.83) under low P conditions, suggesting that P utilization through

34
better dry matter accumulation and partitioning of above-ground dry matter to grain is the
most critical factor in expression of high P efficiency. Alternatively, other late P efficient
accessions SC-JC-11-Mz4 (38), PM-TA-16 Mz21 (41), Camémbaro CRM-00-05 (118), H-
791 99-04-05 (125), and Macho IV-03 (239) can be selected showing a significant
relationship between shoot biomass and nodal root number (r2 = 0.56) and nodal root
branching (r2 = 0.60) or between grain yield and nodal root branching (r2 = 0.54) and nodal
root length (r2 = 0.46) suggesting that P acquisition ability of genotypes is a decisive factor in
expression of high P efficiency. It seems likely that P efficiency mechanisms may be different
among the accessions of a given species.

2.3.3. Mechanisms involved in P efficiency: root growth and allocation of dry matter

Genetic differences in P efficiency were associated with variation for root growth and
root architecture. Accessions that were P efficient (less shoot biomass depression under low
P) maintained a higher root to shoot ratio under low P. Phosphorus efficient accessions had
greater biomass allocation to roots than inefficient accessions (Fig. 3), and better maintenance
of biomass allocation to roots under P stress (r2 = 0.01 to r2 = 0.04). Increased biomass
allocation to root growth is beneficial for P acquisition, since P is relatively immobile in soil,
but may slow overall plant growth because of the increased respiratory burden of root tissue
[7]. Plants with low rates of root respiration may be able to maintain greater root biomass than
inefficient genotypes without increasing overall root carbon costs.

2.3.4. Mechanism involved in P efficiency: root architecture and morphology

The response of axile roots (seminal and nodal) and the length and number of lateral
roots to P stress varies substantially among maize genotypes [29]. Genotypes with increased
or sustained elongation of axile roots and lateral root development under P deficiency had
superior ability to acquire P and maintain growth (r2 = 0.50 to r2 = 0.60). In this crop, a large
proportion of the mature root system consists of nodal roots, so the maintenance of nodal root
formation, when overall growth is inhibited by P-deficiency, could result in an increased
proportion of root length in the adventitious root system. In this study, we confirm that
enhanced nodal rooting and greater nodal branching (nodal root laterals) is indeed important
for plant adaptation to low phosphorus in maize (Fig. 3).

Nodal rooting was significantly correlated with plant growth in the field among
contrasting accessions and across locations (Table 5). Efficient accessions with greater nodal
rooting and lateral branching at low P had greater biomass then did inefficient accessions with
reduced nodal root formation and lateral branching (Fig. 3). The greater nodal rooting of P
efficient accessions under low P could be explained by the greater overall biomass and the
weak allometric relationship of plant biomass with root biomass (r2 = 0.04). Therefore, if
nodal roots in the topsoil are advantageous for acquiring phosphorus under limited P
availability, as suggested by [1], this weak allometric relationship would facilitate the
selection of efficient accessions with high nodal rooting in the field. Genotypic differences in
the specific root length and P concentration of the lateral roots of maize has been associated
with reduced biomass and P investment to the extension to lateral roots [29]. Therefore, these
results suggest that enhanced lateral rooting under P stress may be harnessed as a useful trait
for the selection and breeding of more P-efficient maize genotypes.

35
 6 ** ** ** 30
HP LP

length (cm)
Nodal root
weight (g plant 4 20
Root dry
-1
)
2 10

0 0
50 ** **
40

Nodal root
weight (g plant

number
Shoot dry

25 20
-1
)

0 0

Nodal root branching

0.16 2.4
Root to shoot

(Scale 1-4) 1.6 **

ratio

0.08
0.8

0.00 0.0
120 7
density (Scale 1-
Basal root hair
weight (g

*** ***
plant -1)

80 5
Seed

9)

40 3

0 1
5 7
length (Scale 1-

*** ***
Basal root hair
Yield (t ha -1)

5
3
3

0 1
1 2 3 1 2 3
FIG. 3. Variation of ten parameters of maize plants under low P and high P conditions with low,
medium and high P efficiency. Phosphorus efficiency was ranked according to principal component
analysis and cluster analysis of 50 common accessions. 1 to 3 representing P efficiency from low to
high. Bars are the standard error of means of 14, 22, and 14 accessions, respectively. Within the P
efficient group, bars with *, **, and *** indicate significance at P<0.05, P<0.01, P<0.001,
respectively.

36
 TABLE 4. SELECTION OF 50 MAIZE ACCESSIONS ACCORDING TO P EFFICIENCY INDEX (PEI) IN PONZOMARAN (EXPERIMENT 1, 2007),
BONILLA (EXPERIMENT, 2008), AND COMBINED EXPERIMENTS

ID Accession M† BG‡ Experiment 1 (2007) Experiment 2 (2008) Experiments 1 + 2

ID PEI# Cluster§ ID PEI Cluster ID PEI Cluster
118 Camémbaro CRM-00-05 L AC 118 1.02 E 237 0.97 E 237 1.08 E
185 CIMMYT-1 M-00-04-05 E HxC 185 0.70 E 127 0.93 E 199 0.89 E
182 Macho-III-05 I S 182 0.68 E 198 0.88 E 198 0.76 E
119 Coangatzio CR-03-04-05 L AC 119 0.59 E 199 0.85 E 239 0.64 E
215 Zacapu Cr-03-05 I AC 215 0.57 E 239 0.78 E 127 0.55 E
180 Macho-III-03 I S 180 0.56 E 115 0.54 E 75 0.52 E
239 Macho-IV-03 L S 239 0.56 E 216 0.53 E 241 0.52 E
235 Macho-I-05 E S 235 0.53 E 84 0.53 E 182 0.48 E
181 Macho-III-04 I S 181 0.51 E 83 0.48 E 144 0.48 E
237 Macho-II-04 E S 237 0.49 E 65 0.45 E 119 0.45 E
233 Macho-I-03 E S 233 0.47 E 135 0.43 E 235 0.40 E
199 CCHEDE CM-00-04-05 E AC 199 0.47 E 75 0.35 ME 140 0.40 E
241 Macho-IV-05 L S 241 0.45 E 140 0.31 ME 78 0.33 E
198 CBVA-AS-1-P CM-01-04-05 E HxC 198 0.43 E 79 0.28 ME 185 0.32 E
6 SHUI-MRC-2 L C 6 0.27 ME 218 0.27 ME 216 0.23 ME
140 Cruz Gorda CrM-00-04-05 E AC 140 0.20 ME 111 0.23 ME 115 0.22 ME
216 Tepetate Cr-03-05 I AC 216 0.20 ME 106 0.22 ME 181 0.22 ME
20 UR-RP-7 E C 20 0.14 ME 102 0.21 ME 118 0.17 ME
75 ZR-SGC-6 L C 75 0.13 ME 63 0.18 ME 180 0.17 ME
78 CB-GMP-11 I C 78 0.12 ME 57 0.16 ME 57 0.13 ME
227 San Isidro CM-03-04 I AC 227 0.08 ME 180 0.10 ME 113 0.11 ME
127 DP X Tromba L HxC 127 0.08 ME 117 0.07 ME 65 0.10 ME
113 Coangatzio CM-00-04-05 L AC 113 0.05 ME 241 0.06 ME 63 0.09 ME
234 Macho-I-04 E S 234 0.04 ME 119 0.06 ME 111 0.08 ME
62 PICH-PFC-1 I C 62 0.03 ME 235 0.05 ME 135 0.08 ME
65 PICH-MZU-12 I C 65 –0.05 ME 144 0.02 ME 233 0.06 ME
115 Santa Clara CM–03–04-05 L AC 115 –0.06 ME 236 –0.03 ME 102 0.04 ME
218 Turiran CM-03-05 I AC 218 –0.08 ME 233 –0.05 ME 84 0.03 ME
238 Macho-II-05 E S 238 –0.08 ME 113 –0.07 ME 20 –0.02 ME
240 Macho-IV-04 L S 240 –0.13 ME 78 –0.07 ME 215 –0.04 ME

37
 TABLE 4. SELECTION OF 50 MAIZE ACCESSIONS ACCORDING TO P EFFICIENCY INDEX (PEI) IN PONZOMARAN (EXPERIMENT 1, 2007),
BONILLA (EXPERIMENT, 2008), AND COMBINED EXPERIMENTS (continued)

109 Paso del muerto CM-03-04-05 L AC 109 –0.15 ME 99 –0.07 ME 236 –0.09 ME
117 H-7545 M-00-04-05 L HxC 117 –0.16 ME 20 –0.08 ME 79 –0.09 ME
236 Macho-II-03 E S 236 –0.18 ME 85 –0.12 ME 117 –0.10 ME
144 San Gregorio CM-00-04-05 E AC 144 –0.19 ME 138 –0.17 ME 83 –0.12 ME
135 HV-313XDE M-00-04-05 E HxC 135 –0.23 ME 182 –0.33 I 6 –0.15 ME
63 CB-RSL-2 L C 63 –0.23 ME 238 –0.35 I 218 –0.19 ME
83 OP-4 I C 83 –0.24 ME 40 –0.37 I 238 –0.33 I
57 CB-GMP-9 L C 57 –0.27 ME 181 –0.41 I 234 –0.36 I
40 PA-JRG- 13- Mz 4 L C 40 –0.28 ME 185 –0.42 I 40 –0.38 I
111 CRM-00-04-05 L AC 111 –0.29 ME 131 –0.43 I 131 –0.40 I
79 PICH-EGM-4 I C 79 –0.29 ME 234 –0.45 I 106 –0.41 I
84 AJ-MAM-5 L C 84 –0.29 ME 215 –0.51 I 240 –0.51 I
131 El Tigre CM-00-04-05 E AC 131 –0.30 ME 227 –0.56 I 62 –0.55 I
102 NPZ-GAS-2 I C 102 –0.30 ME 124 –0.68 I 227 –0.55 I
124 Corupo CM-99-04-05 L AC 124 –0.31 ME 62 –0.73 I 138 –0.57 I
106 UR-VR-6 I C 106 –0.68 I 240 –0.74 I 109 –0.81 I
138 DE X HV313 M-00-04-05 E HxC 138 –0.71 I 118 –0.77 I 85 –0.82 I
91 SANAZ-EGS-6 I C 91 –1.10 I 91 –0.78 I 99 –0.86 I
85 PICH-MZU-17 I C 85 –1.13 I 109 –0.79 I 91 –1.00 I
99 TZU-ATM-7 E C 99 –1.18 I 6 –0.96 I 124 –1.22 I

(PEI > 0.51) (PEI > 0.67) (PEI > 0.56)

(–0.96 < PEI < –0.14) (–0.58 < PEI < 0.08) (–0.63< PEI < 0.04)
(PEI < –0.96) (PEI < –0.58) (PEI < –0.63)
†Maturity: Early (E), Intermediate (I), Late (L)
‡BG: Breeding groups: Landrace (C), Advanced Landraces (AC), Hybrids x Landraces (HXC), and Synthetics (S)
#Phosphorus efficiency was expressed as P efficiency index (PEI) gained from the principal component analysis
§Rating of maize accessions as efficient (E), moderately efficient (ME), and inefficient (I) was determined according to the order of the PEI value gained from the principal
component analysis and classification analysis

38
 Several lines of evidence show that root hairs contribute to P acquisition [14]. Low P
availability increases the length and density [30] of root hairs. In this study we observed large
variation for the ability to develop root hairs along the nodal root axis and nodal first-order
laterals (Fig. 3). Whereas several inefficient accessions had shorter and fewer root hairs in
both regions, efficient accessions developed much longer and denser hairs on nodal roots. The
presence of denser root hairs of main axis (r2 = 0.67 to r2 = 0.81) and nodal first-order laterals
(r2 = 0.71), was associated with plant performance at low P. Genetic variation in root-hair
length and density in maize is controlled by several major quantitative trait loci (QTL) [31],
suggesting that this trait could be selected in breeding programs through marked-assisted
selection.

TABLE 5. PHENOTYPIC CORRELATIONS AMONG NODAL ROOT TRAITS AND SHOOT

BIOMASS: NUMBER OF NODAL ROOTS (NODAL_NO), NODAL ROOT LENGTH
(NODAL_RL), NODAL BRANCHING (NODAL_BR), NODAL ROOT ANGLE (NODAL_RA),
SHOOT DRY WEIGHT (SHDW), AND GRAIN YIELD (GY)

HP/LP† Nodal_No Nodal_RL Nodal_Br Nodal_Ra ShDW Gy

Experiment 1 (n = 242)
Nodal_No 0.49*** 0.25*** 0.39*** 0.37*** 0.38*** 0.10
Nodal_RL 0.42*** 0.39*** 0.50*** 0.48*** –0.09 0.24**
Nodal_Br 0.55*** 0.51*** 0.49*** 0.51*** 0.10 0.21**
Nodal_Ra 0.33*** 0.31*** 0.46*** 0.32*** –0.02 0.07
ShDW 0.49*** –0.03 0.21** –0.02 0.50*** 0.09
Gy 0.18* 0.07 0.28** 0.10 0.16 0.68***
Experiments 1 + 2 (n = 50)
Nodal_No 0.32** 0.36** 0.32** 0.13 0.41** 0.09
Nodal_RL 0.19 0.39*** 0.52*** 0.31** 0.00 0.21*
Nodal_Br 0.25* 0.54*** 0.26* 0.24 –0.03 0.18
Nodal_Ra 0.05 0.21 0.45** 0.32** 0.04 0.25*
ShDW 0.48*** –0.01 –0.01 –0.24* 0.36** 0.05
Gy 0.01 0.38*** 0.35** –0.06 0.02 0.54***
†For each environment, values below the diagonal represent correlations within the low P treatment; values
above the diagonal represent correlations within the high P treatment; values on the diagonal (italic) correspond
to across-P treatment correlations
*, **, ***, Significant at P<0.05, P<0.01, P<0.001

3. IDENTIFY AND VALIDATE PHENOTYPIC TRAITS CONFERRING P

ACQUISITION EFFICIENCY IN MAIZE PLANTS 2

3.1. Introduction

Nutrient use efficiency is defined as the ability of a genotype to acquire plant nutrients
from the medium and/or to incorporate or utilize them in shoot and root biomass production
or economic yield [32]. Phosphorus use efficiency can be divided into P acquisition efficiency
(PAE) and P utilization efficiency (PUE) [27]. PAE refers to mobilizing P from poorly
soluble sources or to take up the soluble P available in the soil solution, and PUE is the ability
to produce biomass or yield efficiently using the scarce acquired P. Enhancing P use
efficiency by plants can be achieving through improving P acquisition and/or utilization [6,
33].

2
Full details of this study are in preparation as BAYUELO-JIMÉNEZ, et al., Phosphorus efficiency and
responsiveness in maize landraces from the Purhépecha Plateau: relationships between plant growth, root traits
and phosphorus uptake.

39
 As described in the general introduction, the main mechanisms related to increased P
acquisition efficiency include root architecture, increased proliferation and elongation of root
hairs, root morphology, mycorrhizal associations, high affinity transporters, and the genotypic
capacity to secrete organic compounds in the rhizosphere like phosphatases and organic acids
[34]. A higher P utilization efficiency is mainly attributed to efficient translocation and use of
the stored P in plants [35]. Higher internal utilization efficiency has also been attributed to a
higher grain yield per unit of P in the grain (quotient of utilization) and to a higher ability to
transfer nutrient from shoot to grains, called P harvest index [32].

The existence of considerable variation on root architectural traits that increase topsoil
foraging could enhance the feasibility of improved crops for enhancing mineral nutrient use
efficiency [2]. Identification of cultivars with greater tolerance to suboptimal phosphorus
nutrient levels offers considerable promise for increasing the crop production potential of
marginal low fertility land throughout the world [28]. In assessing maize genotypic variation
for tolerance to P deficiency, great variation was observed among existing landraces of
Michoacan state in their tolerance to P deficiency [36]. Phosphorus efficient landraces of the
Central Mexican Highlands can be of benefit in improving the use of native soil P and
residues of P applied as fertilizer on both low P soils and on soils adequately supplied with P.
To achieve any progress in this area, there is a need to first identify genotypes that differ in
their ability to use P efficiently and that respond differently to applied phosphorus. The
present study was undertaken to identify plant growth traits associated with P acquisition
efficiency (PAE) and/or P utilization efficiency (PUE) in maize landraces growing on P-
deficient Andisol in the Central Mexican Highlands.

3.2. Materials and methods

Three experiments were conducted under low and high P fertilization and rain-fed
conditions, in farmers’ fields in Pontzomaran, San Juan Tumbio, and Bonilla, in the central
highlands of Michoacan, during the 2007 and 2008 growing season. The geographical co-
ordinates, altitude, soil type and the annual rainfall for each site are given in Table I.
Experimental design in all cases was randomized complete blocks with four replications in a
split plot arrangement of treatments, with P level as the main plots and accessions as the sub-
plots. Crop sequences and fertilizer additions were identical as described in Section 2.2.2.

Fifty and twenty-six local maize accessions were grown in 2007 in Ponzomaran (Exp.
1) and San Juan Tumbio (Exp. 2), respectively. Common accessions were planted in the
succeeding crop cycle, Experiment 3 in Bonilla (2008) (n = 50), which allowed a comparison
across locations (Table 2). One plant per replication per accession was harvested between 78
and 89 DAP, which correspond to the silking stage [37]. Analyses for root architecture (crown
roots and root hairs) and biomass production were performed as described previously. Root
and shoot dry tissue samples were ground and analyzed for P content [28]. The phosphorus
acquisition efficiency (PAE) is a measure of the P absorption per unit root dry weight (DW)
(mg P g-1 root DW), and was calculated from total P per plant divided by root DW (g) [28].
The phosphorus utilization efficiency (PUE) is a measure of DW return per unit P uptake (g
DW mg-1 P), and was calculated as total plant DW divided by P content per plant [28]. The
phosphorus efficiency of maize accessions was determined by the phosphorus efficiency
index (PEI) [25] and assessed using principal component analysis [24].

40
3.3. Results and discussion
3.3.1. Phenotypic traits conferring P acquisition efficiency

A number of physiological traits could contribute to P efficiency, by improving P

acquisition from the soil or by improving the utilization of acquired P in growth [28, 32].
Comparisons of maize accessions in three soil environments and maturity types showed that
differences in P acquisition were due to a large variation for root growth, root to shoot ratio,
root and shoot P content (Tables 6, 7 and 8) and root architecture (data not shown). In general,
greater root and shoot dry weight and grain yield were observed in plants with late maturity
under low soil P. Therefore, late P-efficient accessions were the most efficient in terms of
biomass produced per unit of P absorbed (Tables 7 and 8), produced the largest root system
(25.4 cm root-1 and 70.5 nodal roots plant-1) (Table 9), had the largest root uptake per unit root
weight (17 to 24.5 mg P g-1 root DW), and had more nodal rooting than did P-inefficient
accessions (85, 99, 59, 78, 71 and 89% in nodal root length, nodal root number, nodal root
angle, and root hair density and length, respectively).

TABLE 6. ROOT AND SHOOT GROWTH PARAMETERS AND GRAIN YIELD OF MAIZE
ACCESSIONS OF DIFFERING MATURITY IN PONZOMARAN (EXPERIMENT 1, 2007), SAN
JUAN TUMBIO (EXPERIMENT 2, 2007), AND BONILLA (EXPERIMENT 3, 2008) WITH LOW
P (LP) AND HIGH P ADDITION (HP)

Experiment P Root dry weight Shoot dry weight Grain yield

level (g plant-1) (g plant-1) (Mg ha-1)
Early Middle Late Early Middle Late Early Middle Late
1 (2007) LP† 7.2a 6.4 a 6.5a 69.7a 61.0 b 65.1a 2.8a 3.0a 3.3a
HP 9.8a 9.0 a 10.4a 83.4a 77.2 b 83.4a 3.1b 3.0b 3.6a
2 (2007) LP 10.4b 10.5 b 11.4a 126.2b 120.9b 150.0a 3.3a 3.3b 3.8a
HP 9.8b 10.2 b 12.9a 129.0b 122.6b 143.2a 3.1b 3.3b 3.9a
3 (2008) LP 9.1b 10.1 b 11.7a 151.5c 167.5b 189.3a 3.9 a 3.8 a 4.5a
HP 11.6b 12.3 b 13.9a 173.9c 190.4b 205.7a 4.2 a 4.0 a 3.9a
†Within rows, means followed by the same letter are not significantly different according to LSD (P<0.05).

When LP availability limits plant growth, the dry weight produced will depend on P
uptake and the efficiency of internal P use for dry matter production [39]. Dry weight
accumulation in LP level was tightly correlated with P uptake, but not with tissue P
concentration (Fig. 4, Table 11). This suggests that tissue P concentration is not a reliable
criterion in assessing maize genotypes for P deficiency. Therefore, tolerance to P deficiency is
entirely dependent on genotypic differences in P uptake.

P uptake itself is a function of the root size and root efficiency [27], where root
efficiency was defined as P uptake per unit dry weight (Table 7). Genotypic variation existed
for both traits and high P acquisition can be achieved, either due to large roots (P-efficient
accessions 40, 75, 109, 113, 115, 182, 199 and 234) (Table 9) or with a more efficient root
system (6, 40, 75, 78, 109, 182, 199 and 234). Phosphorus-efficient accessions, particularly of
late maturity, had a large root system (12.3 to 16.5 g plant -1) and their P uptake (301.7 mg
plant-1) and PAE (23.4 mg P g-1 root DW) was above average under LP conditions (Table 8).
As a result, the correlation between P uptake and root efficiency was high (Table 11). This
indicates that a better adaptation to P deficiency as a result of a large root system may be a
more reliable criterion to identify efficient accessions to LP soils.

41
TABLE 7. SHOOT P CONCENTRATION, ROOT AND SHOOT P CONTENT OF MAIZE
ACCESSIONS OF DIFFERING MATURITY IN PONZOMARAN, SAN JUAN TUMBIO, AND
BONILLA WITH LOW P (LP) AND HIGH P ADDITION (HP)

Experiment P Shoot P concentration Root P content Shoot P content

(mg P g-1) (mg) (mg)
Early Middle Late Early Middle Late Early Middle Late
1 (2007) LP† 0.98a 1.00 a 0.98a 5.8 a 5.1 a 5.7a 68.6a 60.7a 65.2a
HP 1.01a 0.99 a 1.07a 7.8 b 7.4 b 9.4a 85.4a 76.2b 88.8a
2 (2007) LP 1.07a 1.08 a 1.08a 5.6 b 6.8 a 7.7a 134.3b 134.9b 187.0a
HP 1.12a 1.18 a 1.18a 5.7 b 6.9 a 7.2a 144.6b 147.7b 197.0a
3 (2008) LP 1.51a 1.50 a 1.49a 5.8 b 7.2 a 6.7a 229.0b 252.7a 280.2a
HP 1.56a 1.49 a 1.58a 7.2 a 7.6 a 8.2a 271.4b 284.1a 301.7a
†Within rows, means followed by the same letter are not significantly different according to LSD (P<0.05).

TABLE 8. TOTAL P UPTAKE, P ACQUISITION EFFICIENCY (PAE), AND P UTILIZATION

EFFICIENCY (PUE) OF MAIZE ACCESSIONS OF DIFFERING MATURITY IN
PONZOMARAN, SAN JUAN TUMBIO, AND BONILLA WITH LOW P (LP) AND HIGH P
ADDITION (HP)

Experiment P Total P uptake PAE PUE

level (mg plant-1) (mg P g-1 root DW) (g DW mg-1 P)
Early Middle Late Early Middle Late Early Middle Late
1 (2007) LP† 74.7a 65.7 b 70.8ab 11.1a 12.3 a 12.2a 1.08a 1.1 a 1.1a
HP 93.9ab 83.7 b 98.0 a 9.9a 9.9 a 10.7a 1.06a 1.1 a 0.1a
2 (2007) LP 139.9 141.8 b 194.7a 14.4b 13.3 b 19.5a 1.00a 1.0 a 0.9b
HP 150.7 155.4 b 204.1a 15.8b 15.7 b 18.1a 0.94a 0.9 a 0.8b
3 (2008) LP 234.8 259.9ab 287.0a 27.3a 28.0 a 26.4a 0.69a 0.7 a 0.7a
HP 278.7 291.8ab 319.9a 24.8a 25.3 a 23.3a 0.68a 0.7 a 0.7a
†Within rows, means followed by the same letter are not significantly different according to LSD (P<0.05).

Root system architecture, morphology, and biochemistry, are key traits for optimizing
P acquisition, and thus their P use efficiency and responsiveness [6]. Architectural traits
associated with enhanced topsoil foraging include shallower growth angles of axial roots,
enhanced adventitious rooting, a greater number of axial roots, and greater dispersion of
lateral roots [2]. Maize genotypes with increased or sustained elongation of nodal roots and
lateral root development under P deficiency had superior ability to acquire P and maintain
growth [36]. This study confirmed that enhanced nodal rooting and greater nodal branching
(85 to 99%) is indeed important for plant adaptation to LP in maize. Nodal rooting was
significantly related with P acquisition and plant growth in the field among P efficient groups
(Table 10). P-efficient accessions [i.e. Paramuén Cr-03-JR10 (40), ZR-6 (75), Paso del
Muerto Cr-03-TA (109), and Macho III-05 (182)] with greater nodal rooting and lateral
branching at LP had greater biomass and P uptake efficiency (r2 = 0.53 to r2 = 0.94) then did
inefficient accessions with reduced nodal root formation and lateral branching (r2 = 0.18 to r2
= 0.23). The greater nodal rooting of P-efficient accessions under LP could be explained by
the greater overall biomass and the weak allometric relationship of plant biomass with root
biomass (r2 = 0.05). Therefore, if nodal roots in the topsoil are advantageous for acquiring P
under limited P availability, this weak allometric relationship would facilitate the selection of
efficient accessions with high nodal rooting in the field.

The presence of root hairs conferred a marked benefit under LP conditions, as

evidenced by increased plant biomass, P uptake and reproductive output (grain yield) of P
efficient accessions (Table 10), and decreased those parameters on P-inefficient ones (r2 =

42
0.17, r2 = –0.34, r2 = 0.19). The competitive advantage of root hairs is presumably due to
greater P uptake [40]. Phosphorus mobility in soil is governed by diffusion rather than mass
flow [14], and therefore root P uptake is limited by localized P depletion around the root.
Root hairs extend the P depletion zone from the root epidermis, thereby increasing the rate of
P uptake and the total amount of P accessible by the root. Several lines of evidence show that
root hairs contribute to P acquisition [40]. Root hairs, particularly denser and longer root hairs
of nodal first-order laterals were associated with plant performance, grain yield and root P
acquisition at LP (Table 10). Late P-efficient accessions Paramuén Cr-03-JR10 (40), Paso del
Muerto Cr-03-TA (109), Paso del Muerto CM-00 (113), and Santa Clara CM-03 (115) can be
selected showing a significant relationship between root hairs and root P acquisition,
suggesting that P acquisition ability of genotypes is a decisive factor in expression of high P
efficiency.

3.3.2. Phenotypic traits conferring P utilization efficiency

Phosphorus utilization efficiency represents the amount of dry matter produced per
unit of P absorbed (g DM mg-1 P) [41]. Any species able to maintain metabolic activities at
low tissue P concentration and produce more dry matter per unit of P absorbed is considered
efficient in P utilization. In this study, there were accessions in which P content in shoot had a
highly significant and positive correlation with root dry weight (r = 0.54**) and shoot dry
weight (r = 0.78***) suggesting that accessions with higher root dry weight accumulated
higher amount of shoot P and produced higher shoot dry matter at LP. Thus, under P-stress,
better P acquisition and PUE by the P-efficient accessions for biomass synthesis collectively
formed the basis of higher shoot dry matter production, evidencing that P uptake and PUE are
important plant traits for selecting low P-tolerant accessions (Table 8).

Thus, under P-stress, better P acquisition and PUE by the P-efficient accessions for
biomass synthesis collectively formed the basis of higher shoot dry matter production,
evidencing that P uptake and PUE are important plant traits for selecting low P-tolerant
accessions (Table 9). In addition, PUE and whole–plant P uptake were positively correlated
with each other under LP and HP conditions (Table 11), suggesting lack of any interactions
between these two parameters under either of the growth conditions. This, therefore, suggests
the possibility of a combination of these two parameters to improve growth responses of
accessions to P. Finally, late P-efficient accessions Paramuén Cr-03-JR10 (40), ZR-6 (75),
CM-754 (117), Macho IV-03 (239), and early-P efficient accessions M-I-03 (233), M-I-04
(234), and Macho-II (236) had an excellent relationship between grain yield and shoot
biomass (r2 = 0.77***) under LP conditions, indicating that P utilization through better dry
matter accumulation and partitioning of above-ground dry matter to grain is the most critical
factor in expression of high P efficiency of these accessions.

Phosphorus efficiency is a very complex phenomenon affected by a large number of

plant mechanisms and various physiological and biochemical traits associated with P
acquisition from soil and P utilization at the cellular level [28]. As presented in Table 10,
accessions can be selected showing an excellent correlation between P efficiency and P
content leading to a suggestion that P acquisition is a decisive factor in expression of high P
efficiency. Alternatively, from Table 8 and 9, accessions can be selected which are more or
less the same in P content, but differing greatly in P efficiency, leading to a suggestion that P
utilization ability is the most critical factor in expression of high P efficiency. In both cases,
root system architecture and morphology affected the ability of a plant to acquire P from the
soil, and thus their P use efficiency and responsiveness to external P. It seems likely that P
efficiency mechanisms may be different among the accessions of a given species.

43
 300 HP LP (a)

efficiency
(b) (c)

Low P
150
R 2 = 0.54 R2 = 0.98 R 2 = 0.63
**
2 2 2
R = 0.55 R = 0.49 R = 0.69
Shoot dry weight (g)

0
300

Moderate P
efficiency
150
R2 = 0.65 R2 = 0.76 R2 = 0.70
**
2 2 2
R = 0.80 R = 0.69 R = 0.32
***
0

efficiency
300

High
150
R2 = 0.82 R 2 = 0.97 R2 = 0.76
2 2 2
R = 0.89 R = 0.76 R = 0.69

P
0 ***

0 250 500 0 250 500 0 250 500

Total phosphorus

FIG. 4. Relationship between total P uptake and shoot biomass of low, moderate and high P efficient
accessions grown under high- (HP) and low- (LP) phosphorus availability measured at the silking
stage in (a) Experiment 1 (2007), (b) Experiment 2 (2007), and (c) Experiment 3 (2008). Each data
point represents the mean of four replicates of 12, 24, and 14 accessions; 7, 15, and 4 accessions; 14,
28, and 8 accessions, respectively. *, **, and *** indicate significance at P<0.05, P<0.01, P<0.001.
FIG. 4. Relationship between total P uptake and shoot biomass of low, moderate and high P efficient
accessions grown under high- (HP) and low- (LP) phosphorus availability measured at the silking
stage in (a) Experiment 1 (2007), (b) Experiment 2 (2007), and (c) Experiment 3 (2008). Each data
point represents the mean of four replicates of 12, 24, and 14 accessions; 7, 15, and 4 accessions; 14,
28, and 8 accessions, respectively. *, **, and *** indicate significance at P<0.05, P<0.01, P<0.001.

3.3.3. Phosphorus efficiency and P responsiveness

Large genotypic variation within maize landraces was confirmed for plant growth in
LP soil. Fifty common maize accessions were grouped into 3 categories of P efficiency based
on growth, grain yield, tissue P concentration and P content, PAE and PUE parameters at LP
and their relative values to those at HP. Our results indicated that 12 accessions across
locations had the lowest growth and yield reduction and the highest levels of P efficiency (PEI
> 0.54) under low P. Among accessions, Paramuén Cr-03-JR10, PICH-1, PICH-4, PICH-17,
Paso del Muerto Cr-03-TA, San Gregorio CM 00, Macho III-05, M-I-0, Macho I-05, and
Macho-II 03 had consistently higher PEI in LP (Table 11). When the combination of PEI with
P responsiveness at HP is considered, accessions SHUI-2 (6), Paramuén Cr-03-JR10 (40),
ZR-6 (75), CB-2 (78), Paso del Muerto Cr-03-TA (109), CCHEDE CM-00 (199), Macho III-
05 (182), M-I-03 (233), M-I-04 (234), and M-IV-03 (239) were the best accessions for P-

44
deficient acidic soil of this region (Table 11). These accessions were categorized as the most
P efficient under LP and as the most responsive to increased P availability. The natural
genetic variation observed among maize landraces demonstrates the potential for breeding
cultivars with improved P efficiency, which will ultimately utilize applied inorganic P
fertilizers more efficiently. In addition, the relationship between PEI and plant content and dry
matter showed a significant correlation (r = 0.65 to r = 0.93) at LP. Such a positive correlation
indicates that PEI based on dry matter and plant content uptake can be an appropriate index
for P efficiency evaluation.

4. USE OF 32P ISOTOPE TECHNIQUES TO IDENTIFY GENOTYPES FOR

SUPERIOR PHOSPHORUS ACQUISITION AND / OR UTILIZATION

4.1. Introduction

Andisols contains considerable amounts of P but a large proportion is bound to

different soil constituents, forming complexes of limited bioavailability [42]. This type of soil
is commonly referred to as a P-fixing soil and the concentration of phosphate in the soil
solution is suboptimal for crop production. A first line of strategy for soils with low total P
content is regular amendment with small doses of P fertilizer. However, in soils with high
total P content that fix most of the P, P fertilizer will be equally fixed. In this type of soil,
plants do respond to P fertilizer application, but annual P applications are needed to sustain
crop yields [34].

The use of isotopes has enabled tracing of the dynamics of fertilizers in the soil-plant
system [43]. The use of 32P isotope as a tracer applied to the soil with phosphate fertilizer
permits the detection of exchangeable phosphate ions in the solution and those absorbed by
the plant [44]. The method of isotopic dilution is a useful, practical and economical
methodological alternative when it is not possible to mark the nutrient of interest isotopically.
The method consists in marking the soil with the isotope to differentiate the origin of the
nutrient: soil and fertilizer [43]. The objective of the present study was to assess the P uptake
and recovery from the P fertilizer by selected maize accessions using the 32P isotope dilution
technique.

4.2. Materials and methods

The efficiency of P uptake was determined for five maize accessions from the
Purhepecha region represented by two early (San Gregorio and synthetic M-I-04); two
intermediate (PICH-4 and Zacapu), and one late (DP x Tromba) accessions and a commercial
variety (cv. Leopardo) used as a check (Table 2). These genotypes differed in both P
efficiency and P responsiveness at early vegetative stages [46]. This study was conducted in
the National Nuclear Research Institute (ININ), Mexico, in a greenhouse with an atmosphere
between 14.3 and 26.5oC, 50 to 70% relative humidity, and a fall photoperiod (13h day/11h
night). Artificial light for 12 hours per day from cool-white, fluorescent tubes supplemented
the fall daylight. Plants were grown in pots with 6 kg of Vitric Andisol soil of low available
soil P, which was obtained from a plot on Laderas, Pátzuaro, Michoacan, where maize
landraces are cultivated (Table 1).

The experiment design was completely randomized with a factorial arrangement of (3

× 6) treatments: three levels of P supply and 6 genotypes. There were four replications for a
total of 72 units. Each experimental unit was composed for one pot and included P additions
as diammonium phosphate (NH4)2HPO4 at rates of 0.0 g P pot-1, 0.22 g P pot-1 and 0.44 g P
pot-1. These rates are equivalent to 0, 25 kg P ha-1 and 50 kg P ha-1.

45
46
TABLE 9. SHOOT DRY WEIGHT, P ACQUISITION EFFICIENCY (PAE), AND P UTILIZATION EFFICIENCY (PUE) OF COMMON MAIZE
ACCESSIONS GROWN IN A P-DEFICIENT SOIL ITH LOW P (LP) OR WITH HIGH P ADDITION (HP), IN TWO SEASONS BETWEEN 2007 AND
2008. ACCESSIONS ARE RANKED BY P EFFICIENCY INDEX (PEI)

ID Accessions M‡ BG# Shoot dry weight (g plant-1) PAE (mg P g-1 root DM) PUE (g DM mg-1 P)
LP HP PEI§ SdShΨ PRɸ LP HP SdPAE PR LP HP SdPUE PR
40 Paramuén Cr-03-JR10 L C 153.7 195.9 0.75 3.38 ER 17.2 18.2 0.55 ER 0.83 0.87 1.24 ER
109 Paso del Muerto Cr-03 L AC 155.6 173.9 0.66 1.70 ER 21.8 19.7 0.44 ER 0.72 0.88 –0.16 ENR
78 CB-2 I C 106.6 133.9 0.37 0.21 ER 17.4 19.8 0.92 ER 0.78 0.92 –0.48 ENR
236 M-II-03 E S 116.1 139.3 0.30 0.63 ER 16.2 15.4 –0.18 ENR 0.84 0.91 –0.52 ENR
233 M-I-03 E S 105.5 146.8 0.20 0.98 ER 18.4 16.3 –0.02 ENR 0.76 0.87 0.31 ER
239 M-IV-03 L S 112.3 131.9 0.10 0.10 ER 16.2 17.7 –0.01 ENR 0.81 0.89 0.19 ER
75 ZR-6 L C 142.2 146.7 0.05 1.03 ER 18.7 17.7 0.27 ER 0.90 0.89 0.46 ER
234 M-I-04 E S 141.2 137.7 0.04 0.58 ER 22.2 20.0 1.05 ER 0.93 0.95 –0.63 ENR
182 M-III-05 I S 129.9 127.9 0.64 –0.43 ENR 20.0 19.2 0.57 ER 0.84 0.86 0.45 ER
144 San Gregorio CM-00 E AC 101.7 106.5 0.60 –1.14 ENR 19.5 14.4 –0.62 ENR 0.85 1.18 –2.56 ENR
79 PICH-4 I C 135.8 124.0 0.54 –0.12 ENR 17.4 14.9 –0.29 ENR 0.87 0.86 0.03 ER
235 M-I-05 E S 123.8 127.7 0.41 –0.09 ENR 17.0 15.7 –0.28 ENR 0.83 0.83 0.61 ER
62 PICH-1 I C 114.2 122.5 0.33 –0.39 ENR 14.6 15.6 –0.31 ENR 0.87 0.89 –0.18 ENR
199 CCHEDE CM-00 E AC 124.5 117.5 0.19 –0.48 ENR 15.1 18.7 0.75 ER 1.00 0.96 1.35 ER
6 SHUI-2 L C 102.5 116.8 0.15 –0.45 ENR 15.4 20.2 1.18 ER 0.92 0.89 0.99 ER
Average 114.9 128.6 17.9 16.7 0.92 0.88
LSD (0.05)† 52.4 52.6 9.0 7.9 0.23 0.21
†To compare paired values among accessions
‡Maturity: early (E), intermediate (I), late (L)
#BG: Breeding groups: landrace (C), advanced landrace (AC), hybrids x landrace (H x C), and synthetic (S)
§Phosphorus efficiency index (PEI) obtained from the principal component analysis
ΨStandardized value of shoot dry weight (SdSh), P acquisition efficiency (SdPAE), and P utilization efficiency (SdPUE) under high P conditions. Data from the 3 individual
experiments were standardized by dividing relative values by the standard deviation of the trial
ɸPhosphorus responsiveness (PR) is expressed by shoot biomass or grain yield under HP level. Four categories, efficient and responsive (ER), non-efficient and responsive
(NER), non-efficient and non-responsive (NENR), and efficient and non-responsive (ENR)
 TABLE 10. PHENOTYPIC CORRELATIONS AMONG ROOT TRAITS: NODAL ROOT LENGTH (NOD_RL), NUMBER OF NODAL ROOTS
(NOD_NO), NODAL BRANCHING (NOD_BR), NODAL ROOT ANGLE (NOD_RA), ROOT HAIR DENSITY (RHD_BNR), (RHD_MNR) AND LENGTH
(RHL_BNR), (RHL_MNR) FROM THE BASAL AND MIDDLE REGION OF NODAL ROOT FIRST ORDER LATERALS AND MEASURES OF P
EFFICIENCY: SHOOT DRY WEIGHT, GRAIN YIELD, P ACQUISITION EFFICIENCY, AND P UTILIZATION EFFICIENCY OF P-EFFICIENT
ACCESSIONS ACROSS LOCATIONS

Parameter Shoot dry weight Grain yield P acquisition efficiency P utilization efficiency
(g plant-1) (Mg ha-1) (mg P g-1 root DW) (g DW mg-1 P)
P Early Middle Late Early Middle Late Early Middle Late Early Middle Late
Nod_RL HP –0.49† 0.02 –0.85 –0.08 –0.01 –0.09 –0.04 –0.02 –0.54 0.04 0.00 0.45
LP 0.50 0.01 0.76 –0.72 0.67 0.88 –0.96 0.25 0.11 0.13 –0.63 0.11
Nod_No HP 0.01 0.13 0.19 –0.21 0.02 0.00 –0.27 0.79 0.35 0.28 –0.36 –0.73
LP 0.94 0.53 0.75 –0.21 0.32 0.56 –0.89 0.02 0.43 0.61 0.48 0.01
Nod_Br HP 0.05 0.28 0.73 0.46 –0.23 0.04 0.52 0.06 0.42 –0.54 0.13 –0.23
LP 0.57 0.82 0.31 –0.90 0.35 0.04 –0.82 0.22 0.73 0.02 0.52 –0.38
Nod_ra HP 0.56 0.01 –0.40 0.11 –0.09 –0.05 –0.07 0.05 –0.49 –0.06 –0.14 0.89
LP –0.25 –0.10 –0.75 –0.54 0.59 –0.80 0.03 0.85 –0.02 0.66 0.75 0.50
RHD_bnr HP 0.45 –0.17 –0.07 0.05 0.02 0.45 0.02 –0.05 –0.32 –0.02 0.62 0.29
LP 0.17 –0.38 –0.01 0.63 0.08 –0.80 0.02 0.00 0.61 0.57 –0.15 –0.08
RHL_bnr HP –0.98 –0.21 0.95 –0.86 0.07 0.03 –0.81 –0.16 0.70 0.80 0.96 –0.57
LP 0.44 –0.12 0.44 0.34 –0.05 –0.96 –0.02 0.04 0.50 0.83 –0.36 –0.02
RHD_mnr HP 0.03 –0.25 –0.26 –0.11 –0.03 0.60 –0.15 –0.14 –0.34 0.16 0.61 0.03
LP 0.31 –0.52 0.34 0.47 –0.24 –0.30 0.02 –0.05 0.97 0.72 –0.05 –0.67
RHL_mnr HP –0.53 –0.12 0.79 –0.94 –0.04 0.20 –0.97 –0.07 0.55 0.97 0.90 –0.69
LP 0.74 –0.12 0.81 0.46 –0.16 –0.82 0.19 –0.12 0.69 0.99 –0.20 –0.12
†Correlation coefficients in bold type are significant at P<0.05

47
48
TABLE 11. PHENOTYPIC CORRELATIONS AMONG ROOT P CONCENTRATION (RPCT), ROOT P CONTENT (RPC), SHOOT P CONCENTRATION
(SPCT), SHOOT P CONTENT (SPC), TOTAL P CONTENT (TPC), ROOT DRY WEIGHT (RDW), SHOOT DRY WEIGHT (SHDW), TOTAL DRY
WEIGHT (TDW), P ACQUISITION EFFICIENCY (PAE), P UTILIZATION EFFICIENCY (PUE), AND GRAIN YIELD (GY)

HP/LP† RPCt RPC SPCt SPC TPC RDW ShDW TDW PAE PUE GY
RPCt 0.15 0.58*** –0.10 0.06 0.01 0.06 0.11 0.11 0.00 –0.01 –0.08
RPC 0.54*** 0.43** 0.04 0.37* 0.38* 0.82*** 0.39* 0.47** –0.38* 0.11 0.01
SPCt 0.13 0.02 0.20 0.61*** 0.51** 0.14 0.10 0.09 0.37* 0.90*** 0.27
SPC 0.18 0.28* 0.66*** 0.52*** 0.88*** 0.38* 0.81*** 0.81*** 0.44** 0.60*** 0.21
TPC 0.22 0.36* 0.65*** 0.99*** 0.44** 0.37* 0.72*** 0.73*** 0.52** 0.68*** 0.19
RDW 0.11 0.88*** –0.03 0.54** 0.32** 0.46** 0.35* 0.45** –0.56*** 0.12 0.07
ShDW 0.13 0.31* 0.14 0.78*** 0.77*** 0.30* 0.47** 0.99*** 0.32** 0.16 0.04
TDW 0.14 0.41** 0.13 0.77*** 0.78*** 0.41* 0.99*** 0.43* 0.11 0.25 –0.13
PAE 0.06 –0.61*** 0.43** 0.39* 0.42** –0.76*** 0.40* 0.16 0.34* 0.54*** 0.11
PUE 0.16 0.02 0.98*** 0.66*** 0.65*** –0.05 0.19 –0.22 0.48** 0.17 0.25
Gy 0.15 0.16 –0.01 0.22 0.19 0.07 0.26 0.25 0.05 0.00 0.57***
†For each environment, values below the diagonal represent correlations within the lowP treatment; values above the diagonal represent correlations within the high P treatment;
values on the diagonal (italic) correspond to across-P treatment correlations
*, **, ***, Significant at P<0.05, P<0.01, P<0.001, respectively
 The Andisol soil was labelled with 32P by applying a high activity (10 mCi). The tracer
was added in 300 ml of water at pH 6.0 to ensure uniform labelling of soil at concentrations
of 0.395 mCi pot-1 and 0.791 mCi pot-1. Dilute solution was applied into the soil at a depth of
10 cm, near to the plant root zone. Plants were irrigated with distilled water applied to the soil
to field capacity (33 kPa) to avoid leaching of the radioactive 32P-labelled marker and other
nutrients. Two seeds were planted in every pot. Five days after emergency, seedlings were
thinned to one plant per pot. N was added as a solution of NH4NO3 at 60 kg−N ha-1 to the pots
in which maize was grown. The N was split applied, 33% at planting, 33% at 14 days after
planting (DAP), and 33% at 28 DAP.

One sample per replication per maize accession was harvested at 39 DAP; shoots were
cut at 1 cm above the soil surface, and dried in an oven at 70 oC during 2 days. After
determining the shoot dry weight, shoot were cut in pieces < 5 mm and a sample of 0.5 g of
every shoot was digested in hot H2SO4.The P concentration in the digests were determined
colorimetrically [38]. 32P activity in the digest was determined by liquid scintillation counting.
Samples were counted in a Beckman LS-60000LL liquid scintillation system. The content of
P in plant was calculated by multiplying the P concentration by the dry weight (P
concentration in plant (mg P g-1) × dry matter yield (g P pot-1) / 100). Based on the isotopic
dilution method, the proportion of P taken up by the plants from the fertilizer and soil were
calculated [43]. The percentage of P derived from the fertilizer (% Pddf) was obtained
dividing the specific activity (SA) in the plant by SA in the fertilizer × 100. Thus, the
percentage of P derived from the soil (% Pdds) was obtained by difference: 100 – (% Pdds).
The amount of P fertilizer extracted by the crop (P-fertilizer yield) was calculated by
multiplying the total P yield and the percentage of P derived from fertilizer: P-fertilizer yield
(mg P pot-1) = (Total P yield × % Pddf) / 100. The P-fertilizer extracted by the plant relative
to the dose of P applied (efficiency P-fertilizer) is known as P use efficiency: P-fertilizer
efficiency (%) = (P-fertilizer yield / dose of P applied) x 100.

4.3. Results and discussion

Due to the very low available soil P utilized in this study (6 mg P kg-1 soil), which is
typical of Mexican volcanic soils, the response of the maize to fertilizer application was
evident (Fig. 5). The dry matter (DM) production increased from 1.88 g (control) to 2.81 and
5.01 g for the rates of 25 and 50 mg P kg-1 soil, respectively. The increases in both DM and P
content of all accessions, except Zacapu and DP x Tromba, resulted in doubling the DW
accumulated by the crop, when the P rate was increased from 25 to 50 mg P kg-1 soil as
diammonium phosphate. In DP x Tromba, total P content was comparable and responded
similarly to P application (Fig. 5). The smallest P contents were obtained in LP treatment in
cv. Leopardo.

The mean specific activity of P in DP x Tromba was significantly smaller than all
other accessions which had the highest value under HP and LP conditions (Fig. 6). If all the
species drew their P from the same pool of available P, then the specific 32P activity of the P
in their shoots should be comparable, although concentrations of P in the shoots and the
amounts of P accumulated may differ [45]. Consequently, the lower specific radioactivity of
the P taken up by DP x Tromba indicates that it was able to access a pool of soil P that was
less available to all other accessions.

49
 10
Total dry weight (g plant-1) HP LP
a)
8
***
6
*** ***
4 **
2

12
b)
Total P uptake (mg)

9
*** ***
***
6 ***
*
3

0
Gregorio M-I-04 Zacapu PICH-4 DP-Tromba Leopardo
NER NER NENR NER ER NER

FIG. 5. Total dry weight (a) and total P content (b) for maize accessions grown in a Vitric
Andisol. The error bar represents standard deviation from the mean. Within maize
accessions, bars with *, **, and *** indicate significance at P<0.05, P<0.01, P<0.001,
respectively. Categories represented by efficient and responsive (ER), inefficient and
responsive (NER), and inefficient and non-responsive (NENR).

The P uptake derived from soil (% Pdds) and from fertilizer (% Pddf) was
significantly affected by the P supply, accession, and interaction (P<0.0001), suggesting that
phosphate fertilization improved P uptake from soil or fertilizer due to the initially low P
content in the soil (Table 1). The % Pdff values for HP were, in general, larger than in LP for
all accessions. However, San Gregorio and PICH-4 showed the highest uptake derived from
fertilizer at LP. Under LP condition, significantly larger % Pdds and lower P fertilizer yield in
the plant was observed for DP x Tromba (Fig. 6). Different % Pdds in maize accessions and
cultivar indicate that chemically different pools of soil P were utilized, with DP x Tromba
accessing a larger pool than all other accessions. The increase in % Pdds indicates that when P
deficiency occurs, DP x Tromba can exert certain mobilization mechanisms and access non-
labile P in the Andisol soil.

In comparing P-fertilizer yield and the efficiency of P fertilizer, large differences

among accessions were found (Fig. 6). San Gregorio, M-1-04, and PICH-4 produced the
highest values of P-fertilizer yield and % P-fertilizer use efficiency when the P-fertilizer was
applied at the seedling stage. The % P-fertilizer use efficiency in these accessions was 14, 5.3
higher than cv. Leopardo at HP conditions. These accessions were the most responsive to high
P rates of application indicating to be more demanding for their growth and development, as
identified in previous studies in the field [46].

50
 30000 1000
HP LP (a) LSD 0.05 (LP) =102.3 (d)
LSD 0.05 (LP) =1082 LSD 0.05 (HP) =181.1

Total P yield
750

(mg P pot-1)
LSD 0.05 (HP) =2700
20000
SA (Bq mg-1 P)

500

10000
250

0 0

40 200
(b) (e)
LSD 0.05 (LP) =1.53 LSD 0.05 (LP) =10.1

Fertilizer P yield
30 LSD 0.05 (HP) =3.82 150 LSD 0.05 (HP) =57.4

(mg P pot-1)
% Pddf

20 100

10 50

0 0

150 50
(c)
Efficiency P fertilizer (%)
LSD 0.05 (LP) =1.53 LSD 0.05 (LP) = 4.87 (f)
40
LSD 0.05 (HP) =3.82 LSD 0.05 (HP) =13.8
100
30
% Pdds

20
50

10

0 0
Gregorio M -I-04 Zacapu PICH-4 DP Leopardo Gregorio M -I-04 Zacapu PICH-4 DP Leopardo

FIG. 6. Specific radioactivity of the P taken by the plant (a), %Pddf (b), %Pdds (c), total P yield (d),
fertilizer P yield (e), and fertilizer P efficiency (f) for maize accessions grown in an Andisol.

We have shown that the accession DP x Tromba is able to access non-labile P.

However, different varieties need to be tested for their ability to mobilize non-labile P. Recent
studies have investigated genotypic differences of maize in P efficiency [36]. The authors
showed that there exist differences in P efficiency under low P conditions. Particularly, they
showed that the development of an extensive root system with dense root hairs was one of the
main strategies of acquiring soil P for P-efficient accessions. In our study, DP x Tromba was
able to access non-labile P, but this did not enable it to increase P uptake. It is therefore
relevant to investigate more maize accessions for their ability to access and utilize non-labile
P.

Compared to the results for P-inefficient accessions, the significantly lower values of
the specific activity of P in the late P-efficient DP x Tromba and the higher values of P
derived from soil by this accession clearly show for the first time that DP x Tromba utilizes
soil P from a normally non-labile soil P pool that is not utilized by P-inefficient accessions
and cv. Leopardo. In conclusion, these data demonstrate that P-efficient accession has a
specific mechanism to mobilize and use a fraction of soil that is not utilized by other
accessions of the given species.

51
5. CONCLUSIONS

The first two studies described above were aimed to assess the genotypic variation in
root architecture and plant growth traits associated with P acquisition efficiency of maize
landraces to a P-deficient Andisol in the Central Mexican Highlands. The Mexican Highlands,
including the Purhepecha Plateau, with strong dominance of traditional varieties, are an
important repository of genetic resources. Controlled environment and field studies confirmed
that root architecture in maize landraces was responsive to low P availability and that
variation in these traits within landraces contributed to differing capacities for P acquisition.
P-efficient accessions adapted to low P produced more nodal and lateral roots and long and
dense root hairs that greatly increase the soil volume that roots can exploit. In this Andisol, Pi
was low, independent of P fertilization. The only possibility for the maize plants to absorb
more P was to access more soil volume by increased root growth. Nodal rooting was best
correlated with P acquisition efficiency and grain yield, followed by root hair density and
length on the nodal first-order laterals. A more P uptake efficient plant would be a
combination of the high P acquisition and the extended root system seen in P efficient
accessions of late maturity. Hence, root traits of more P acquisition efficiency in maize
landraces exist, opening the possibility to breed for more P acquisition efficient varieties.

The final study was aimed to determine the ability of maize accessions to acquire
adequate amounts of P from fertilizer and soil that are very low in available P, by measuring
the specific radioactivity of plant P in P-efficient and P-inefficient accessions grown on a 32P-
labeled soil. Lower values of the specific radioactivity of P in the late P-efficient DP x
Tromba and the higher values of P derived from soil by this accession clearly show, for the
first time, that P-efficient accessions utilize soil P from a normally non-labile soil P pool that
is not utilized by P-inefficient accessions. These data demonstrated that a P-efficient
accession has a specific mechanism to mobilize and use a fraction of soil that is not utilized
by other accessions of the same species. Hence, there is a scope for identifying P mobilizing
accessions through the use of isotopic dilution techniques. The results also showed that
accessions San Gregorio, M-1-04, and PICH-4 were the most responsive to high P application
indicated by the highest P-fertilizer use efficiency.

ACKNOWLEDGEMENTS

The authors are grateful to the International Atomic Energy Agency (IAEA), Vienna,
Austria for financial support under Research Contract MEX-13800/R0 of the Co-ordinated
Research Project “Selection and Evaluation of Food Crop Genotypes Tolerant to Low
Nitrogen and Phosphorus Soils through the Use of Isotopes and Nuclear-Related Techniques”.

REFERENCES

[1] ALCALÁ DE JESÚS, M., et al., Clasificación de los suelos de la Meseta Tarasca,
Michoacán, Terra 19 (2001) 217–239.
[2] LYNCH, J.P., Roots of the second green revolution, Aust. J. Bot. 55 (2007) 493–512.
[3] LYNCH, J.P., BROWN, K.M., “Root strategies for phosphorus acquisition”, The
Ecophysiology of Plant-Phosphorus Interactions (WHITE, P.J., HAMMOND, J.P.,
Eds), Springer, Netherlands (2008) 83–116.
[4] SCHAFFERT, R.E., et al., “Genetic control of phosphorus uptake and utilization
efficiency in maize and sorghum under marginal conditions”, Molecular Approaches
for the Genetic Improvement of Cereals for Stable Production in Water-Limited
Environments, (RIBAUT, J.M., POLAND, E., Eds), CIMMYT, El Batan, Mexico
(1999) 79–85.

52
[5] PANDEY, S., et al., Genetics of tolerance to soil acidity in tropical maize, Crop Sci.
34 (1994) 1511–1514.
[6] MANSKE, G.G.B., et al., Traits associated with improved P-uptake efficiency in
CIMMYT´s semidwarf bread wheat grown on an acid Andisol in Mexico, Plant Soil
221 (2000) 189–204.
[7] NIELSEN, K.L., et al., Effects of phosphorus availability and vesicular-arbuscular
mycorrhizas on the carbon budget of common bean (Phaseolus vulgaris L.), New
Phytol. 139 (1998) 647–656.
[8] ZHU, J., et al., Detection of quantitative trait loci for seminal root traits in maize (Zea
mays L.) seedlings grown under differential phosphorus levels, Theor. Appl. Genet.
113 (2006) 1–10.
[9] HINSINGER, P., Bio-availability of soil inorganic P in the rhizosphere as affected by
root induced chemical changes: a review, Plant Soil 237 (2001) 173–195.
[10] BATES, T., LYNCH, J.P., Root hairs confer a competitive advantage under low
phosphorus availability, Plant Soil 236 (2001) 243–250.
[11] RAGOTHAMA, K.G., “Phosphorus and plant nutrition: An overview”, Phosphorus
Agriculture and the Environment, (SIMS, J.T., Ed.), American Society of Agronomy
and Soil Science Society of America, Madison, Wisconsin (2005) 355–378.
[12] MILLER, C.R., et al., Genetic variation for adventitious rooting in response to low
phosphorus availability: potential utility for phosphorus acquisition from stratified
soils, Funct. Plant Biol. 30 (2003) 973–985.
[13] ZHU, J., et al., Topsoil foraging and phosphorus acquisition efficiency in maize (Zea
mays L.), Funct. Plant Biol. 32 (2005) 749–762.
[14] JUNGK, A., Root hairs and the acquisition of plant nutrients from soil, J. Plant Nutr.
Soil Sci. 164 (2001) 121–129.
[15] MANO, Y.M., et al., Identification of QTL controlling adventitious root formation
during flooding conditions in teosintle (Zea mays spp. huehuetenangensis) seedling.
Euphytica 142 (2005) 33–42.
[16] OCHOA, I.E., et al., QTL Analysis of adventitious root formation in common bean
(Phaseolus vulgaris L.) under contrasting phosphorus availability, Crop Sci. 46 (2006)
1609–1621.
[17] GAHOONIA, T.S., et al., Genetic variation in root traits and nutrient acquisition of
lentil accessions, J. Plant Nut. 29 (2006) 643–655.
[18] GAHOONIA, T.S., et al., Phosphorus (P) acquisition of cereal cultivars in the field at
three levels of P fertilization, Plant Soil 211 (1999) 269–281.
[19] SÁNCHEZ, G.J.J., et al., Isozymatic and morphological diversity in the races of maize
of Mexico, Econ. Bot. 54 (2000) 43–59.
[20] BELLON, M.R., RISOPOULUS, J., Small-scale farmers expand the benefits of
improved maize germplasm: A case study from Chiapas, Mexico, World Develop. 29
(2001) 799–811.
[21] VIEIRA, R.F., et al., Method for evaluation of root hairs of common bean genotypes.
Pesq. Agropec. Bras. 42 (2007) 1365–1368.
[22] HOCHHOLOLDINGER, F., et al., Genetic dissection of root formation in maize (Zea
mays) reveals root-type specific development programmers, Ann. Bot. 93 (2004) 359–
368.
[23] HUNT, R., Basic Growth Analysis, Unwin Hyman Ltd., London (1990).
[24] SAS INSTITUTE, SAS User´s Guide: Statistics, SAS Inst., Cary, NC (2000).
[25] PAN, X.W., et al., Assessment on phosphorus efficiency characteristics of soybean
genotypes in phosphorus-deficient soils. Agric. Sci. Chin. 7 (2008) 958–969.
[26] ROMERSBURG, H.C., Cluster Analysis for Researchers, Lifetime Learning,
Publications, Belmont, CA (1988).

53
[27] WISSUWA, M., AE, N., Genotypic variation for tolerance to phosphorus deficiency
in rice and the potential for its exploration in rice improvement, Plant Breed. 120
(2001) 43–48.
[28] FAGERIA, N.K., The role of nutrient efficient plants in improving crop yields in the
twenty first century, J. Plant Nutr. 31 (2008) 1121–1157.
[29] ZHU, J., LYNCH, J.P., The contribution of lateral rooting to phosphorus acquisition
efficiency in maize (Zea mays L.) seedlings, Funct. Plant Biol. 31 (2004) 949–958.
[30] MA, Z., et al., Regulation of root hair density by phosphorus availability in
Arabidopsis thaliana, Plant Cell Environ. 24 (2001) 459–467.
[31] ZHU, J., et al., Mapping of QTL controlling root hair length in maize (Zea mays L.)
under phosphorus deficiency, Plant Soil 270 (2005) 299–310.
[32] BALIGAR, V.V., et al., Nutrient use efficiency in plants, Commun. Soil. Sci. Plant
Anal. 32 (2001) 921–950.
[33] MANSKE, G.G.B., et al., Importance of P uptake efficiency versus P utilization for
wheat yield in acid and calcareous soils in Mexico, Eur. J. Agron. 14 (2001) 261–274.
[34] RAMAEKERS, L., et al., Strategies for improving phosphorus acquisition efficiency
of crop plants, Field Crops Res. 117 (2010) 169–176.
[35] WANG, X., et al., Acquisition or utilization, which is more critical for enhancing
phosphorus efficiency in modern crops? Plant Sci. 179 (2010) 302–306.
[36] BAYUELO-JIMÉNEZ, J.S., et al., Genetic variation for root traits of maize (Zea mays
L.) from Purhepecha Plateau, under contrasting phosphorus availability, Field Crops
Res. 121 (2011) 350–362.
[37] RITCHIE, S., HANWAY, J., How a Corn Plant Develops, Special Report No. 48,
Iowa State University of Science and Technology, Cooperative Extension Science,
Ames, Iowa (1993).
[38] MURPHY, J., RILEY, J.P., A modified single solution method for the determination
of phosphate in nature waters, Anal. Chim. Acta 27 (1962) 31–36.
[39] SEPEHR., et al., Genotypic variation in P efficiency of selected Iranian cereals in
greenhouse experiment, Int. J. Plant Prod. 3 (2009) 17–28.
[40 VESTERAGER, J.M., et al., Variation in phosphorus uptake and use efficiencies
between pigeonpea genotypes and cowpea, J. Plant Nutr. 29 (2006) 1869–1888.
[41] WISSUWA, M., AE, N., Further characterization of two QTLs that increase
phosphorus uptake of rice (Oryza sativa L.) under phosphorus deficiency, Plant Soil
237 (2001) 275–286.
[42] DRIESSEN, P., et al., Lecture Notes on the Major Soils of the World, FAO, Rome
(2001).
[43] IAEA, Use of Nuclear Techniques in Studies of Soil-Plant Relationships: Manual,
IAEA Training Course Series 2, International Atomic Energy Agency, Vienna, Austria
(1990) 220 p.
[44] FARDEAU, J.C., et al., The role of isotopic techniques in the evaluation of the
effectiveness of P fertilizers, Fert. Res. 45 (1996) 101–109.
[45] LARSEN, S. The use of 32P in studies on the uptake of phosphorus by plants, Plant
Soil 4 (1952) 1–10.
[46] MAGDALENO-ARMAS, M.L., Evaluación genotípica de la arquitectura radical de
maíz en etapa vegetativa temprana: Mecanismo adaptativo para eficiencia a fósforo en
un suelo ácido, B.S. Dissertation, Universidad Michoacana de San Nicolás de Hidalgo,
México (2010) 65 p.

54
 SELECTION AND EVALUATION OF MAIZE GENOTYPES TOLERANCE
TO LOW PHOSPHORUS SOILS

J.C. YANG, H.M. JIANG, J.F. ZHANG, L.L. LI, G.H. LI

Institute of Agricultural Resources and Regional Planning,
Chinese Academy of Agricultural Sciences,
Beijing,
China

Abstract

Maize species differ in their ability to take up phosphorus (P) from the soil, and these
differences are attributed to the morphology and physiology of plants relative to their germplasm base.
An effective method of increasing P efficiency in maize is to select and evaluate genotypes that can
produce a high yield under P deficient conditions. In this study, 116 maize inbred lines with various
genetic backgrounds collected from several Agricultural Universities and Institutes in China were
evaluated in a field experiment to identify genotypic differences in P efficiency in 2007. Overall, 15
maize inbred lines were selected from the 116 inbred lines during the 5-year field experimental period
based on their 100-grain weight in P-deficient soil at maturity, when compared to the characteristics
exhibited in P-sufficient soil. All of the selected lines were evaluated in field experiments from 2008
to 2010 for their tolerance to low-P at the seedling and maturity stages. Inhibition (%) was used and
defined as the parameter measured under P limitation compared to the parameters measured under P
sufficiency to evaluate the genotypic variation in tolerance. Inhibition of root length, root surface area,
volume, root: shoot ratio and P uptake efficiency could be used as indices to assess the genotypic
tolerance to P limitation. Low-P tolerant genotypes could uptake more P and accumulate more dry
matter at the seedling stage. A strong relationship between the total biomass and root length was
exhibited. In order to understand the mechanisms of the genotypic tolerance to low-P soil to utilize P
from the sparing soluble P forms, 5 maize genotypes selected out of the 15 maize inbred lines,
according to the four quadrant distribution, was used as the criteria in a 32P isotope tracer experiment
to follow the recovery of 32P in soil P fractions. The 32P tracer results showed a higher rate for water-
soluble P transformation to slowly available P in P deficient soil than in soil with sufficient P. The L-
values showed that different genotypes had different soil P-use efficiency and low-P tolerance
mechanisms. A low-P tolerant cultivar DSY-32 regulated soil P use efficiency and plant P content
according to the L-value under exogenous P fertilizer application. However, another low-P tolerant
cultivar, DSY-2, utilized soil P more efficiently, regardless of the L-value. In conclusion, the study
exploited the physiological-biochemical mechanism on P-uptake, and P-transport of selected maize
genotypes in low-P soil under field conditions, and the 32P tracer technique proved to be a valuable
tool that sought physiological explanations for superior genotype performance.

1. INTRODUCTION

Phosphorus (P) is one of the essential nutritional elements for all organisms. However,
P deficiency in many regions of the world, especially in developing countries, limits plant
growth and crop productivity [1, 2]. Plants are almost entirely dependent on P absorbed from
soil [3], and while total phosphate is quite abundant in many soils, it is largely unavailable to
plants (means “P-deficiency in heredity”) [4]. In fact, P utilization efficiency is only 15% and
no more than 25%, even including subsequent utilization by plants [5]. To increase crop
productivity, farmers often apply substantial amounts of phosphate fertilizers to compensate
for the low-P use efficiency of crops [6]. However, applying fertilizers in large amounts not
only exhausts limited P resources, but also leads to environmental pollution. Therefore,
development of low-P tolerant cultivars may help reduce the demand for P, since numerous
studies have shown that differences in tolerance to low-P stress exist among genotypes within
crop species [7].

55
 Maize is an important grain and forage crop worldwide after wheat and rice, which is
also sensitive to P and subjected to P-deficiency in heredity [8]. Due to population growth and
the generally low level of available P in soils, it is difficult to increase global maize
production by extending the area of cultivated land. As a result, a complementary approach
that has received more attention is the adaptation of crops to unfavourable soil conditions by
selecting, and evaluating genotypes with enhanced nutrient use efficiency in soils with a low
nutrient status, and/or requiring moderately low external inputs to induce expression of their
genetic potential for adequate production. This strategy is now considered a promising,
energy-efficient, eco-friendly and socio-economically feasible approach [9]. Recently, several
researchers have begun screening and improving the tolerance to P deficiency in maize
cultivars [3, 7, 10], but most of these were carried out under hydroponic or pot experimental
conditions. Due to differences in the growing environment between greenhouse and field
conditions, cultivars may vary greatly in response to low-P supply, and P efficiency may
differ among maize inbred lines. Therefore, special attention should be given to screening
maize genotypes that are tolerant to low-P soils under field conditions. Although the
workloads are large and there are a great number of uncontrollable conditions in field
experiments, the results of such studies more closely approximate the actual conditions. Plants
differ greatly in their ability to grow on low-P soils, because they have developed specific
physical-chemical and biological mechanisms to utilize P compounds in such soils [11].
Quantifying soil P into different pools could identify the availability of these fractions to the
plant. Naruzzaman et al. [12] have shown genotype variation of crops in their ability to access
and utilize sparingly soluble forms of soil P (Ca-P, Al-P, Fe-P), which has been proposed as a
possible means for overcoming P deficiency stress and to optimize P fertilizer use where P is
poorly available.

In a pre-screening experiment, 116 maize inbred lines with various genetic

backgrounds collected from several Agricultural Universities and Institutes in China were
employed in a field experiment in 2007, which resulted in selection of 15 maize inbred lines
by evaluation in a two year field experiment during 2008 and 2009, to identify an index
system for assessment of low-P tolerant maize genotypes. To verify that P uptake from
sparingly soluble P forms by different maize genotypes can be used as a criterion to evaluate
maize tolerance to low available P soils, 5 lines were selected out of the 15 maize inbred lines
and used in a 32P isotopic tracer experiment to determine recovery of 32P in different soil P
fractions in 2010.

2. METHODOLOGY

2.1. Experiment 1: Field experiment

2.1.1. Experimental site

A study was carried out in 2007, 2008 and 2009 at Langfang experimental station,
located in Hebei Province (E 116°35'; N 39°36'). The frost period at this site extends to 182
days, mean annual sunshine duration is 2660 hours and mean annual evapotranspiration is
1909.6mm. Annual precipitation is 554.9 mm, 70-80% of which distributed during June to
August. The annual mean temperature is 11.9oC, the mean monthly maximum 26.2oC, in July
and minimum –4.7oC, in January. Soil chemical characteristics were pH 8.5, available P 4.91
mg kg-1, available N 55.8 mg kg-1, available K 92.6 mg kg-1 and organic matter 14.38 g kg-1.
The soil was characterized as typically P deficiency.

56
2.1.2. Plant materials

A total of 116 maize inbred lines with various genetic backgrounds were collected
from several Agricultural Universities and Institutes in China and employed in field
Experiment 1 in 2007. Overall, 15 maize inbred lines were selected from Experiment 1 based
on the difference in their 100-grain weight when grown in P-deficient soil and P-sufficient
soil at the mature stage. Specifically, the low-P tolerant genotypes DSY-30, DSY-2, DSY-31,
DSY-20, DSY-21, DSY-39, DSY-101, DSY-33, DSY-32, DSY-23 and DSY-93, and the low
P sensitive genotypes, DSY-113, DSY-79, DSY-129 and DSY-48 were selected for
experiments 2 and 3, respectively, in 2008 and 2009.

2.1.3. Field management

Maize was sown on 25th April 2007, 5th May 2008 and 13th May 2009, respectively.
Row width was 50 cm, and distance between plants within a row was 30 cm, resulting in an
overall planting density of 12 plants m-2 with pest control carried out as required. Water was
applied by sprinkler irrigation as needed.

2.1.4. Experimental design

Two treatments with different P application levels of 0 and 52 kg P ha -1 corresponded

to the treatments of P0 and P1; i e, P-deficiency and P-sufficiency, a randomized complete
block design with three replication. Urea, superphosphate, and potassium chloride were
applied with N (225 kg−N ha-1 in total, 1/2 N of which was applied basally, while the
remaining 1/2 is applied as top-dressing during the bell-mouthed stage), P (applied basally)
and K (87 kg K ha-1 applied basally). To prevent impurity, maize cover package pollination
was utilized.

2.1.5. Measurement and statistical analyses

Ten representative plants (48 plants per treatment) were sampled at maturity, and the
100-grain weights of the maize inbred lines were measured in 2007. During 2008 and 2009,
plants were harvested at both the seedling and maturity stages. Plants collected during the
seedling stage were separated into roots and shoots to assess morphological traits, dried at
65°C in a forced-air oven for 48 h and weighed to determine the root and shoot dry weight.
The collected roots were placed in a transparent water filled tray (15 × 25 cm) on a scanner
(EPSON PERFECTION 4990 PHOTO, Model J1318) to facilitate root spreading. The system
was used to scan all fine root fragments and the program calculated root traits such as root
length, root volume and root surface area for all fine roots. The data obtained were then
analysed using the digital image analysis system, WinRHIZO Pro 2007a. Plants harvested at
the mature stage were divided into roots and shoots and then dried at 65°C in a forced-air
oven for 48 h, weighed and the 100-grain weight determined. P nutrient efficiency methods
[13] indicated that resource capture and P efficiency were highly interrelated. In this study, P
uptake efficiency was measured as the amount of P in the plant materials, including the roots
and shoots at the seedling stage (mg P plant -1). The P content in the plant material was
determined using the Mo-Sb-Vc colorimetric method [14] after digestion in sulphuric acid-
perchloric acid [15].

The inhibition (%) due to P limitation was defined as follows:

Inhibition (%) = {(P1 – P0) / P1} x 100

57
 Root: shoot ratio was calculated as (root dry weight in g plant -1) / (shoot dry weight in
g plant-1).

2.2. Experiment 2: 32P isotopic tracer experiment

2.2.1. Soil and plant materials

The soil was air-dried, ground and passed through a 2 mm sieve. Each pot contained 2
kg of the soil. Based on the results of a previous 3-year field experimental selection, 5 typical
cultivars were employed: genotypes DSY-30 and DSY-2, responsive and tolerant to low-P
soil (named as low-P tolerant genotype); DSY-32, non-responsive and tolerant (named as
low-P tolerant genotype); DSY-48, non-responsive and sensitive and DAY-79, responsive and
sensitive (named as low-P sensitive genotype).
32
2.2.2. P carrier-free solution preparation

A 32P carrier-free solution was prepared (100 ml) with a specific radioactivity of 185
MBq ml-1. Four ml (7.4 MBq) of the solution was added individually to the soil of each pot,
and the solution and soil were homogenized. The pots were covered with Al foil to avoid
light, for equilibration and the analysis of P fractionations.

2.2.3. Treatments

To evaluate the P pool of low-P soil for different maize genotypes, the pots were
separated into two groups: one group was planted and the other group was without plants.
Each group included treatment 1 (the P0 treatment), with no P application but 200 mg N kg-1
and 166 mg K kg-1, and treatment 2 (the P1 treatment), which included 66 mg P kg-1, 200 mg
N kg-1 and 166 mg K kg-1. Each treatment was applied in 3 replications. One week after
preparing the pots, they were placed in a greenhouse. Soil samples were taken on days 0, 3, 7,
14 and 25 , and plant samples were taken only on day 25.

2.2.4. Measurements

2.2.4.1. Inorganic P fractionations

The sequential P fractionation of Ca2-P, Ca8-P, Al-P, Fe-P, O-P (occluded P) and
Ca10-P was performed according to [16].

Ca2-P: One g of an air-dried soil sample was passed through a 100-mesh sieve and put
into a 100 ml centrifugation tube. Fifty milliliters of NaHCO3 (0.5 mol l-1, pH = 7.5) was
added. The sample was shaken for 1 h at 20-25°C and centrifuged at 3500 rev. min-1 for 8
min. The upper solution of the extract was collected in a 50-ml flask for Ca2-P analysis.

Ca8-P: The soil was washed twice with 25 ml of 95% ethanol after the extraction of
Ca2-P, and 50 ml CH3COONH4 (0.5 mol l-1, pH =4.2) was added. The soil was dispersed
homogenously and put aside for 4 h. The soil was shaken for 1 h at 20-25°C and centrifuged
at 3500 rev. min-1 for 8 min. The upper solution of the extract was collected in a 50 ml flask
for Ca8-P analysis.

Al-P: One g of an air-dried soil was passed through a 100-mesh sieve and placed in a
100-ml centrifugation tube. Fifty ml of NH4Cl (1 mol l-1] was added, and the sample was
shaken for 30 min at 20-25°C. The sample was centrifuged at 3500 rev. min-1 for 8 min. The
upper solution was discarded, and 50 ml NH4F (0.5 mol l-1, pH = 8.2) was added. The solution

58
was shaken for 1 h at 20-25°C, and centrifuged at 3500 rev. min-1 for 8 min. The upper
solution of the extract was collected in a 50-ml flask for Al-P analysis.

Fe-P: The soil was washed twice with 25 ml of a saturated NaCl solution after the
extraction of Al-P. The upper solution was discarded after centrifugation at 3500 rev. min-1
for 8 min. Fifty ml of NaOH (0.1 mol l-1] was added. The sample was shaken for 2 h at 20-
25°C and put aside 16 h. The sample was shaken for 2 h at 20-25°C once more, and
centrifuged for 10 min at 4500 rev. min-1. The upper solution was collected in a flask, and 1.5
ml H2SO4 was added. The solution was homogenized by shaking and set aside overnight. The
solution was filtered, and the filtrate was collected in a flask for Fe-P analysis.

O-P: The soil was washed twice with 25 ml of a saturated NaCl solution after the
extraction of Fe-P. The upper solution was discarded after centrifugation for 8 min at 3500
rev. min-1. Forty ml of Na3C6H5O7·2H2O (0.3 mol l-1) was added, and the sample was stirred
until the soil was dispersed homogenously. One g of Na2S2O4 was added, and the solution
was placed in a bain-marie at 80-90°C, until the temperature of the solution inside of the tube
and the temperature of the water outside were balanced. The solution was stirred for 15 min,
and 10 ml NaOH (0.5 mol l-1) was added. The solution was stirred for 10 min and centrifuged
for 10 min at 4500 rev. min-1. The upper solution was collected in a 100 ml volumetric flask.
The soil was washed twice with 20 ml of a saturated NaCl solution and added to the solution
after centrifugation. The volume of the extract was kept constant at 100 ml with distilled
water. Ten ml of the extract was removed and placed into a 50 ml flask. Ten ml of a mixed
acid solution (H2SO4: HClO4: HNO3 = 1: 2: 7, in volume) was added, and the solution was
put into a small funnel. The solution was boiled until HNO3 and HClO4 totally decomposed
and a back-flow of H2SO4 appeared. The crystalline solid was boiled and dissolved with 50
ml of distilled water after the solution was cooled. The extract was collected, and the volume
was made constant at 100 ml with distilled water for O-P analysis.

Ca10-P: Fifty ml of H2SO4 (0.5 mol l-1] was added after the extraction of O-P. The
solution was shaken for 1 h at 20-25°C and centrifuged for 10 min at 4500 rev. min-1. The
upper solution was collected in a flask for Ca10-P analysis. The P content in all extracted
fractions was determined using the MO-SB colorimetric method.

2.2.4.2. Measurement of 32P radioactivity

32
P radioactivity in the fractions of Ca2-P, Ca8-P, Al-P, Fe-P, O-P and Ca10-P was
measured using a liquid scintillation counter (LS-6500, BECKMAN, USA). The maize L-
value (µg P g-1 soil) was calculated according the followed formula:

L-value = (32P Bq g-1 soil) / (32P Bq µg-1 P in plant)

2.3. Data statistics

All data are presented as the means of three replicates with standard errors.
Differences between treatment means were compared by the least significant difference
(LSD) test at P<0.05.

3. RESULTS

3.1. Experiment 1: Field experiment

3.1.1. Screening of maize genotypes tolerant to low-P soil

59
 The frequency distribution of the inhibition (%) of the 100-grain weight of 116 maize
cultivars under P limitation (P0) compared with P sufficient (P1) is shown in Fig. 1. Overall,
the results were found to be normally distributed (KS test, Z value = 0.928, P value = 0.356).
Maize genotypes tolerant to low-P soils were defined according to the distribution of
inhibition that increased under low-P limitation, and a change in the inhibition that was lower
than –5%. Low-P sensitive genotypes were selected when the inhibition was higher than 10%.

Four quadrant analyses for the selected typical maize genotypes showed that tolerance
to low-P was primarily distributed in the first (responsive and efficient) and fourth quadrants
(non-responsive and efficient). However, low-P sensitive maize genotypes were mainly
distributed in the second (responsive and inefficient) and third quadrants (non-responsive and
inefficient) (Fig. 2). Based on the depression of the 100-grain weight and the four quadrant
analyses, the following typical 15 inbred lines were selected out of the 116 maize varieties for
further analysis: genotypes tolerant to low-P, DSY-30, DSY-2, DSY-31, DSY-20, DSY-21,
DSY-39, DSY-101, DSY-33, DSY-93, DSY-23 and DSY-32; genotypes sensitive to low-P ,
DSY-113, DSY-79, DSY-129 and DSY-48.

Genotypes tolerant to low-P produced higher 100-grain weight than low-P sensitive
genotypes under low-P limitation in the field experiments 1, 2 and 3 (Fig. 3). 100-grain
weight of DSY-30 (genotype tolerant to low-P) had the highest value, 39.2 g, 37.4 g and 38.4
g in the three years of the experiment, respectively. It was significantly higher than DSY-113,
DSY-79, DSY-129 and DSY-48 (genotypes sensitive to low-P) (P<0.05). 100-grain weight of
DSY-48 was significantly lower than other maize genotypes in the three years of the
experiment (P<0.05).

3.1.2. Root architectural traits at seedling stage in field experiments 2 and 3

3.1.2.1. Adventitious root angle

Root architectural plasticity might be an important factor in the acquisition by plants

of immobile nutrients such as P. The results of two field experiments suggested that
adventitious root angle correlated with adaptation to low-P. Adventitious root angles of maize
genotypes under the P0 treatment were more or less pronounced than those under the P1
treatment (Fig. 4).

Effect of limiting P on the adventitious root angles of typical maize genotypes were
showed in Fig. 5. Adventitious root of low-P sensitive genotypes had deeper growth angles
than low-P tolerant genotypes in the field experiment 3 (Fig. 5B). Adventitious root angles
DSY-30, DSY-2, DSY-21, DSY-33, DSY-93, DSY-20 and DSY-79 in low-P tolerant
genotypes were lower than mean value (56.38°). Our results demonstrated that variation for
adventitious root angles existed in maize, which P could modulate root shallowness
independently, and that a shallow root system was beneficial for plant performance in maize
at low-P limitation.

3.1.2.2. Root length, root surface area, root volume

Root morphological parameters such as decreased root length, root surface area and
root volume are shown in Table 1. In 2008, increased root lengths were observed for maize
varieties grown under low-P limitation. Specifically, 64% of the genotypes that were tolerant
to low-P showed increased root lengths under low-P limitation, while only 25% of the
genotypes that were sensitive to low-P showed increased root length under low-P treatment.
The inhibition of DSY-93 was significantly higher than that of other genotypes. The total root

60
surface area and root volume of low-P tolerant genotypes were higher than those of low-P
sensitive genotypes under low-P limitation, except for DSY-113. These findings indicated
that low-P tolerant genotypes could accelerate root growth under low-P conditions, resulting
in roots exploring more soil volume for the uptake of nutrients.

25

20
F re q u e n c y

15

10

Me a n = -1 . 0 6 6 1
S td . D e v. =
0 1 0 .9 8 6 7 7
N = 116
-4 0 . 0 0 -2 0 .0 0 0 .0 0 2 0.00 4 0.00

In h ib itio n o f 1 0 0 - g ra in we ig h t (% )
FIG. 1. Frequency distribution of the inhibition of the 100-grain weight of 116 maize inbred lines in
field experiment.

Our data also showed that Inhibition of root length ranged from –44.0 to 47.7% in
2009. The total root length of low-P tolerant genotypes increased dramatically under P-
deficient conditions such as DSY-30, DSY-2, DSY-31, DSY-21, DSY-39, DSY-33 and DSY-
32 that developed larger root lengths than low-P sensitive genotypes. The inhibition of root
length in DSY-21 (low-P tolerant genotype) was significantly lower than those of other
genotypes, which indicated that low-P tolerant genotypes developed larger root length under
low-P limitation. On the contrary, low-P sensitive genotypes developed smaller root lengths
under P deficient conditions compared to normal P conditions. The above results indicated
that root length of low-P tolerant genotypes were essential for their high ability to acquire soil
P. Comparing the inhibition of total root surface area and that of total root volume between
low-P tolerant and sensitive genotypes, the results showed that the inhibition of root surface
area and total root volume ranged from –45.5 to 47.1%, and –49.6 to 58.3%, respectively. The
inhibitions were lower in low-P tolerant genotypes than in low-P sensitive genotypes.

61
 40 Responsive and inefficient Responsive and efficient
100-grain weight at normal P (g) DSY-30

35 DSY-23 DSY-33
DSY-2

DSY-21
30 DSY-101
DSY-20 DSY-39

DSY-113 DSY-93
DSY-79
25 DSY-31

DSY-129
DSY-32
20

DSY-48
Non-responsive and inefficient Non-responsive and efficient
15 20 25 30 35 40 45
100-grain weight at low P (g)
FIG. 2. 100-grain weight of typical inbred lines of maize under P0 and P1 treatments in field
experiment 1.

TABLE 1. CHANGES IN ROOT MORPHOLOGICAL CHARACTERISTICS OF SEEDLINGS IN

FIELD EXPERIMENTS 2 AND 3

Genotype No. Inhibition (%) of

Root length Root surface area Root volume
2008 2009 2008 2009 2008 2009
P tolerant DSY-30 –6.9 –6.7 –16.1 –12.2 –25.6 –20.0
DSY-2 –56.7 –41.6 –10.1 –45.5 22.1 –49.6
DSY-31 11.0 –29.0 18.1 –28.5 25.0 –27.8
DSY-21 –17.3 –44.0 5.5 17.3 23.8 52.5
DSY-39 –7.4 –19.2 8.3 –14.6 21.2 –10.0
DSY-101 –1.1 29.7 –5.4 27.9 –9.9 26.0
DSY-33 50.3 –13.0 34.4 –10.7 13.0 –8.3
DSY-32 34.1 1.4 40.0 –8.0 45.0 –18.0
DSY-23 11.8 47.7 5.3 50.4 –1.0 53.0
DSY-93 –127.4 22.0 –53.2 46.5 –3.0 54.3
DSY-20 –45.5 14.9 –47.6 –1.1 –49.7 –20.1
P sensitive DSY-113 –68.9 34.8 –56.4 41.0 –45.6 46.7
DSY-79 17.6 32.9 26.7 32.2 34.5 31.5
DSY-129 28.2 24.5 22.7 29.1 16.7 33.4
DSY-48 2.3 32.9 3.3 47.1 4.4 58.3

62
 50 50
A P tolerant
B P tolerant
a P sensitive
a P sensitive
40 40 a
b b b ab
b ab ab
b b ab ab
bc bc b
30 c c 30 b b b b
c
bc
d d
20 d d 20 c

100-grain weight (g)

100-grain weight (g)

10 10

0 0

DSY-2

DSY-2
DSY-30
DSY-31
DSY-21
DSY-39
DSY-33
DSY-32
DSY-23
DSY-93
DSY-20
DSY-79
DSY-48

DSY-30
DSY-31
DSY-21
DSY-39
DSY-33
DSY-32
DSY-23
DSY-93
DSY-20
DSY-79
DSY-48
DSY-101
DSY-113
DSY-129

DSY-101
DSY-113
DSY-129
Genotype Genotype

50
C P tolerant
a P sensitive
a
40 ab
ab
ab b
30 b b b
b c
c c
c
20
d

100-grain weight (g)

10

DSY-2

DSY-30
DSY-31
DSY-21
DSY-39
DSY-33
DSY-32
DSY-23
DSY-93
DSY-20
DSY-79
DSY-48

DSY-101
DSY-113
DSY-129

Genotype
FIG. 3. Effects of low-P limitation on 100-grain weight of typical maize genotypes in different years. Each column represents the mean of three plants ± SD.
A: the field experiment 1 (in 2007), B: the field experiment 2 (in 2008), C: the field experiment 3 (in 2009).

63
 64
Adventitious root angle ( 0 )
0
10
20
30
40
50
60
A 70

D S Y-30
D S Y-2
D S Y-31
D S Y-21
D S Y-39
D S Y-101
D S Y-33
D S Y-32
D S Y-23
Genotype

D S Y-93
D S Y-20
D S Y-113
D S Y-79
D S Y-129
D S Y-48
P1
P0
B

Adventitious root angle (0)

± SD. A: the field experiment 2 (in 2008), B: the field experiment 3 (in 2009).
00
20
40
60
80

DSY-30
DSY-30
DSY-2
DSY-2
DSY-31
DSY-31
DSY-21
DSY-21
DSY-39
DSY-39
DSY-101
DSY-101
DSY-33
DSY-33
DSY-32
DSY-32
DSY-23
DSY-23

Genotype
DSY-93
DSY-93
DSY-20
DSY-20
DSY-113
DSY-113
DSY-79
DSY-79
DSY-129
DSY-129
DSY-48
P1
P0

DSY-48
P1
P0

FIG. 4. Adventitious root angle of typical maize genotypes under low-P limitation and normal P conditions. Each column represents the mean of three plants
 A
B

FIG. 5. Adventitious root angle of typical maize genotypes under low-P limitation. Red line: the mean value of adventitious root angle. Each column represents
the mean of three plants ± SD. A: the field experiment 2 (in 2008), B: the field experiment 3 (in 2009).

65
3.1.3. P nutrient characteristics of maize genotypes at the seedling stage in the field
experiments 2 and 3

3.1.3.1. Root: shoot ratio

As expected, the root/shoot ratio increased in response to P deficiency (Fig. 6). The
reason for this phenomenon is that P deficiency during the early stage affected the plant
growth, resulting in greater restriction of the shoots than the roots because carbon dioxide was
converted to carbohydrates through photosynthesis and transported from the shoots to the
roots. The root/shoot ratio of low-P tolerant genotypes was higher under low-P limitation
condition than that of low-P sensitive genotypes, whatever in 2008 or 2009. These findings
suggested that the Inhibition of the root/shoot ratio was more stable during the seedling stage.

There was a close relationship between the plant’s P uptake and the soil P supply
level. P uptake efficiency of different maize genotypes were lower under P deficiency as
compared with that under P sufficiency in both field experiments (Fig. 7). Under P deficiency,
P uptake efficiency in DSY-21 and DSY-23 (low-P tolerant genotypes) was significantly
higher than that DSY-113 (low-P sensitive genotype) in the field experiment II (Fig. 7A). In
comparison with low-P sensitive genotypes, the low-P tolerant genotypes were of higher P
content under P deficiency, indicating that P uptake ability of low-P tolerant genotypes were
relatively stronger in the field experiment III (Fig. 7B). P uptake efficiency in DSY-30, DSY-
2, DSY-39, DSY-101, DSY-93 (low-P tolerant genotypes) was significantly higher than
DSY-113, DSY-79, DSY-129, DSY-48 (low-P sensitive genotypes). The Inhibition of P
uptake efficiency in low-P tolerant was much lower than most of low-P sensitive genotypes in
the field experiment II and III, indicating that P uptake ability of low-P tolerant was relatively
strong under low-P limitation.

3.1.3.2. Relationship between root length and biomass

Correlation analysis was performed to reveal relationship with root length and the total
biomass (Fig. 8). Under P-deficiency condition, root length was significantly correlated with
total biomass. The correlation coefficient was 0.61 (P<0.05). It was obvious that the biomass
of maize genotype for tolerance to low-P limitation had the tendency to associate with this
morphological characteristic. Otherwise, it may be used as a fast screening protocol to
evaluate the maize lines at the seedling stage under low-P limitation in the field evaluation.

3.2. Experiment 2: 32P isotope tracer

3.2.1. Selecting plant materials

To study the nutrient acquisition from sparingly soluble P forms and determine L-
values, the 32P tracer technique was used [17]. Based on 100-grain weight of three field
experiments (2007, 2008 and 2009), DSY-30, DSY-2, DSY-32 were selected as low-P
tolerant genotypes, and DSY-48 and DSY-79 were selected as low-P sensitive genotypes for
the tracer experiment (Fig. 9, red circle: maize inbred lines).

66
67
A

Inh ibitio n o f R o ot/S h o ot (% )

-140
-120
-100
-80
-60
-40
-20
0
20
40

D S Y -3 0
D S Y -2
D S Y -3 1
D S Y -2 1
D S Y -3 9
D S Y -10 1
D S Y -3 3
D S Y -3 2
D S Y -2 3
Genotype

D S Y -9 3
D S Y -2 0
D S Y -11 3
D S Y -7 9
D S Y -12 9
P tolerant
P sensitive

D S Y -4 8
B

Inhibition of R oot/Shoot (%)

-140
-120
-100
-80
-60
-40
-20
0

DSY-30
DSY-2
DSY-31
DSY-21
DSY-39
DSY-101
DSY-33
DSY-32

FIG. 6. Inhibition of root: shoot ratio. A: the field experiment 2 (in 2008), B: the field experiment 3 (in 2009).
DSY-23

Genotype
DSY-93
DSY-20
DSY-113
DSY-79
DSY-129

P tolerant
P sensitive
DSY-48
68
B 60
A 70 P0 A
P0 P1
ABC A
60 P1 50
AB AB
A ABCD A A AB C
50 AB 40
ABCD C BC ABC
ABC ABC ABCD a ab bcdabc ab
40 C C
cde CD CD
30 de de CDE e
BCD de
30 a
CD ab D ab DEFEF
ab a 20 fg F f fg
ab g fg
20 ab ab ab
ab ab
b ab ab
b 10
10

0 0
80
P tolerant P tolerant
80
P sensitive P sensitive
60

60
40

20
40

20
-20

Inhibition of P uptake efficiency (%) P uptake efficiency (mg/plant)

-40

Inhibition of P uptake efficiency (%) P uptake efficiency (mg/plant)

0
DSY-2

DSY-2
DSY-30
DSY-31
DSY-21
DSY-39
DSY-33
DSY-32
DSY-23
DSY-93
DSY-20
DSY-79
DSY-48

DSY-30
DSY-31
DSY-21
DSY-39
DSY-33
DSY-32
DSY-23
DSY-93
DSY-20
DSY-79
DSY-48
DSY-101
DSY-113
DSY-129

DSY-101
DSY-113
DSY-129
Genotype Genotype

FIG. 7. P uptake efficiency and inhibition of typical maize inbred lines. A: the field experiment 2 (in 2008), B: the field experiment 3 (in 2009).
Note: Different uppercase or lowercase letters indicate significant differences (P<0.05).
 1.2

log BM -P0 1.0

0.8

0.6 y = 0.5696x - 0.7968

R² = 0.3726 (P <0.05)

0.4
2.4 2.6 2.8 3.0 3.2
log Root length-P0
FIG. 8. Relationship between the root length and biomass at the seedling stage.

3.2.2. Dynamic variations of the P fractionations

To better understand the transformation dynamics of fertilizer P to different P

fractions, the 32P tracer technique was employed. Without P fertilizer, the 32P activity in Ca2-P
decreased rapidly during the early period, but then decreased slowly. However, a reverse
trend was observed in Ca8-P, i.e., a rapid increase during the early period followed by a slow
increase. The 32P activity in Fe-P and Al-P increased steadily at a moderate rate and reached a
steady state. However, the 32P activity in O-P and Ca10-P increased slowly and steadily during
the entire test period of 25 d (Fig. 10A). When P fertilizer was added to the soil, the
transformation dynamics of 32P in all P fractions in the soil were consistent with the
transformation dynamics without P fertilizer (Fig. 10B). However, the 32P activity in different
P fractions showed different characteristics. First, the 32P activity in Ca2-P was significantly
higher than activity without P fertilizer. Second, the 32P activity in slowly available P (Ca8-P,
Al-P and Fe-P) and unavailable P (O-P and Ca10-P) was significantly lower than without P
fertilizer. These results suggested that the rate of water-soluble P that transformed to slowly
available and unavailable P in the soil with deficient P was higher than in the soil with
sufficient P. The possibility of water-soluble P transformation to unavailable P (such as O-P)
in the soil with sufficient P was lower than in the soil with deficient P.

69
70
FIG. 9. 100-grain weight of typical inbred lines of maize under P0 and P1 treatments in field experiment 1, 2 and 3. A: the field experiment 1(in 2007), B: the
field experiment 2 (in 2008), C: the field experiment 3 (in 2009).
 FIG. 10. Temporal changes in the 32P specific activity of the inorganic P fractions. A: no P-fertilizer; B: with P-fertilizer.

71
3.2.3. Inorganic P and L-value

This experiment was performed using radioactive 32P to estimate the plant available
soil P. The L-value has considerable theoretical advantages as a measurement of plant
available soil P. As shown in Tables 2 and 3, the DSY-30 and DSY-2 cultivars had higher L-
values and intermediate P consumption, which indicated they activated soil P in different
forms, and absorbed it under low-P limiting conditions. These cultivars showed a slight
dependency on insoluble and soluble P; therefore, DSY-30 and DSY-2 were likely to be the
low-P tolerant genotypes.

The L-value of DSY-32 was the highest under low-phosphorus limitation, which
indicated that DSY-32 activated and utilized different Pi forms in soil, and reduced the
dependence on soil soluble P (Table 2). After P fertilizer was applied, the L-value of DSY-32
decreased rapidly (Table 3), which suggested that more P from fertilizer was utilized by DSY-
32. Therefore, DSY-32 was a typically low-P tolerant genotype.

TABLE 2. L-VALUES OF MAIZE GENOTYPES WITHOUT EXTERNAL P FERTILIZER

Variety Specific activity of L-value

Plant (Bq µg-1) Soil (Bq g-1 soil) (µg P g-1 soil)
DSY-30 22.4 ± 2.5† 4216 ± 144 190 ± 28
DSY-2 20.0 ± 0.1 4484 ± 154 225 ± 7
DSY-32 7.8 ± 1.0 3965 ± 136 513 ± 45
DSY-79 11.4 ± 0.8 3572 ± 122 315 ± 32
DSY-48 64.9 ± 2.9 3623 ± 124 56 ± 1
†Values following means are ± standard errors

TABLE 3. L-VALUES OF MAIZE GENOTYPES WITH EXTERNAL P FERTILIZER

Variety Specific activity of L-value

Plant (Bq µg-1) Soil (Bq g-1 soil) (µg P g-1 soil)
DSY-30 16.8 ± 0.9† 4015 ± 339 239 ± 32
DSY-2 14.0 ± 1.0 3517 ± 120 252 ± 26
DSY-32 20.4 ± 1.7 3793 ± 130 251 ± 5
DSY-79 8.3 ± 0.4 3821 ± 131 460 ± 17
DSY-48 28.8 ± 2.7 5089 ± 312 134 ± 36
†Values following means are ± standard errors

Although the L-value of DSY-79 was higher than DSY-30 and DSY-2, the
radioactivity of the shoot was also higher, indicating that DSY-79 required much higher P
content in the plant to maintain normal crop growth. Therefore, DSY-79 was a typical low-P
sensitive genotype. The L-value of DSY-48 was significantly lower than the other maize
genotypes with or without external P fertilizer. One probable explanation for this result is that
DSY-48 had a high dependency on soil available P and was sensitive to the supply of soil
available P.

4. DISCUSSION

The primary objective of most maize breeding programs is the evolution of high
yielding and well adapted cultivars. Breeding for improved varieties is a continuous process
and requires, primarily, a thorough knowledge of the genetic mechanism governing yield and
yield components [18]. Tang et al. [19] reported that grain yield had positive significant
correlation with 100-grain weight (P<0.05). Kanaka [20] also reported that over dominance

72
was present for 100-grain weight. The 100-grain weight was also likely to play a major role in
the tolerance to low-P limitation, and could also be used as the best parameter to evaluate P
efficiency in maize [21], which was supported by our data. 100-grain weight of low-P tolerant
genotypes were higher than DSY-113, DSY-79, DSY-129 and DSY-48 (low-P sensitive
genotypes) under low-P limitation in our study, which suggested that 100-grain weight could
be an indicator for screening tolerant maize genotypes. Similar ‘specific mechanism(s)’ has
been described [22]: a truly tolerant genotype requires less nutrient than a sensitive genotype
for normal metabolic processes. Therefore, a genotype with a small decrease in yield with a
decrease in P supply is more tolerant to P limitation than one with a large decrease, provided
the compared germplasm could achieve similar yield when sufficient P is available.

Root architectural plasticity may be an important factor in the acquisition of immobile

nutrients such as P by plants. It was well known that plant tolerance to P limitation is closely
related with root characteristics [23]. A study conducted by [24] found that adventitious
rooting enhanced P acquisition with shallow growth angles. These early reports were
supported by our results: that an adventitious root angle may be correlated with low-P
adaptation. Adventitious root angles of maize genotypes were observed more frequently in the
P0 groups than the P1 groups. This is likely because P availability is greatest near the surface
in most natural and agricultural soils [25]. The results also showed that low-P tolerant
genotypes had shallower adventitious root angles than low-P sensitive under low-P limitation,
which indicated that maize genotypes with shallower roots had greater growth in low-P soil
than deep-rooted genotypes [26].

With less mobile ions like P, uptake was often closely related to root length [27].
There have been a number of reports on root length under P-deficient conditions. Root
elongation by P deficiency was observed in Arabidopsis [28], barley [29], horsegram [30],
and rice [31]. The screening of maize varieties in the present study revealed root elongation
induced by P deficiency. Varieties in the low-P tolerant groups were found to elongate their
roots specifically under P deficient conditions, which could result in roots exploring more soil
volume (Table 1). The result was consistent with previous reports of maize root system
growth and development as influenced by P deficiency [32]. P uptake is also dependent on
root surface area and volume in contact with soils [33]. Our results suggested that low-P
tolerant genotypes could accelerate the growth of root surface and volume under low-P
limitation.

Effects of P deficiency on root biomass are more controversial. In a P starvation

experiment, increased root dry weight was found in 12 day-old maize plants [33], and it was
also reported that tolerant varieties of broad bean usually have higher root biomass [34, 35].
In contrast, no effect of P deprivation was observed on maize root biomass [36]. However,
authors generally agree that P deficiency in maize leads to a higher root: shoot ratio [37]. The
previous reports were supported by our results: that the root: shoot ratio of low-P tolerant
genotypes was higher under low-P limitation than that of low-P sensitive genotypes, in both
experiments in 2008 and 2009. This response indicated that dry matter accumulated more in
the roots of low-P tolerant genotypes than that of low-P sensitive genotypes. We suggest that
the relationship between root length and the total biomass for those tolerant and sensitive
varieties may be used as a fast screening protocol at the seedling stage in the field evaluation.

An understanding of the relationship with morphological and physiological

characteristics is fundamental for identification and utilization of P efficient germplasm [3]. In
this study, changes of root length induced by P deficiency were reflected in the biomass
correspondingly. Close relationships of the total biomass with root length were revealed. The

73
finding is expected since longer roots could be more efficient in absorbing P from soils low in
available P, which affected biomass accumulation and distribution. This suggested that root
length was a reliable trait in screening the germplasm for P efficiency.

Zhang et al. [1] reported that in the seedling stage, P uptake efficiency was the main
contributor to P tolerance, because a higher P uptake ability under P limitation may also
contribute to the tolerance to P limitation for a germplasm, regardless of its nutrient
requirement. In our study, there existed significant genotypic differences in P uptake
efficiency among different maize inbreds (Fig. 7). Low-P tolerant genotypes could keep the
higher uptake efficiency and absorb more P from soil to satisfy the demand of their growth
under low-P conditions, while low-P sensitive genotypes could not.

The elucidation of the transformation dynamics of P fractions is important for a better

understanding of the availability of different P fractions. In different soil types, the
availability of a certain P fraction to a plant differs significantly. Jiang and Gu [38] reported
that the Ca2-P fraction was the main one for rapid availability of P to plants, whereas Ca8-P,
Fe-P and Al-P were slowly available, and Ca10-P and O-P were unavailable. Previous studies
have shown that the primary P form was Ca2-P after P fertilizer was applied to calcareous soil
[39, 40]. Our results are consistent with this conclusion. Additionally, our data using the 32P
tracer technique indicated that the 32P activity reached a maximum before 3 d and then
decreased rapidly during 3 to 7 d after 32P-labeled fertilizer P was applied to Langfang low-P
soil. Simultaneously, the 32P in other P fractions, including Ca8-P, Fe-P, Al-P, O-P and Ca10-P
increased at different rates, i.e., Ca2-P transformed to other P fractions over time. Therefore,
we proposed that Ca2-P in this soil was a rapidly transforming phase and the main P fraction
available to the plant. A continuously increasing trend of 32P activity was observed in Ca8-P,
Fe-P and Al-P during the entire 25 d test period, and this rate was moderate. Therefore, Ca8-P,
Fe-P and Al-P should be described as moderate transforming phases. Among these three P
fractions, the 32P activity and its increasing rate in Ca2-P and Ca8-P exhibited maxima, which
might be partly due to the higher concentration of Ca than Fe and Al in calcareous soil.
Therefore, Ca8-P might be another important available P fraction in Langfang low-P soil.
During the test period of 25 d, the 32P activity in Ca10-P and O-P increased very slowly, which
suggested that Ca10-P and O-P were slow transforming phases (Fig. 9). Additionally, after
fertilizer P application, the 32P activity and its increasing rate in O-P decreased rapidly, which
indicated that a larger proportion of water-soluble P in P deficient soil without external P
fertilizer would transform to the plant-unavailable O-P fraction, i.e., a plant would have more
difficultly acquiring P from P deficient soil. The result presented further evidence that
external P fertilizer was essential to vigorous crops growing in the Langfang low P soil.

In low-P soil, the P efficiency of plants differs among genotypes within a given plant
species [11, 41, 42]. The screening criterion for low-P tolerant materials differs among plant
species [43–45]. This experiment was performed using radioactive 32P to estimate the plant
available soil P. The L-value has considerable theoretical advantages as a measurement of
plant available P from the soil. The L-values were determined after maize plants were grown
in Langfang low-P soil for 25 d to evaluate the genotype P efficiency. Together with our data,
the P efficiency for five maize genotypes was significantly different. DSY-32, which is a
typically low-P tolerant genotype, actively regulated the P utilization ratio between the soil P
and the exogenous fertilizer P. When exogenous P was supplied, DSY-32 preferentially
absorbed exogenous P, otherwise it would try to exploit soil P when no exogenous P was
supplied. However, another low-P tolerant genotype, DSY-2, exhibited a different low-P
tolerant pathway that efficiently utilized soil P regardless of exogenous P application. This
result suggested that different maize genotypes have very different low-P tolerant pathways

74
(Tables 2 and 3). Therefore, it was viable to fully exploit the limited P resources in low-P soil
by the screening and planting of low-P tolerant crop species. Additionally, the results showed
that the L-value was a very useful parameter for evaluation of plant P efficiency.

5. CONCLUSIONS

This field study clearly demonstrated that maize genotypes differ in their ability to
take up P from the low-P soil, and that these differences were attributed to the morphology
and physiology of the plants relative to their germplasm base. Based on these results, an
effective method of increasing P efficiency was to develop P tolerant cultivars that could
achieve a high yield under P deficiency. The results indicated that soil P availability during
maize seedling development was critical for early growth and grain yield of maize. The
inhibition of the root: shoot ratio, root length, root surface area, root volume and P uptake
efficiency were tentatively defined as screening indices for low-P tolerant genotypes during
the seedling stage. The relationship between root length and total biomass for those typical
genotypes could be used as a very fast screening protocol at the seedling stage in the field
evaluation under low-P limitation. In addition, the 100-grain weight was defined as the
screening indexes of low-P tolerant genotypes during the mature stage.

The 32P tracer technique provides a powerful alternative to better understand soil P
availability and sources of P pools in a low P soil-plant system. The results indicated that the
rate of water-soluble P that transformed to slowly available and unavailable P in the soil with
deficient P was higher than in the soil with sufficient P. Ca2-P was a quick transforming
phase; Ca8-P might be another important available P fraction while Ca10-P and O-P were slow
transforming phases. The L-value determination showed low-P tolerant cultivars regulated
soil P use efficiency and plant P content.

The study exploited the physiological-biochemical mechanism on P uptake and P

transport of selected maize genotypes in low-P soil, through a field experiment associated
with the use of 32P tracer. Since P deficiency had slight effects on low-P tolerant genotypes as
compared with low-P sensitive genotypes, it was demonstrated that differences of tolerance to
P deficiency existed among different maize genotypes.

REFERENCES

[1] ZHANG, L.M., et al., Phosphorus nutrient characteristics of different maize (Zea mays
L.) inbreds for tolerance to low-P stress, Agric. Sci. Chin. 4 (2005) 281–287.
[2] JENNIFER, L., et al., Phosphatase under-producer mutants have altered phosphorus
relations. Plant Physiol. 135 (2004) 334–345.
[3] YAO, Q.L., et al., The Effects of low phosphorus stress on morphological and
physiological characteristics of maize (Zea mays L.) landraces, Agric. Sci. Chin. 6
(2007) 559–566.
[4] LIU, J.Z., et al., Utilization of plant potentialities to enhance the bio-efficiency of
phosphorus in soil, Ecoagric. Res. 2 (1994) 16–23.
[5] LI, S.X., The current state and prospect of plant nutrition and fertilizer science, Plant
Nutr. Fert. Sci. 5 (1999) 193–205.
[6] WANG, Q.R., et al., Screening Chinese wheat germplasm for phosphorus efficiency in
calcareous soils, J. Plant Nutr. 28 (2005) 489–505.
[7] YAN, L., et al., Rhizosphere effect and root growth of two maize (Zea mays L.)
genotypes with contrasting P efficiency at low P availability, Plant Sci. 167 (2004)
217–223.

75
[8] USUDA, H., KOUSUKE, S., Phosphorus deficiency in maize. V. Leaf phosphate
status, growth, photosynthesis and carbon partitioning, Plant Cell Physiol. 32 (1991)
497–504.
[9] GUO, Y.C., et al., Screening methodology for rice (Oryza sativa) genotypes with high
phosphorus use efficiency at their seedling stage, Chin. J. Appl. Ecol. 13 (2002) 158–
159.
[10] ZHANG, J., et al., Genotypic differences in phosphorus uptake by maize inbred lines,
Guizhou Agric. Sci. 32 (2004) 22–23.
[11] HIRADATE, S., et al., Strategies of plants to adapt to mineral stress in problem soils,
Adv. Agron. 96 (2007) 66–133.
[12] NARUZZAMAN, M., et al., Distribution of carboxylates and acid phosphatase and
depletion of different phosphorus fractions in the rhizosphere of a cereal and three
grain legumes, Plant Soil 281 (2006) 109–120.
[13] MORRIS, R.A., GARRITY, D.P., Resource capture and utilization in intercropping:
non-nitrogen nutrients, Field Crops Res. 34 (1993) 319–334.
[14] BARRY, D.A.J., MILLER, M.H., Phosphorus nutritional requirement of maize
seedlings for maximum yield, Agron. J. 81 (1989) 95–99.
[15] HE, L.Y., LIANG, H.D., The improving method of sulfuric acid – perchloric acid
digesting plant sample and its studies for preventing nitrogen from volatile loss, Chin.
J. Anal. Chem. 20 (1992) 1277–1280.
[16] GU, Y.C., JIANG, B.F., Assay method of inorganic phosphorus fractionation in
calcareous soils, Soils 22 (1990) 101–102.
[17] ADU-GYAMFI, J.J., et al., Variations in phosphorus acquisition from sparingly
soluble forms by maize and soybean in low- and medium-P soils using 32P, Proc. Int.
Plant Nutr. Colloq. (2009) 1317 p.
[18] MUHAMMAD, S., et al., Genetic analysis for various quantitative traits in maize (Zea
mays L.) inbred lines, Int. J. Agric. Biol. 4 (2002) 379–382.
[19] TANG, J.H., et al. Genetic dissection for grain yield and its components using an
“immortalized F2 population” in maize, Acta Agron. Sinica 33 (2007) 1–7.
[20] KANAKA, S.K., Genetic analysis of ten quantitative characters in grain sorghum, Ind.
Thesis Abstr. 8 (1983) 408.
[21] HOU, Y.N., et al., P content and grain characters of different maize genotypes under
low-phosphorus soil environment, Acta Agric. Nucl. Sinica 23 (2009) 327–333.
[22] GOURLEY, et al., Defining phosphorus efficiency in plants, Plant Soil 155/156 (1993)
289–292.
[23] NI, J.J., et al., Rice seedling tolerance to phosphorus stress in solution culture and soil,
Nutr. Cycl. Agroecosyst. 51 (1998) 95–99.
[24] MILLER, C.R., et al., Genetic variation for adventitious rooting in response to low-
phosphorus availability: Potential utility for phosphorus acquisition from stratified
soils, Funct. Plant Biol. 30 (2003) 973–985.
[25] HUANG, P.M., “Soil chemistry”, Handbook of Soil Science, (Summer, M.E., Ed.),
CRC Press, Boca Raton (2000) B1-B24.
[26] ZHU, J.M., et al., Mapping of QTLs for lateral root branching and length in maize
(Zea mays L.) under differential phosphorus supply, Theor. Appl. Genet. 111 (2005)
688–695.
[27] ATKINSON, D., “Influence of root system morphology and development on the need
for fertilizers and the efficiency of use”, Plant roots: The Hidden Half, (WAISEL, Y.,
ESHEL, A., KAFKAKI, U., Eds), Marcel Dekker Inc., New York (1991) 411–451.
[28] MA Z., et al., Regulation of root elongation under phosphorus stress involves changes
in ethylene responsiveness, Plant Physiol. 131 (2003) 1381–1390.

76
[29] STEINGROBE, B., et al., Root production and root mortality of winter barley and its
implication with regard to phosphate acquisition, Plant Soil 237 (2001) 239–248.
[30] ANURADHA, M., NARAYANAN, A., Promotion of root elongation by phosphorus
deficiency, Plant Soil 136 (1991) 273–275.
[31] KIRK, G.J.D., DU, L.V., Changes in rice root architecture, porosity, and oxygen and
proton release under phosphorus deficiency, New Phytol. 135 (1997) 191–200.
[32] MOLLIER, A., PELLERIN, S., Maize root system growth and development as
influenced by phosphorus deficiency, J. Exp. Bot. 55 (1999) 487–489.
[33] ANGHINONI, I., BARBER, S.A., Phosphorus influx and growth characteristics of
corn roots as influenced by phosphorus supply, Agron. J. 72 (1980) 685–688.
[34] FAWOLE, I., et al., Genetic control of root development in beans (Phaseolus vulgaris
L.) grown under phosphorus stress, J. Am. Soc. Hort. Sci. 107 (1982) 98–100.
[35] GHADERI, A.H., LOWER, R.L., Gene effects of some vegetative characters of
cucumber, J. Am. Soc. Hort. Sci. 104 (1979) 104–144.
[36] KHAMIS, S., et al., CO2 assimilation and partitioning of carbon in maize deprived of
orthophosphate, J. Exp. Bot. 41 (1990) 1619–1625.
[37] ROSOLEM, C.A., et al., Root growth and mineral nutrition of corn hybrids as affected
by phosphorus and lime, Commun. Soil Sci. Plant Anal. 25 (1994) 2491–2499.
[38] JIANG, B.F., GU, Y.C., A suggested fractionation scheme of inorganic phosphorus in
calcareous soils, Sci. Agric. Sinica 20 (1989) 156–165.
[39] LI, Z.Y., Study on P-fixation intensity and P-fixation stroma of calcareous soils, Chin.
J. Soil Sci. 23 (1992) 190–192.
[40] LU, R.K., Research progress of soil phosphorus chemistry, Progr. Soil Sci. 18 (1990)
1–5.
[41] DAI, K.J., et al., Research advances in plant and its low-phosphorus environment -
inducement, adaptation and counter measures, Chin. J. Ecol. 25 (2006) 1580–1585.
[42] OZTURK, L., et al., Variation in phosphorus efficiency among 73 bread and durum
wheat genotypes grown in a phosphorus-deficient calcareous soil, Plant Soil 269 (2005)
69–80.
[43] YAN, X., et al., Phosphorus efficiency in common bean genotypes in contrasting soil
types. I. Vegetative response, Crop Sci. 35 (1995) 1086–1093.
[44] YAN, X., et al., Phosphorus efficiency in common bean genotypes in contrasting soil
types.Ⅱ. Vegetative response, Crop Sci. 35 (1995) 1094–1099.
[45] LI, Z.G., et al., On research progress of different crops genotypes phosphorus nutrition,
J. Inner Mongolia Univ. Nation. 17 (2002) 307–312.

77
 PHOSPHORUS USE EFFICIENCY BY BRAZILIAN UPLAND RICE
GENOTYPES EVALUATED BY THE 32P DILUTION TECHNIQUE

V.I. FRANZINI, F.L. MENDES

Brazilian Agricultural Research Corporation,
EMBRAPA-Amazonia Oriental,
Belém, PA,

T. MURAOKA, E.C. DA SILVA

Center for Nuclear Energy in Agriculture,
University of São Paulo,
Piracicaba, SP,

Brazil

J.J. ADU-GYAMFI
Soil and Water Management & Crop Nutrition Laboratory,
International Atomic Energy Agency,
Vienna,
Austria

Abstract

The objectives of this work were to identify the most efficient upland rice genotypes in
phosphorus (P) utilization, and to verify if P from the seed affects the classification of upland rice
genotypes on P uptake efficiency. The experiment was conducted in a greenhouse of the Center for
Nuclear Energy in Agriculture (CENA/USP), Piracicaba, São Paulo, Brazil, using the 32P isotope
technique, and plants were grown in pots with samples of dystrophic Typic Haplustox (Oxisol). The
experimental design was completely randomized with four replications. The treatments consisted of 47
upland rice genotypes and two standard plant species, efficient or inefficient in P uptake. The results
were assessed through correlation and cluster analysis (multivariate). The Carisma upland rice
genotype was the most efficient in P uptake, and Caripuna was the most efficient on P utilization. The
P derived from seed does not influence the identification of upland rice genotypes in P uptake
efficiency.

1. INTRODUCTION

In low input farming systems, phosphorus (P) is one of the most important factors
worldwide limiting crop yields. In Brazil, upland rice is typically grown in P deficient soils
with high P-fixing capacity [1] and without fertilization in agricultural frontier areas, mainly
in the Cerrado region. Low P input as fertilizer is one of the main factors that explain low
upland rice grain yields in Brazil, on average 2 t ha-1 [2], compared to lowland rice of 4.5 t ha-
1
[3]. Furthermore, upland rice is an important crop in Brazil, because it is grown by resource
poor farmers as a subsistence crop, representing a minimum input production system.

There is a large natural inter- and intra-specific genetic variation for plant traits that
are associated with P uptake efficiency, and development of transgenic plants can be used as a
strategy for improving P uptake efficiency of crops that may represent a sustainable solution
to increase yields [4]. Although success in developing nutrient efficient crop genotypes has
been limited, this strategy should continue to receive top priority during the 21st century [5].

79
 Upland rice genotypes differ in P use efficiency and efficient rice genotypes can be
used in breeding programs [1, 6]. Identifying rice genotypes more efficient in P uptake is the
first step to a successful breeding program, and is a strategy to reach high economic yields in
low input systems. High P content in rice grains (the majority as phytate) contributes little to
human nutrition because micronutrients such as iron and zinc are binding to phytate [7]. In
addition, continued removal of P from the fields in rice grain at harvest results in depletion of
soil P reserves in low input agricultural systems [8].

There are several definitions and calculation methods for P use efficiency (divided
into P uptake efficiency and P utilization efficiency). Here we define P uptake efficiency as
the ability of upland rice genotypes to take up P from soil assessed by the 32P dilution
technique [9], and P utilization efficiency as the ability to produce grain yield under low
available P supply [10]. The advantage of the 32P isotopic dilution method compared to others
is the possibility of eliminating the influence of seed derived P when comparing P uptake
efficiency by crop species or genotypes, by the L-value [9, 11, 12].

The ability of different plant species (canola, white lupin, pigeon pea, soybean,
sunflower and wheat) for absorbing less available forms of soil P was compared using the 32P
isotope dilution technique, and it was observed that the white lupin was more efficient [9].
Using this same technique in a study of 22 plant species [13], it was observed that white
lupin, upland rice, eucalyptus, cotton and pigeon pea were the most efficient in P uptake,
while sunhemp, cowpea and soybean were classified as less efficient species. An important
factor usually not considered in studies assessing genotypic variation for P uptake efficiency
is the P content present in seeds. Genotypes that have seeds with higher P content can be
classified, by mistake, as more efficient in P uptake than others with less P content of seeds.
Comparing wheat genotypes for their efficiency in P uptake, it was found that the tolerance to
P deficiency was higher in genotypes of durum wheat (Triticum durum L.) in relation to
wheat (Triticum aestivum L.) [14, 15]. In this study, the greater tolerance of durum wheat
genotypes was attributed to higher P content in seeds in relation to wheat.

The objectives of this study were to identify upland rice genotypes more efficient in P
uptake using the 32P isotope dilution technique and upland rice genotypes more efficient in P
utilization. Furthermore, we aim to verify if P from the seed affects the classification of
upland rice genotypes on P uptake efficiency.

2. MATERIALS AND METHODS

2.1. Experimental

The experiment was conducted in the greenhouse at Center for Nuclear Energy in
Agriculture (CENA/USP), located at latitude 22°42'30'' S, longitude 47°38'01'' W and 554 m
altitude, in Piracicaba, São Paulo, Brazil. The study were performed in 3.0 l plastic pots, lined
with polyethylene bags, containing 2.5 kg of air-dried soil, collected from the 0 to 0.20 m of a
dystrophic Typic Haplustox [16]. The soil samples were dried, sieved in a 2 mm mesh sieve
and homogenized. The soil had 280, 70 and 650 g kg-1 content of clay, silt and sand,
respectively, and the following chemical characteristics: pH (0.01 mol l-1 CaCl2, 4.5; organic
matter, 18.0 g dm-3 ; resin extracted P, 5 mg dm-3; K, 0.6 mmolc dm-3 ; Ca, 11.5 mmolc dm-3 ;
Mg, 5.2 mmolc dm-3; H + Al, 35.4 mmolc dm-3 ; CEC, 52.7 mmolc dm-3 ; sum of bases, 17.3
mmolc dm-3 ; base saturation, 32.8%, according to methodology described by [17]; and P by
Mehlich-1, 3 mg dm-3 [18].

80
 After application of lime (Calcium Carbonate Equivalent = 110%) to raise the base
saturation to 50% for the upland rice, according to the official recommendation of Bulletin
100 [19], the soil was incubated for 30 days and the moisture content was maintained at
approximately 70% of water holding capacity.

To evaluate the efficiency of genotypes of upland rice for P uptake, a mixture of triple
superphosphate (20 mg P kg-1 soil) as a source of readily available P to plants, and Patos
phosphate rock (150 mg P kg-1 soil) were applied to raise the total P content of soil to 170 mg
P kg-1 in each pot. N and K were applied at rates of 200 mg N kg -1 as urea and 200 mg K kg-1
as potassium sulphate. Fertilization with micronutrients, in the three experiments was done
applying nutrient solution in all treatments at rates of 0.5 mg B kg-1, 1.5 mg Cu kg-1, 3.0 mg
Fe kg-1, 2.0 mg Mn kg-1, 3.0 mg Zn kg-1 and 0.1 mg Mo kg-1.

The experimental design was completely randomized with four replications. The
treatments consisted of 47 upland rice genotypes and two standard species described in the
literature as efficient or inefficient in P uptake: Sunhemp (Crotalaria juncea L.) as inefficient
in absorbing P [12] and white lupin (Lupinus albus L.) as efficient [19, 12].

The upland rice genotypes evaluated were: Araguaia, Arroz Preto, Beira Campo, Bico
Ganga, BRS Aimoré, BRS Apinajé, BRS Aroma, BRS Bonança, BRS Caripuma, BRS
Colosso, BRS Curinga, BRS Monarca, BRS Pepita, BRS Primavera, BRS Sertaneja, BRS
Soberana, BRS Talento, BRSMG Conai, BRSMG Relâmpago, Cabaçu, Cambará, Canastra,
Carajás, Carisma, Cateto Seda, Centro América, Cuiabana, Douradão, Guanai, Guape,
Guaporé, IAC 25, IAC 47, IAC 60 dias, IAC 202, IAC 1246, Ipê, Jaguary, Maravilha,
Maranhão, Montaninha 90 dias, Progresso, Rio Paranaíba, Rio Verde, Tangará, Xingú and
Zebu.

The soil was labeled with 32P by applying a solution with 9.25 MBq of 32P and 0.2 mg
P kg-1 carrier. Eight seeds of the upland rice varieties or five seeds of the two standards were
sown in each pot, and the final population was thinned to three plants pot -1. Soil moisture was
maintained at approximately 70% of water retention capacity during the experiment.

The above-ground part of the plant of each genotype was taken at two samplings: (i)
first, two plants, from the total of three cultivated in each plot, were harvested at 40 days after
emergence, and (ii) the one remaining plant at the stage of panicle full maturity. The plant
samples were separated into shoots (leaves, stems, rachis and rice husks) and grain.

The seed-contained P was discounted for calculating the L-value, which was used to
compare the efficiency of P uptake among the genotypes, considering that from the total P
stored in the seeds of rice genotypes, 60% is used for plant growth [11], i.e., 40% of seed P is
not used by the plant (remains in the cotyledon).

2.2. Calculations of phosphorus and 32P

With shoot dry matter (Sdm), grain weight (Gw), P concentration in Sdm and Gw, the
P contents in the shoot and in grain were calculated:

P uptake = Pconcentration x Sdm , where Sdm is shoot dry matter.

P content = Pconcentration x Gw , where Gw is grain weight.

81
 With the data of plant P content and the 32P activity of the plant, the specific activity
(SA), the L-value, and the L-value subtracting the amount of seed-derived P from the total P
content of the shoot were calculated [9, 10].
32
P
SA = 31
P

where SA is specific activity (dpm µg-1 P); 32

P is radioisotope activity in the plant
(dpm); 31P is plant P content (µg P plant-1);

 SA 
Lvalue = X  0 − 1
 SA 
 p 

where L-value (mg P kg-1 soil); SA0 is specific activity of the applied solution (dpm
µg P); SAp is specific activity of plant (dpm µg-1 P); X is amount of applied P;
-1

 (X − Z ) 
L − s value =  Y T − X 
 YT 

where: L-s value is L-value subtracting P in the plant derived from the seed (mg of P
kg-1 soil); Y is the 32P activity in the applied solution (dpm); XT is plant P uptake (mg); YT is
32
P activity in the plant shoot dry matter (dpm); X is the rate of 31P carrier applied pot-1 (mg);
Z is the total P content derived from seed (mg).

2.3. Statistical analysis

The results of Sdm, Gw, P concentration and P content in the shoot or in the grain,
specific activity (SA), L-value and L-value subtracting the P derived from the seed (L-s value)
were submitted to analysis of Pearson linear correlation and hierarchical cluster analysis with
the objective for grouping the similar genotypes. Cluster analysis of upland rice genotypes
was carried out with the SAS 9.1 - “Statistical Analysis System” [20] and SYSTAT version
10.2 software programs, using the UPGMA (un-weighed pair group arithmetic average
clustering) The cluster analysis was preceded by the standardization of data before the
Euclidian distances calculation, as the variables presented different scales. After
standardization, all the variables were equally important in the determination of these
distances. Final results of the groups were presented as dendrograms. The P uptake efficiency
by plants is inversely proportional to SA and directly proportional to L-value and L-s value.

The upland rice genotypes were grouped into four or five groups, aiming at achieving
greater homogeneity within each group and greater heterogeneity among the different groups.
The results are presented and discussed in three parts: (1) first sampling - shoot; (2) second
sampling - shoot, and (3) second sampling - grain. The term shoot dry matter (Sdm) refers to
all above ground plant organs (leaves, stalks, husks and rachis) except the grain.

3. RESULTS AND DISCUSSION

3.1. First sampling - shoot

Plant data for the first sampling are given in Table 1.

82
TABLE 1. MEAN SHOOT DM YIELD (SDM) OF 47 UPLAND RICE GENOTYPES, P
CONCENTRATION (P CONC), P UPTAKE, SPECIFIC ACTIVITY (SA), L-VALUE AND L-
VALUE DISCOUNTING THE P FROM THE SEED (L-S VALUE) IN THE FIRST SAMPLING

Genotype Sdm P conc P uptake SA L-value L-s value

-1 -1 (mg pot- -1 -1
(g pot ) (g kg ) 1 (dpm µg P) (mg kg soil) (mg kg-1 soil)
)
Cuiabana 2.07 1.53 3.16 137.29 12.34 12.18
Caripuna 2.50 1.45 3.63 213.81 7.86 7.87
Relâmpago 2.63 1.44 3.77 200.52 8.39 8.38
Maravilha 2.66 1.58 4.20 103.04 16.81 16.75
Xingú 2.69 1.44 3.88 119.29 14.25 14.12
Ipê 2.75 1.60 4.39 152.76 11.08 10.98
Aroma 2.78 1.46 4.06 173.75 9.71 9.72
Canastra 2.80 1.34 3.76 136.66 12.40 12.32
Carisma 2.81 1.72 4.81 80.37 21.24 21.16
Carajás 2.91 1.49 4.32 158.15 10.69 10.66
IAC 202 2.95 1.48 4.38 140.00 12.10 12.09
Colosso 2.95 1.64 4.83 135.93 12.47 12.46
Progresso 2.95 1.52 4.49 105.28 16.17 16.09
Araguaia 2.99 1.55 4.63 129.26 13.14 13.12
Rio Verde 3.00 1.47 4.41 160.72 10.61 10.57
Arroz Preto 3.02 1.25 3.77 95.90 17.67 17.50
Bonança 3.05 1.46 4.44 152.64 11.11 11.10
Zebu 3.05 1.55 4.73 106.84 15.95 15.81
Guaporé 3.10 1.58 4.88 152.11 11.13 11.11
Talento 3.10 1.47 4.55 105.13 16.20 16.09
Pepita 3.10 1.72 5.33 130.04 13.05 12.95
Sertaneja 3.11 1.43 4.41 97.20 17.53 17.50
Primavera 3.14 1.53 4.78 126.48 13.43 13.36
Tangará 3.15 1.26 3.95 214.96 7.81 7.77
Apinajé 3.20 1.56 4.99 122.64 13.85 13.80
Montanhinha 90 dias 3.21 1.43 4.61 171.16 9.86 9.85
Guanai 3.23 1.34 4.32 153.87 10.99 10.99
Conai 3.24 1.60 5.18 138.10 12.27 12.27
Cambará 3.27 1.54 5.03 127.68 13.29 13.23
Monarca 3.28 1.25 4.11 164.62 10.27 10.25
Curinga 3.29 1.48 4.88 162.81 10.38 10.45
Douradão 3.39 1.49 5.02 154.87 10.96 10.92
IAC 47 3.42 1.37 4.68 132.41 12.81 12.73
Beira Campo 3.45 1.34 4.59 218.79 7.67 7.65
Aimoré 3.50 1.59 5.57 173.49 9.73 9.70
Soberana 3.51 1.51 5.28 169.53 9.96 9.95
Rio Paranaíba 3.52 1.27 4.48 102.31 16.71 16.58
Cateto Seda 3.54 1.30 4.60 186.54 9.03 9.01
Centro América 3.66 1.44 5.28 97.96 17.38 17.29
IAC 1246 3.73 1.28 4.78 158.50 10.67 10.57
Cabaçu 3.80 1.35 5.14 157.02 10.77 10.76
Maranhão 3.82 1.31 5.00 134.41 12.62 12.62
Bico Ganga 3.87 1.45 5.61 109.53 15.52 15.48
IAC 25 3.92 1.47 5.77 142.28 11.91 11.88
IAC 60 dias 3.94 1.26 4.94 145.46 11.76 11.75
Jaguary 4.04 1.33 5.39 195.96 8.61 8.60
Guape 4.36 1.29 5.63 187.21 8.96 8.91
Average 3.22 1.45 4.65 145.43 12.32 12.27
CV (%) 10.90 10.22 10.38 12.21 12.78 12.65

83
 The results obtained with the two standard species were: (i) White lupin - Sdm = 1.15
g pot , P in Sdm = 2.03 mg pot-1, SA = 39.06 dpm mg-1 P, L-value = 43.99 mg P kg-1 soil and
-1

L-s value = 2.6 mg P kg-1 soil, (ii) Sunhemp - Sdm = 5.16 g pot-1, P content in Sdm = 5.77 mg
pot-1, SA = 193.32 dpm mg -1 P, L-value = 6.88 mg P kg-1 soil and L-s value = 6.67 mg P kg-1
soil. The white lupin plant was, as expected, more efficient in absorbing P (the lowest SA, and
highest L-value L and L-s value) than all upland rice genotypes evaluated in this study. The
Sdm of rice is one of the main parameters related to grain yield of this crop, and P increases
due to an increase in the number of tillers and leaf area [21]. The values of Sdm of 47 upland
rice genotypes correlated significantly and negatively with Sdm P concentrations (–0.466***)
and positively with Sdm P contents (0.785***). Therefore, the dilution effect was observed in
Sdm P, i.e., increasing Sdm decreased the Sdm P concentrations, although the total P uptake
was higher. From these three variables, the cluster analysis identified the following five
groups of upland rice genotypes (Fig. 1):

1st: Aimoré, Soberana, Centro América, Bico Ganga, IAC 25, Jaguary and Guape;
2nd: Cabaçu, Maranhão, IAC 1246, IAC 60 dias, Rio Paranaíba, Cateto Seda, Beira Campo,
IAC 47 and Guanai;
3rd: Monarca, Tangará, Arroz Preto, Canastra, Aroma, Xingú, Relâmpago and Caripuna;
4th: Maravilha, Ipê, Progresso, Carajás, IAC 202, Rio Verde, Bonança, Talento, Sertaneja,
Montaninha 90 dias, Douradão, Curinga, Araguaia, Zebu, Primavera, Guaporé,
Apinajé, Cambará, Conai, Colosso, Carisma and Pepita;
5th: Cuiabana.
Aimoré
Soberana
C. América
Bico Ganga
IAC 25
Jaguary
Guape
Cabaçu
Maranhão
IAC 1246
IAC 60 dias
RioParanaíba
Cateto Seda
Beira Campo
IAC 47
Guanai
Monarca
Tangará
Arroz Preto
Canastra
Aroma
Xingú
Relâmpago
Caripuna
Maravilha
Ipê
Progresso
Carajás
IAC 202
Rio Verde
Bonança
Talento
Sertaneja
Montanhinha
Douradão
Curinga
Araguaia
Zebu
Primavera
Guaporé
Apinajé
Cambará
Conai
Colosso
Carisma
Pepita
Cuiabana

0 1 2 3
Euclidean distance

FIG. 1. Dendrogram resulting from the hierarchical cluster analysis of 47 genotypes of upland rice,
based on the variables of shoot dry matter (Sdm), concentration and accumulation of P in Sdm. First
plant sampling.

84
 Among all the correlations between variables of upland rice genotypes, taken in the
first sampling, the SA and L-value (–0.962 ***) x the L-s value (–0.960***), and L-value x
the L-s value (0.999***) were the variables that showed higher Pearson correlation
coefficients. By hierarchical cluster analysis with both variables SA and L-value (Fig. 2) as
with the SA and L-s value (Fig. 3), upland rice genotypes were classified for the P uptake
efficiency in the following four groups:

1st: very efficient, Carisma;

2nd: efficient, Arroz Preto, Sertaneja, Centro América, Maravilha, Rio Paranaíba, Talento,
Progresso, Zebu and Bico Ganga;
3rd: medium efficiency, Xingú, Apinajé, Primavera, Cambará, Araguaia, Pepita, IAC 47,
Maranhão, Colosso, Canastra, Cuiabana, Conai, IAC 202, IAC 25, IAC 60 dias,
Guaporé, Bonança, Ipê, Guanai, Douradão, Cabaçu, Carajás, IAC 1246, Rio Verde,
Curinga, Monarca, Soberana, Montaninha 90 dias, Aimoré and Aroma;
4th: less efficient, Cateto Seda, Guape, Jaguary, Relâmpago, Caripuna, Tangará and Beira
Campo.

The Carisma genotype was the best for P uptake efficiency, and did not form a group
with any other genotype (Figs. 2 and 3). Furthermore, we observed that the two dendrograms
(Figs. 2 and 3) are similar, meaning that there was no difference in P uptake efficiency among
groups of upland rice genotypes based on L-values or P in the plant derived from seed,
because the genotypes grouped by SA, L-value and L-s value were similar.

3.2. Second sampling - shoot

The Sdm values correlated significantly and positively with Sdm P concentrations
(0.486***) and P content in Sdm P (0.884***). Therefore, there was a response in shoot
production to an increase of P concentration in plant tissue. The dendrogram obtained by
grouping these three variables, in the second sampling, is shown in Fig. 4. The 47 upland rice
genotypes were classified into four groups:

1st: Cuiabana, Ipê, Cabaçu and Zebu (genotypes with higher values of Sdm, Sdm P
concentration and P uptake);
2nd: Cateto Seda, Beira Campo, Guaporé, Xingú, Sertaneja, Araguaia, Caripuna, Rio
Parnaíba, IAC 47, Maranhão, IAC 1246, Arroz Preto, Guape, Monarca, Canastra,
Maravilha, Rio Verde, Progresso, Curinga, Bonança, Pepita, Montaninha 90 dias,
Carisma, Jaguary, Carajás, Aroma, IAC 202, Talento, Cambará, IAC 25 and Soberana;
3rd: Tangará, Relâmpago, Aimoré, Conai, Douradão, Centro América, Colosso, Primavera,
Apinajé, Guanai and IAC 60 dias (genotypes with lower values of Sdm, P
concentration and content in Sdm);
4th: Bico Ganga. It did not group with any other genotypes, because although it had high
Sdm production, P concentration and P uptake were low (Fig. 4).

85
 Carisma
Arroz Preto
Sertaneja
C. América
Maravilha
RioParanaíba
Talento
Progresso
Zebu
Bico Ganga
Xingú
Apinajé
Primavera
Cambará
Araguaia
Pepita
IAC 47
Maranhão
Colosso
Canastra
Cuiabana
Conai
IAC 202
IAC 25
IAC 60 dias
Guaporé
Bonança
Ipê
Guanai
Douradão
Cabaçu
Carajás
IAC 1246
Rio Verde
Curinga
Monarca
Soberana
Montanhinha
Aimoré
Aroma
Cateto Seda
Guape
Jaguary
Relâmpago
Caripuna
Tangará
Beira Campo

0,0 0,5 1,0 1,5 2,0

Euclidean distance

FIG. 2. Dendrogram resulting from hierarchical cluster analysis of 47 genotypes of upland rice, based
on specific activity (SA) and L-value. First plant sampling.

3.3. Second sampling – grain

The grain dry matter, P concentration and P content of grain are given in Table 2. The
Gw values correlated significantly and negatively with its P concentrations (–0.512***) and
positively with its P contents (0.711***). The grain P concentration decreased with increasing
Gw due to the dilution effect of P in vegetal tissue. The positive correlation between grain
yield and its P content indicates that it is possible to increase grain production of upland rice
with increasing plant P content, as observed by [23] for common bean, suggesting the use of
bean genotypes more efficient in P utilization to increase grain yield.

86
 Carisma
Arroz Preto
Sertaneja
C. América
Maravilha
RioParanaíba
Talento
Progresso
Zebu
Bico Ganga
Xingú
Apinajé
Primavera
Cambará
Araguaia
Pepita
IAC 47
Maranhão
Colosso
Canastra
Conai
Cuiabana
IAC 202
IAC 25
IAC 60 dias
Guaporé
Bonança
Ipê
Guanai
Douradão
Cabaçu
Carajás
IAC 1246
Rio Verde
Curinga
Monarca
Soberana
Montanhinha
Aimoré
Aroma
Cateto Seda
Guape
Jaguary
Relâmpago
Caripuna
Tangará
Beira Campo

0,0 0,5 1,0 1,5 2,0

Euclidean distance

FIG. 3. Dendrogram resulting from hierarchical cluster analysis of 47 genotypes of upland rice, based
on specific activity (SA) and L-value discounting the P in plant derived from seed (L-s value). First
plant sampling.

From the analysis of hierarchical clustering of variables Gw concentration and content

of P, the following five groups were identified, homogeneous and distinct from varieties of
upland rice genotypes (Fig. 5) and were classified as:

1st: highly productive and highly rich in grain P content (genotype Caripuna);
2nd: very productive and very rich in grain P content (genotypes Bico Ganga, Sertaneja,
IAC 202, Colosso and Rio Parnaíba);
3rd: productive and rich in grain P (genotypes Cambará, Relâmpago, Tangará, Aroma,
Monarca, Guanai, Progresso, Cuiabana, Aimoré, Arroz Preto, Rio Verde, Xingú, Zebu,
Cabaçu, Araguaia, Douradão, Centro América, Maravilha, IAC 25, IAC 60 dias,
Primavera, Bonança, IAC 1246, Ipê, Montaninha 90 dias, IAC 47, Talento, Carisma,
Conai, Pepita, Beira Campo, Apinajé, Guaporé, Maranhão, Jaguary, Curinga, Canastra
and Cateto Seda);
4th: moderately productive and moderately rich in P in the grains (genotype Guape);
5th: less productive and low grain P (genotypes Soberana and Carajás).

87
 Cuiabana
Ipê
Cabaçu
Zebu
Cateto Seda
Beira Campo
Guaporé
Xingú
Sertaneja
Araguaia
Caripuna
RioParanaíba
IAC 47
Maranhão
IAC 1246
Arroz Preto
Guape
Monarca
Canastra
Maravilha
Rio Verde
Progresso
Curinga
Bonança
Pepita
Montanhinha
Carisma
Jaguary
Carajás
Aroma
IAC 202
Talento
Cambará
IAC 25
Soberana
Tangará
Relâmpago
Aimoré
Conai
Douradão
C. América
Colosso
Primavera
Apinajé
Guanai
IAC 60 dias
Bico Ganga

0,0 0,5 1,0 1,5 2,0

Euclidean distance

FIG. 4. Dendrogram resulting from hierarchical cluster analysis of 47 genotypes of upland rice, based
on shoot dry matter (Sdm), concentration and accumulation of P in Sdm. Second plant sampling.

The Guape genotype was not grouped with any other genotype due to its low yield,
but high accumulation of P in the grain, indicating that this genotype was not efficient in
converting the plant accumulated P. Caripuna showed Gw similar to other genotypes, but was
not grouped with any other, as the accumulation of P in the plant was higher than of other
upland rice genotypes. In this experiment, we observed higher Gw and its P content in Rio
Parnaíba compared to Araguaia, and these genotypes were classified as very productive and
productive, respectively. Differences in P uptake and grain yield among upland rice genotypes
grown in soil with low available P (P Mehlich-1 = 2.2 mg kg -1) were also observed in the field
[24].

88
TABLE 2. MEAN GRAIN DRY MATTER YIELD 47 UPLAND RICE GENOTYPES, P
CONCENTRATION AND P CONTENT IN GRAIN IN THE SECOND SAMPLING

Genotype Grain yield P concentration P content

(g plant-1) (g kg-1) (mg plant-1)
Soberana 10.66 2.01 21.35
Guape 11.21 2.45 27.47
Carajás 11.47 1.88 21.45
Zebu 12.43 2.24 27.77
Cabaçu 12.46 2.10 26.21
Xingú 12.64 2.22 28.06
Araguaia 12.83 2.04 26.12
Centro América 12.88 1.92 24.73
Douradão 12.88 2.03 26.12
Arroz Preto 13.25 2.29 30.34
IAC 60 dias 13.51 1.98 26.75
Rio Verde 13.61 2.22 30.18
Primavera 13.64 2.02 27.56
IAC 25 14.04 1.97 27.69
Montanhinha 90 dias 14.10 2.12 29.89
Progresso 14.15 1.82 25.68
Bonança 14.17 2.02 28.65
IAC 1246 14.25 2.05 29.21
Ipê 14.34 2.08 29.85
Maravilha 14.48 1.94 28.11
Cuiabana 14.74 1.84 27.02
Aimoré 14.81 1.87 27.59
Jaguary 14.81 2.15 31.83
Guanai 14.96 1.71 25.52
Monarca 15.18 1.72 26.00
Carisma 15.25 1.95 29.80
Conai 15.35 1.94 29.71
Relâmpago 15.41 1.76 27.00
Tangará 15.47 1.73 26.78
Aroma 15.49 1.70 26.38
Pepita 15.60 1.93 30.03
Curinga 15.72 2.27 35.60
Beira Campo 15.78 1.91 30.09
Maranhão 15.88 2.09 33.15
Apinajé 16.15 1.94 31.38
Guaporé 16.36 2.00 32.70
Cambará 16.61 1.67 27.68
Canastra 16.79 2.11 35.41
Cateto Seda 17.37 2.20 38.31
Talento 17.60 1.95 34.26
Rio Paranaíba 17.69 1.79 31.63
IAC 47 17.79 1.99 35.44
Caripuna 18.18 2.39 43.34
Colosso 18.52 1.65 30.50
Sertaneja 20.20 1.88 37.91
IAC 202 20.39 1.69 34.43
Bico Ganga 20.59 2.04 41.87
Average 15.14 1.98 29.88
CV (%) 12.32 12.19 12.15

89
 Guape
Soberana
Carajás
Cambará
Relâmpago
Tangará
Aroma
Monarca
Guanai
Progresso
Cuiabana
Aimoré
Arroz Preto
Rio Verde
Xingú
Zebu
Cabaçu
Araguaia
Douradão
C. América
Maravilha
IAC 25
IAC 60 dias
Primavera
Bonança
IAC 1246
Ipê
Montanhinha
IAC 47
Talento
Carisma
Conai
Pepita
Beira Campo
Apinajé
Guaporé
Maranhão
Jaguary
Curinga
Canastra
Cateto Seda
Bico Ganga
Sertaneja
IAC 202
Colosso
RioParanaíba
Caripuna

0 1 2 3
Euclidean distance

FIG. 5. Dendrogram resulting from hierarchical cluster analysis of 47 genotypes of upland rice, based
on grain dry matter yield (Gw), and accumulated P concentration in Gw. Second plant sampling.

Although Carisma was the most efficient upland rice genotype in P uptake (Figs. 2
and 3), it was not classified in the group of the genotypes most productive in grain. Therefore,
considering the definition of efficiency on P utilization by crops [10], Carisma was not the
most efficient in P utilization.

In the second group of upland genotypes Arroz Preto, Sertaneja, Centro América,
Maravilha, Rio Paranaíba, Talento, Progresso, Zebu and Bico Ganga were more efficient in P
uptake (Figs. 2 and 3). Bico Ganga, Sertaneja and Rio Parnaíba were the highlighted
genotypes, because these genotypes were classified in the second group that produced more
grain (Fig. 5).

There was no significant correlation between SA, L-value and L-s value of 47 upland
rice genotypes (measured in the first sampling) with Gw, P concentration and P content in the
grain (measured in the second sampling). This indicates that upland rice genotypes more
efficient in P uptake are not necessarily the most efficient in converting P taken up into grain.

90
 The P amount required by plants can be reduced by using efficient upland rice
genotypes in P use [25]. The identification of upland rice genotypes more efficient in P uptake
and P utilization is a strategy to reduce P fertilizer rates besides allowing its cultivation in
soils poor in P, and yet obtain high economic grain yields.

4. CONCLUSIONS

− The upland rice genotype Carisma was the most efficient in P uptake;

• The Caripuna upland rice genotype was the most productive in grain yield under
conditions of low available soil P (genotype more efficient in P utilization);

• The P derived from seed in the plant, when the 32P L-value technique is used, did not
affect the identification and classification of upland rice genotypes.

ACKNOWLEDGMENTS

The work was supported by IAEA (International Atomic Energy Agency) - research
contract 13779. Vinícius Ide Franzini acknowledges a graduate fellowship from CAPES
(Coordenação de Aperfeiçoamento de Pessoal de Nível Superior, Brazil).

REFERENCES

[1] FAGERIA, N.K., BALIGAR, V.C., Upland rice genotypes evaluation for phosphorus
use efficiency, J. Plant Nutr. 20 (1997) 499–509.
[2] EMBRAPA, Cultivo do Arroz de Terras Altas,
http://sistemasdeproducao.cnptia.embrapa.br/FontesHTML/Arroz/ArrozTerrasAltas/in
dex.htm (2011).
[3] CONAB, Arroz: Comparativo da área, produção e produtividade: safras 1976/77 a
2010/2011, http://www.conab.gov.br/conabweb/index.php?PAG=131 (2011).
[4] RAMAEKERS, L., et al., Strategies for improving phosphorus acquisition efficiency
of crop plants, Field Crops Res. 117 (2010) 169–176.
[5] FAGERIA, N.K., BALIGAR, V.C., LI, Y.C., The role of nutrient efficient plants in
improving crop yields in the twenty first century, J. Plant Nutr. 31 (2008) 1121–1157.
[6] FAGERIA, N.K., WRIGHT, R.J., BALIGAR, V.C., Rice cultivar evaluation for
phosphorus use efficiency, Plant Soil 111 (1988) 105–109.
[7] RABOY, V., Approaches and challenges to engineering seed phytate and total
phosphorus, Plant Sci. 177 (2009) 281–296.
[8] ROSE, T.J., et al., Genotypic variation in grain phosphorus concentration, and
opportunities to improve P-use efficiency in rice, Field Crops Res. 119 (2010) 154–
160.
[9] HOCKING, P.J., et al., “Comparation of the ability of different crop species to access
poorly-available soil phosphorus”, Plant Nutrition for Sustainable Food Production
and Environment, (ANDO, T., et al., Eds), Kluwer Academic Publishers (1997) 305–
308.
[10] GRAHAM, R.D., “Breeding for nutritional characteristics in cereals”, Advances in
Plant Nutrition, Vol. 1, (TINKER, P.B., LAUCHI, A., Eds), Praeger, New York (1984)
57–102.
[11] LARSEN, S., The use of 32P in studies of the uptake of phosphorus by plants, Plant
Soil 4 (1952) 1–10.
[12] BROOKES, P.C., Correction for seed-phosphorus effects in L-value determinations, J.
Sci. Food Agric. 33 (1982) 329–335.

91
[13] MURAOKA, T., et al., “Comparison of the ability of different plant species and corn
hybrids to access poorly-available soil phosphorus in an Oxisol of the Cerrado region,
Brazil”, Management Practices for Improving Sustainable Crop Production in Tropical
Acid Soils, Proceedings Series, IAEA, Vienna (2006) 137–146.
[14] OZTURK, L., et al., Variation in phosphorus efficiency among 73 bread and durum
wheat genotypes grown in a phosphorus-deficient calcareous soil, Plant Soil 269 (2005)
69–80.
[15] GUNES, A., et al., Genotypic variation in phosphorus efficiency between wheat
cultivars grown under greenhouse and field conditions, Soil Sci. Plant Nutr. 52 (2006)
470–478.
[16] DOS SANTOS, H.G., et al., Sistema Brasileiro de Classificação de Solos, 2nd Edn,
EMBRAPA Solos, Rio de Janeiro (2006) 306 p.
[17] VAN RAIJ, B., et al., Análise Química para Avaliação da Fertilidade de Solos
Tropicais, Instituto Agronômico, Campinas (2001) 285 p.
[18] EMBRAPA-SOLOS, Manual de Métodos de Análises de Solos, 2nd Edn, Rio de
Janeiro, (1997) 212 p.
[19] VAN RAIJ, B., et al., Recomendação de adubação e calagem para o Estado de São
Paulo, 2nd Edn, Instituto Agronômico e Fundação IAC, Campinas (Boletim Técnico,
100) (1997) 285 p.
[20] SAS INSTITUTE, SAS User’s Guide: Statistics, vs. 8.2., SAS Institute, Cary (2001).
[21] FAGERIA, N.K., Rice in Cerrado soils with water deficiency and its response to
phosphorus, Pesq. Agropec. Bras. 15 (1980) 259–265.
[22] CRUSCIOL, C.A.C., et al., Doses de fósforo e crescimento radicular de cultivares de
arroz de terras altas, Bragantia 64 (2005) 643–649.
[23] FAGERIA, N.K., et al., “Nutrição de fósforo na produção de feijoeiro”, Simpósio
Sobre Fósforo na Agricultura Brasileira, Associação Brasileira para Pesquisa da
Potassa e do Fosfato, Piracicaba 17 (2004) 435–455.
[24] FAGERIA, N.K., SANTANA, E.P., MORAIS, O.P. Resposta de genótipos de arroz de
sequeiro favorecido à fertilidade do solo, Pesq. Agropec. Bras. 30 (1995) 1155–1161.
[25] FAGERIA, N.K., Nutrient management for improving upland rice productivity and
sustainability, Commun. Soil Sci. Plant Anal. 32 (2001) 2603–2629.

92
 EVALUATION AND SELECTION OF COMMON BEAN (PHASEOLUS
VULGARIS L.) GENOTYPES FOR ROOT TRAITS ASSOCIATED WITH
PHOSPHORUS (P) ACQUISITION EFFICIENCY AND THE USE OF 32P
ISOTOPE IN STUDIES ON P UPTAKE BY ROOT HAIRS

M.A. MIGUEL, C. JOCHUA

Agricultural Research Institute of Mozambique (IIAM),
Maputo,
Mozambique

J.P. LYNCH
Pennsylvania State University,
University Park, PA,
USA

Abstract

Low phosphorus (P) availability is one of the main edaphic constraints limiting crop
production and productivity in most of the tropical agro-ecosystems. Several root traits are known to
be associated with P acquisition efficiency in low P soils. These root traits include root hairs.
Computer modeling, laboratory and field studies show the depletion of 32P-phosphate around roots and
that the depletion zone is influenced by the length and density of root hairs. We conducted a study
involving a series of experiments with the objective of evaluating the variability of root traits
associated with P uptake efficiency among common bean (Phaseolus vulgaris L.) genotypes, and to
understand the mechanisms of long root hairs leading to the increase in P uptake in common bean. The
study included (a) the screening of common bean genotypes in the laboratory and in the field for root
traits, and (b) the use of radioactive phosphorus (32P) in the experiments conducted in the greenhouse.
For laboratory screening, seedlings were germinated in paper rolls in a growth media for 3 days before
evaluation for basal root whorl number (BRWN), basal root number (BRN), basal root growth angle
(BRGA) and root hair length (RHL). Common bean genotypes were planted in the field with low P for
45 days after planting (DAP) before evaluation. For the 32P study four contrasting genotypes for root
hairs were grown for 28 DAP in the greenhouse using 15-20 liter pots filled with a mixture of sand
and vermiculate as the growth media. The radioactive P was incorporated in the growth medium in the
form of alumina-P fertilizer. Normal phosphorus (non-radioactive 31P) was included in the nutrient
solution in the form of calcium phosphate, Ca3(PO4)2, and supplied through irrigation. Screened
genotypes exhibited different root traits associated with P uptake efficiency, and that a given genotype
can have one or more root traits responsible for it P uptake efficiency. Data analysis of radioactivity
present in the plant tissue among contrasting genotypes showed that long root hair genotypes had
greater 32P uptake compared with short root hair genotypes. In addition, a strong positive correlation
(R2 = 0.8703) was observed between specific activity of 32P plant tissue and shoot dry weight of the
four genotypes contrasting for root hairs. These results suggests that (a) a genotype can exhibit one or
more root traits responsible for P uptake efficiency; (b) long root hairs increase total P uptake by
releasing organic compounds (root exudates) that could help to solubilize P otherwise not readily
available to inefficient genotypes with short root hairs.

1. INTRODUCTION

Common bean (Phaseolus vulgaris L.) is one of the most important sources of protein
for over a billion people worldwide, providing in addition to protein, vitamins and minerals
for people in the developing countries situated in the tropics [1, 2]. Meanwhile, bean
production is often limited by low soil fertility, especially low P availability, which is often
found in weathered tropical soils, because of the intense leaching and weathering conditions
and high P fixation by Fe and Al oxides [3]. Low P availability is especially problematic for

93
legumes like common bean, since root nodules responsible for N2 fixation have a high P
requirement. Application of P fertilizer is only a partial solution since they are costly,
nonrenewable, potentially harmful to the environment, and also marginally effective because
of immobilization by the soil. Therefore, several authors have suggested that genetic
improvement for P efficiency in crops would be more economical and practical than reliance
on chemical P fertilizers alone. As it has been widely demonstrated that genotypes vary in
response to high nutrient availability (thus the success of several efforts in that regard, e.g. the
20th century green revolution), we should accept that genotypes also vary in response to low
nutrient availability, which has been demonstrated by several authors. Lynch et al. [4] have
demonstrated genotypic variations in response to P availability in common bean and other
crops, and have referred to nutrient efficiency as the ability of a genotype to grow and yield at
sub-optimal nutrient supply. The root system is an important factor for plant productivity.
Plants evolved a wide range of adaptations to enhance P and water acquisition from the soil.
Important root traits for P acquisition efficiency are root hair length and density.

The study of root hairs is of a great importance in many disciplines of plant science. In
cell biology root hairs are important as single cell models [5] and in plant nutrition as nutrient
uptake organs [6]. The capacity of plants to absorb both water and mineral nutrients from the
soil is related to the plant’s ability to develop an extensive and well-located root system. Of
the total root surface area, root hairs can contribute up to 67%. This is a very valuable
contribution provided by root hairs. Cost / benefit analysis of C respired per unit of P acquired
suggests that the extension of the root surface area through root hairs is an efficient plant
strategy for improving P uptake at low costs in P stressed ecosystems [7]. In addition, it has
been demonstrated that the role of root hairs in P acquisition through studies in the laboratory
and computer modeling showing the depletion of 32P-phosphate around roots, and that the
depletion zone was influenced by the length and density of root hairs [8], involving low C
cost for maintenance.

The present study had two main objectives, namely, (a) to evaluate common bean
materials for root traits conferring P uptake efficiency both in laboratory and in the field; and
(b) to determine and quantify the contribution of root hairs for P acquisition in low P soils.
We hypothesize that (a) different genotypes employ different mechanisms (root traits) P
acquisition efficiency, and (b) the increased P content in plant tissue of efficient genotypes
grown in soils with low P is due to the ability of acquiring P made available through
dissolution of insoluble P by organic compounds from root exudates released in increased
amounts induced by long and dense root hairs.

2. MATERIALS AND METHODS

2.1. Genotype evaluation for root traits associated with P uptake efficiency

One hundred ninety six (196) genotypes selected from CIAT were evaluated in the
field experiments for root traits associated with P acquisition efficiency. G 19833, an Andean
genotype considered to be P efficient, with shallow basal roots, three basal root whorls, 12
basal roots and many adventitious roots, and DOR 364, a P inefficient genotype from the
Mesoamerican gene pool, with deep basal roots, two whorls and 8 basal roots, were included
as checks. The field experiment was conducted from June to August of 2010 in Rock Springs,
Pennsylvania, USA. The experimental design was a randomized complete block design
(RCBD) with 4 replications, and each experimental unit was composed of one row of 1.6 m
with eight plants. The spacing between rows was 0.7 m and between plants in a row was 0.2
m. The experiment was planted under low P availability. Weed and pest control, and irrigation
were applied as needed.

94
 Root crowns of 3 representative plants per replication were harvested 45 DAP. The
following traits were measured: (i) adventitious root length, branching and diameter; (ii) basal
root length, branching and diameter, (iii) primary root length, branching and diameter; (iv)
basal root angle; and (v) number of nodules. Actual counts were taken for total number of
adventitious and basal roots, and basal root whorls. Root length, angle from horizontal and
diameter were measured with a ruler, protractor and caliper, respectively. The root diameter
was measured in the main. Root branching (density) was obtained by counting the number of
lateral roots in a representative segment of 2 cm in each root class. Shoots were dried at 60o
for 2-3 days for determination of the shoot dry weight. Minitab statistical software (2010
Minitab Inc., USA) was used to analyze the data.

2.2. Root hair imagining

Basal roots were briefly stained with diluted Typan blue (0.05%) for better
visualization of root hairs. Root hair images were visualized with a light microscope and
images were captured at 40x magnification. Images were taken 2 cm above emerging new
root hairs. Image analysis software (http://rsbweb.nih.gov/ij/download.html) called Image J,
was used to measure root hair length and density. Root hair length of each genotype was
measured in 5 different representative segments per replication. The root hair density was
measured by counting the number of root hairs in a representative area. Root hair density was
then converted to number of root hairs per mm2. Genotypes were then grouped in three
categories based on root hair length: short (<0.4 mm), intermediate (0.4-0.6 mm) and long
root hairs (>0.6 mm).

2.3. Use of 32P isotope for determination of root hair contribution to P uptake

To test the above mentioned hypothesis, we employed a technique involving 32P in the
form of alumina-P in which common bean genotypes contrasting for root hair characteristics
were used in the study conducted in the greenhouse. The study consisted in using radioactive
phosphorus (32P) using 15-20 liter pots filled with a mixture of sand and vermiculate as the
growth medium. The radioactive P was incorporated in the growth medium in the form of
Alumina-P fertilizer, and normal 31P in the form of calcium phosphate, Ca3(PO4)2, was
included in the nutrient solution and supplied through irrigation. The solid phase buffered
alumina P fertilizer was prepared according to Lynch et al. [9], and regulated the availability
of 32P in the treatments. All other nutrients (including micronutrients) were applied in the
growth mediim as nutrient solution dissolved in the drip irrigation water.

Plants with contrasting root traits were grown for 35 DAP, and plant sub-samples were
collected for laboratory analysis, which consisted of the determination of radioactivity (CPM
g-1 plant tissue), using a liquid scintillation analyzer, and total P content in the plant tissue,
using a colorimetric absorption spectrophotometer (Lambda 25). Specific activity was
calculated using the following formula: SA = C (cps. g-1 DW) / (32P + 31P).

Four Recombinant Inbred Lines (RILs) contrasting in root hair length and density
were used in this study. Two lines were selected from DOR364 x G19833 RIL population and
the other two lines were selected from G2333 x G19839 RIL population of common bean.
The selection of DO364 x G19833 RILs was based on [10] and the selection of the G2333 x
G19839 was based on root hair evaluation at 10 DAP in the field. Long root hairs genotypes
used in this study were RILs DxG53 and GxG41, and short root hair genotypes were DxG11
and GxG23.

95
 The experiment was conducted under controlled greenhouse conditions, using a
Completely Randomized Design Block (CRBD), with a total of 4 genotypes contrasting in
root hair length, making a total of 2 treatments (2 genotypes per treatment), with 3
replications. Pots with the following dimensions: 30 cm deep and 25 cm in diameter were
used. Pots were filled with a mixture of 50% Vermiculite, 45% sand and 5% alumina P,
making a total volume of 20 liters per pot. At 35 DAP, plant shoots were collected from all
treatments, and shoot dry weight determined.

3. RESULTS

3.1. Genotype evaluation for root traits associated to P acquisition efficiency

Field measurements showed significant differences among genotypes for adventitious

root number, branching, and diameter, basal root whorl number, basal root number,
branching, diameter and growth angle, primary root branching and diameter, and number of
nodules (P≤ 0.01) (Fig. 1). Considerable variation in root traits was found among common
bean genotypes evaluated in 8-day old seedlings. Significant differences in 165 genotypes
were detected in basal root whorls number (BRWN), (F value = 8.2***), number of basal
roots (F = 7.7***) and root hair length (F = 6.3***) evaluated in 8-day old seedlings. BRWN
varied from 1 to 4. Most genotypes had 2 whorls. The average root hair length measured on
basal roots varied from 0.19 to 0.78 mm (Table 1), and there was a positive correlation
between root hair length and density.

High correlation between basal root number and basal root whorl number evaluated in
8-day old seedlings was found (R2 = 0.9, P≤ 0.01). Basal root number evaluated in 45-day old
plants (Table 2) was moderately correlated with BRWN measured in 8-day old seedlings (R2
= 0.522, P≤ 0.01). Similarly to data from 8-day old seedlings, a strong and significant
correlation was found between basal root number and BRWN (R2 = 0.88***) measured in the
field. In this study substantial variation occurred in root traits among common bean
genotypes. Useful root traits conferring tolerance to low P such as high number of basal root
whorls, basal root number, basal root growth angle (BRGA) can be found in most of common
bean accessions. Information on genotypic diversity of root traits and sources of useful root
traits is important in breeding programs for development of genotypes adapted to a specific
stress.

96
 ARN ARL ARB ARD
20
20 10 20

10 10
5 10

0 0 0 0
8 16 24 32 40 48 6 9 12 15 18 21 2 4 6 8 10 12 0.2 0.4 0.6 0.8 1.0 1.2

BRWN BRN BRL BRB

16
50 30
20
Frequency

25 15 8
10

0 0 0 0
1.2 1.6 2.0 2.4 2.8 4 5 6 7 8 9 10 11 16 18 20 22 24 26 28 30 5 6 7 8 9 10 11

BRD A ngle PRL PRB

20 20 20
10
10 10 10
5

0 0 0 0
1.2 1.5 1.8 2.1 2.4 2.7 20 30 40 50 60 70 10 12 14 16 18 20 22 5 6 7 8 9 10 11

PRD SDW (g) Nodule

20
20 20

10 10 10

0 0 0
1.5 2.0 2.5 3.0 3.5 4.0 4.5 80 100 120 140 160 0 15 30 45 60 75

FIG. 1. Phenotypic variation of root traits of the genotypes evaluated in the field. Adventitious root
number (ARN); Adventitious root length (ARL); Adventitious root branching (ARB); Adventitious root
diameter (ARD); Basal root whorl number (BRWN); Basal root number (BRN); Basal root length
(BRL); Basal root branching (BRB); Basal root diameter (BRD); Basal root growth angle (Angle);
Primary root length (PRL); Primary root branching (PRB); Primary root diameter (PRD); number of
nodules per plant, and shoot dry weight (SDW). Branching correspond to number of lateral roots in 2
cm root segment. ARN, BRWN and BRN are counts per plant.

3.2. Root hair contribution to P uptake

Genotypes GxG 23 had less shoot dry weight (1.28 g) followed by DxG 11 and
GxG41, both with 1.37 g each. Genotype DxG 53 had the greatest shoot dry weight (1.65 g).
Genotype DxG53, which is a long-haired line, had significantly higher shoot dry weight
compared to the rest of the genotypes (Fig. 2).

TABLE 1. ROOT HAIR LENGTH IN SELECTED GENOTYPES THAT CAN BE USED IN THE
BREEDING PROGRAMS. ROOT HAIR LENGTH (RHL) CLASSIFICATION: SHORT: < 0.4 MM;
LONG: > 0.6 MM

Genotype Gene BRWN BRN Basal Category PR Category

pool RHL RHL
(mm) (mm)
AFR 298 Andean 2 6.5 0.705 Long / high 0.73 Long / high
G 14665 Andean 3.5 13.25 0.65 Long / high 0.795 Long / high
Sel. 63
crema Andean 2.5 8 0.71 Long / high 0.695 Long / high
PVA 773 Andean 3 10 0.295 Short/ low 0.4 Short/ low
SUG 47 Andean 2.75 11.75 0.38 Short/ low 0.4 Short/ low
SEA 5 Mesoam. 2 7.25 0.69 Long / high 0.745 Long / high
VAX 1 Mesoam. 2 8 0.67 Long / high 0.735 Long / high
SXB 418 Mesoam. 2 7.75 0.42 Short/ low 0.33 Short/ low

97
TABLE 2. ROOT TRAITS OF SIX COMMON BEAN GENOTYPES EVALUATED IN THE
FIELD. THE DATA ARE AVERAGE OF 4 REPLICATIONS MEASURED 45 DAYS AFTER
PLANTING

Genotype Basal root Basal root Basal root whorl Adventitious root
growth angle number number number
BAT 477 67.5a 7.2c 4a 5.5b
Tio Canela 62a 7c 3b 6.25b
SEQ 1003 57a 8.25bc 3b 4.25b
Bonus 37b 10.5b 2.5c 8ab
LIC-04-3-1 27b 14a 2c 12.5a
LIC-04-2-1 26b 11b 2c 7.5ab
Means within a column followed by the same letter are not significantly different (P<0.05)

Total radioactivity (RA) of 32P was assessed both in the stems and in the leaves of the
4 genotypes. Efficient (long root haired) genotypes had significantly higher accumulation of
32
P in plant tissue (Fig. 3). 32P activity in the leaves ranged from 7681 to 10775 CPM g-1 DW,
while for the stems 32P activity ranged from 4123 to 6792 CPM g-1 DW.

Efficient genotypes had greater total P compared to inefficient genotypes (Fig. 4). In
addition, long root hair genotypes DxG53 and GxG 41 had greater accumulation of 32P in
plant tissue compared to short root hair genotypes GxG23 and DxG11. In fact efficient
genotypes (GxG 41 and DxG 53) accumulated 25.3 and 33.1% respectively, while inefficient
genotypes (DxG11 and GxG 23) had percentages of 32P accumulation of only 10.0% and
11.2%, respectively (Fig. 5). Efficient genotypes had lower specific activity values compared
to inefficient genotypes, since these were able to uptake P from two different pools with a
greater total P accumulation (Fig. 6).

1.80

1.60

1.40
Shoot dry weight, grams

1.20

1.00

0.80

0.60

0.40

0.20

0.00
GxG 23 DxG 11 GxG 41 DxG 53
Genotypes

FIG. 2. Shoot dry weight of 4 common bean genotypes contrasting for root hairs, grown for 30 days in
the greenhouse under low P availability and exposed to 32P. Genotypes GxG 41 and DxG 53 have long
root hairs, while genotypes GxG 23 and DxG 11 have short root hairs. Genotype DxG 53 (a RIL from
DOR364xG19833), showed significantly high shoot dry weight compared to the rest of the genotypes,
which showed no significant differences among them (Y axis represents shoot dry weight in grams).

98
FIG. 3. Total radioactivity among four contrasting common bean genotypes for root hairs: short-root
hair genotypes (open bars) and long root hair genotypes (solid bars).

Phosphorus Partioning among root categories of

common bean

1.20
T o ta l P c o n te n t, m g / p la n t

1.00

0.80

0.71 P31
0.60
P32

0.40
0.57

0.20
0.29
0.07
0.00
Short root hair Long root hair
Genotypes
!

FIG. 4. Total P content per plant tissue between two common bean categories: Long root hair and
short root hair genotypes. Long root hair category has greater portion of 32P absorbed in plant tissue
compared to short root hair genotypes.

99
 32 31
FIG. 5. Perceptual proportion of P (blue bars) and P (plum bars) in plant tissue of 4 genotypes
contrasting for root hair length.

FIG. 6. Specific radioactivity in plant tissue of 4 genotypes contrasting for root hairs: Genotypes with
short root hairs (open bars) and genotypes with long root hairs (full bars).

100
4. DISCUSSION

The results of genotype screening show that genotypes exhibit a number of root traits
associated with P acquisition efficiency. A given genotype can have one or more root traits
responsible for P acquisition efficiency. These traits include root hair length and density, basal
root growth angle, adventitious rooting and basal root number.

The importance of root hairs for P acquisition in low P environments is very well
documented [2]. Plant roots not only are responsible for absorbing nutrients, but they are also
responsible for secretion of a variety of compounds (exudates) into the rhizosphere, leading to
an increase of the amounts and forms of dissolved ions in the soil solution. Root hairs enable
these processes to occur more effectively. A study involving nearly 20 genotypes contrasting
for root hairs grown in the field with low P availability showed a strong correlation of root
hair length to shoot dry weight and to shoot P content in plant tissue [10]. In fact, it has been
estimated that of the total root surface area, root hairs can contribute up to 67%, a very
significant contribution.

The results of this study attempt to give an explanation about one of the physiological
mechanisms associated with root hairs’ contribution to enhanced P uptake in plants grown in
soils with low P. Less soluble Al-P applied in all of the four treatments could only be readily
available to the plants after being dissolved from its source. We observed that genotypes with
long root hairs were able to dissolve and take up more 32P than genotypes with short root
hairs. This supports our hypothesis that long root hair genotypes are able take up more P, in
part by being able to dissolve P initially not readily available. The plant might achieve this by
releasing some forms of organic compounds (root exudates) capable of dissolving P otherwise
not readily available to the plant, and root hairs might be associated with the increased level
of root exudates occurring in genotypes with long root hairs. The P availability in seeds from
genotypes with long root hairs and genotypes with short hairs requires assessment.

We also observed that 32P was found in grater amounts in the leaves than in the stems
in all the treatments. This was probably due to the fact that P, which is relatively immobile in
the soil, is readily mobile in plant tissue, and therefore, it could have been rapidly translocated
to the growing younger leaves.

The availability of an additional source of P for long root hair genotypes (observed by
an increase in the amount of 32P in plant tissue, which can also be seen by looking at the
proportion of 32P in relation to total P (32P + 31P) in plant tissue (Fig. 5), was expressed by
additional growth and vigor, and ultimately a relative increase in shoot dry matter
accumulation in long root hair genotypes. In addition, a strong correlation between absorbed
32
P and shoot dry weight observed in this study seems to support the idea that an additional
source of P available to hairy genotypes led to an increased accumulation of shoot dry matter,
which can in turn lead to a better plant performance.

5. CONCLUSIONS

According to the results of our studies we can conclude that (a) genotypes have
various root traits associated with P uptake efficiency, and that a given genotype can have one
or more root traits that will help it to acquire P, when grown in low P soils (b) root hairs
confer a significant contribution to P acquisition, confirming that genotypes with long root
hairs have significantly better performance (greater total shoot dry weight and P content and
concentration) compared to genotypes with short root hairs (c) the method that was used in

101
this study involving 32P can also be used to quantify the contribution (degree of importance)
of a particular root class (e.g. adventitious, basal, tap root, etc.) for overall P uptake by plants.

REFERENCES

[1] FAO (FOOD AND AGRICULTURE ORGANIZATION), Production Yearbook, Vol.

44, FAO, Rome, Italy (1990).
[2] BROUGHTON, W.J., et al., Beans (Phaseolus spp.) – model food legumes, Plant Soil
252 (2003) 55–128.
[3] SAMPLE, E.C., SOPER, R.J., RACZ, G.J., “Reactions of phosphate fertilizers in
soils”, The Role of Phosphorus in Agriculture, (KHASAWNEH, F.E., SAMPLE, E.C.,
KAMPRATH, E.J., Eds), American Society of Agronomy, Madison, USA (1980)
263–310.
[4] LYNCH, J.P., The role of nutrient efficient crops in modern agriculture, J. Crop Prod.
1 (1998) 241–264.
[5] RIDGE, W.R. “Root hairs: Cell Biology and development”, Plant Roots: The Hidden
Half, (WAISEL, Y., ESHEL, A., KAFKAFI, U., Eds), Marcel Dekker, Inc., New York
(1996) 127–147.
[6] GAHOONIA, T.S., et al., A root hairless barley mutant for elucidating genetic of root
hairs and phosphorus uptake, Plant Soil 235 (2001) 211–219.
[7] NIELSEN, K.L., et al., Fractal geometry of root systems: Field observations of
contrasting genotypes of common bean (Phaseolus vulgaris L.) grown under different
phosphorus regimes, Plant Soil 206 (1998) 181–190.
[8] BHAT, K.S., NYE, P.H., Diffusion of phosphate to plant roots in soil. I. Quantitative
autoradiography of the depletion zone, Plant Soil 38 (1973) 161–175.
[9] LYNCH, J.P., DEIKMAN, J., Phosphorus in Plant Biology: Regulatory Roles in
Molecular, Cellular, Organismic, and Ecosystem Processes, American Society of Plant
Physiologists (1999).
[10] MIGUEL, M.A., Genotypic variation of root hairs and phosphorus efficiency in
common bean (Phaseolus vulgaris L.), MS Thesis, Pennsylvania State University
(2004).

102
 PHOSPHORUS USE EFFICIENCY BY BRAZILIAN COMMON BEAN
GENOTYPES ASSESSED BY THE 32P DILUTION TECHNIQUE

V.I. FRANZINI
Brazilian Agricultural Research Corporation,
EMBRAPA-Amazonia Oriental,
Belém, PA,

T. MURAOKA
Center for Nuclear Energy in Agriculture,
University of São Paulo,
Piracicaba, SP,

Brazil

J.J ADU-GYAMFI
International Atomic Energy Agency,
Vienna,
Austria

J.P. LYNCH
Pennsylvania State University,
University Park, PA,
USA

Abstract

The objectives of this work were to identify the most efficient common bean (Phaseolus
vulgaris L.) genotypes on phosphorus (P) utilization, and verify if P from the seed affects the
classification of common bean genotypes on P uptake efficiency when the 32P isotopic dilution
technique is used. The experiment was conducted in a greenhouse, and plants were grown in pots with
surface samples of a dystrophic Typic Haplustox. The treatments consisted of 50 common bean
genotypes and two standard plant species, efficient or inefficient in P uptake. The results were
assessed through correlation and cluster analysis (multivariate). Sangue de Boi, Rosinha, Thayú,
Grafite, Horizonte, Pioneiro and Jalo Precoce common bean genotypes were the most efficient on P
uptake, and Carioca 80, CNF 10, Pérola, IAPAR 31, Roxão EEP, Aporé, Pioneiro, Pontal, Timbó and
Rudá were the most efficient in P utilization. The P derived from seed influences the identification of
common bean genotypes for P uptake efficiency.

1. INTRODUCTION

Common bean is one of the main crops grown in the off-season under irrigation in the
Cerrado (Savannah) areas of Brazil [1]. The low P content and the high P fixation capacity in
the Cerrado soils is one of the main limitations to agricultural productivity. To reach
satisfactory production of beans it is necessary to apply high rates of P fertilizer [2]. Besides
liming and fertilizer application to minimize such problems, another strategy would be to
identify and to explore the use of the genotypic differences in common bean for efficiency in
P use (uptake and utilization), which can reduce the expenses with P fertilizers [3, 4]. The P
recovery efficiency by the bean plant is less than 10% in the Cerrado soils, depending on the
application rate [5]. This low value of P efficiency by beans can be increased with the use of
genotypes more efficient in P uptake. There are common bean genotypes which can increase
P uptake significantly through the capacity to modify the rhizosphere by organic acid

103
exudation, surface root architecture and longer basal root hairs under stressed conditions [6].
In other words some common bean cultivars can increase soil P use efficiency by different
adaptations [7, 8].

The objectives of this study were to identify common bean genotypes more efficient in
P uptake using the 32P isotope dilution technique, and also for P utilization. Furthermore, we
investigated whether P from the seed affects the classification of common bean genotypes in
P uptake efficiency when the 32P isotopic dilution technique is used.

2. MATERIALS AND METHODS

2.1. Experimental

The experiment was conducted in the greenhouse at the Center for Nuclear Energy in
Agriculture (CENA / USP), located at latitude 22°42'30'' S, longitude 47°38'01'' W and 554 m
altitude, in Piracicaba, Sao Paulo, Brazil.

The study were performed in 3.0 l plastic pots, containing 2.5 kg of air-dried soil,
collected from the 0 to 0.20 m layer of a dystrophic Typic Haplustox [9]. The soil had 280, 70
and 650 g kg-1content of clay, silt and sand, respectively, and the following chemical
characteristics: pH (0.01 mol l-1 CaCl2), 4.5; organic matter, 18.0 g dm-3 ; P extracted by resin,
5 mg dm-3 ; K, 0.6 mmolc dm-3 ; Ca, 11.5 mmolc dm-3; Mg, 5.2 mmolc dm-3; H + Al, 35.4
mmolc dm-3; CEC, 52.7 mmolc dm-3; sum of bases, 17.3 mmolc dm-3; base saturation, 32.8%,
according to methodology described by [10] and P by Mehlich-1, 3 mg dm-3 [11].

After application of lime (Calcium carbonate equivalent = 110%) to raise the base
saturation to 70%, according to the official recommendation of Bulletin 100 [12], the soil was
incubated for 30 days, maintaining the moisture content at approximately 70% of water
holding capacity. To evaluate the efficiency of bean genotypes for P uptake, a mixture of
triple superphosphate (20 mg P kg-1) as a source of readily available P, and Patos rock
phosphate (150 mg P kg-1) were applied to raise the total soil P content to 170 mg P kg-1
soil in each pot. N and K were applied at rates of 200 mg N kg-1 as urea and 200 mg K kg-1 as
potassium sulfate. Micronutrients were also applied as nutrient solution in all treatments at
rates of 0.5 mg B kg-1, 1.5 mg Cu kg-1, 3.0 mg Fe kg-1, 2.0 mg Mn kg-1, 3.0 mg Zn kg-1 and
0.1 mg Mo kg-1.

The experimental design was completely randomized with four replications. The
treatments consisted of 50 common bean genotypes and two standard species described in the
literature as efficient or inefficient in P uptake: Sunhemp (Crotalaria juncea L.) as inefficient
in absorbing P [13] and white lupin (Lupinus albus L.) as efficient [13, 14]. The common
bean genotypes evaluated in this study were: Carioca 80, Rudá, Aporé, Princesa, Pérola,
Requinte, Pontal, BRS Horizonte, BRSMG Talismã, BRSMG Pioneiro, IPR Colibri, IAC
Tybatã, IAC Alvorada, LP01-38, CV-48, CNF 10, Roxão EEP, Sangue de Boi, Roxão,
Roxinho, Safira, BRS Timbó, Roxo 90, BRS Pitanga, Gen 99TG50-47, Ouro Negro, Ônix,
Macanudo, Chapecó, Xodó, Diamante Negro, Xamego, IAC UNA, BRS Valente, BRS
Grafite, FT Nobre, BRS Triunfo, Thayú, Rosado, Rosinha de Cipó, Rosinha G2, Rosinha,
Rosinha Brilhosa, Rubi, BRS Vereda, IAC Boreal, Gen99TGR1-10, Jalo Precoce, IAPAR 31
and Gen99TG34-50.

The soil was labeled with 32P solution (9.25 MBq of 32P) and 0.2 mg P kg-1 carrier.
Five bean seeds were sown in each pot and the plants were thinned to three plants pot -1. Soil

104
moisture was maintained at approximately 70% of water retention capacity during plant
development.

The above-ground parts of the plants were sampled on two occasions: (i) two plants
from the total of three grown in each plot were harvested at 30 days after emergence, and (ii)
the one remaining plant, at the stage of grain physiological maturity. The shoot samples were
separated into stem, branches, leaves, bark and grain. The P contained in seed was discounted
for calculating the L-value, which was used to compare the efficiency of P uptake among the
genotypes, considering that from the total P stored in the seeds of bean genotypes, 60% is
used for plant growth [15], i.e. 40% of seed P is not used by the plant and remains in the
cotyledon.

2.2. Calculations of P and 32P

With shoot dry matter (Sdm), grain weight (Gw), P concentration in Sdm and Gw the
P content in the shoot and in grain were calculated.

P uptake = P concentration × Sdm , where Sdm is shoot dry matter.

P content = P concentration × Gw, where Gw is grain weight.

With the data of plant P concentration and 32P activity of the plant, the specific
activity, L-value, and the L-value subtracting the amount of seed-derived P from the total P
content of shoots were calculated [14,16].
32
P
SA = 31
P
where SA is specific activity (dpm µg-1P); 32
P is radioisotope activity in the plant
(dpm); 31P is plant P content (µg of P plant-1);
 SA 
Lvalue = X  0 − 1
 SA 
 p 

where L-value (mg P kg-1 soil); SA0 is specific activity of applied solution (dpm µg-1
P); SAp is specific activity of the plant (dpm µg-1P); X is amount of applied P;
 (X − Z ) 
L − s value =  Y T − X 
 YT 

where L-s value is L-value subtracting P in the plant derived from the seed (mg of P
kg soil); Y is 32P activity in the applied solution (dpm); XT is plant P uptake (mg); YT is 32P
-1

activity in the plant shoot dry matter (dpm); X is 31P carrier applied rate pot-1 (mg); Z = total P
content derived from the seed (mg).

2.3. Statistical analysis

Sdm, Gw, P concentration and P content in the shoot or in the grain, specific activity
(SA), L-value and L-value subtracting the P derived from the seed (L-s value) were submitted
to analysis of Pearson linear correlation, and hierarchical cluster analysis with the objective
for grouping similar genotypes. Cluster analysis of bean genotypes was carried out with the
SAS 9.1 - “Statistical Analysis System” [17] and SYSTAT version 10.2 software programs,
using the UPGMA (un-weighed pair group arithmetic average clustering) The cluster analysis

105
was preceded by the standardization of data before the Euclidian distances calculation, as the
variables presented different scales. After standardization, all the variables were equally
important in the determination of these distances. Final results of the groups were presented as
dendrograms. The P uptake efficiency by plants is inversely proportional to SA and directly
proportional to L- and L-s values.

The common bean genotypes were grouped into four or five groups, aiming to achieve
greater homogeneity within each group and greater heterogeneity among the different groups.
The results are presented and discussed in three parts: (1) first sampling - shoot; (2) second
sampling - shoot, and (3) second sampling - grain. The term shoot dry matter (Sdm) refers to
all above ground tissues (stem, branches, leaves and bark of legumes) except the grain.

3. RESULTS AND DISCUSSION

3.1. First sampling - shoot

Plant data for the first sampling are given in Table 1. The results obtained with the two
standard species were: (i) White lupin - Sdm = 0.56 g pot -1, P in Sdm = 1.34 mg pot-1, SA =
17.71 dpm g-1 P, L-value = 47.44 mg P kg-1 soil and L-s value = 29.5 mg P kg-1 soil, (ii)
Sunhemp - Sdm = 3.49 g pot-1, P content in Sdm = 5.40 mg pot-1, SA = 75.49 dpm mg 1 P-, L-
value = 10.26 mg P kg-1 soil and L-s value = 9.95 mg P kg-1 soil. The white lupin plant was,
as expected, more efficient in absorbing P (the lowest SA, and highest L-value and L-s
values) than all common bean genotypes evaluated in this study.

Mean values of shoot dry matter (Sdm), P concentration (P conc) and P uptake (P
uptake) in Sdm, specific activity (SA), L-value and L-value discounting the P from the seed
(L-s value) of 50 common bean genotypes at the first sampling are given in Table 1.. The
values of Sdm correlated significantly and negatively with P concentration (–0.625***) and
positively with P uptake (0.675***). P uptake in Sdm increased with increasing Sdm, but the
P concentration in plant tissue decreased by the dilution effect of P in the Sdm. By cluster
analysis with these three variables (Fig. 1), the following five groups of common bean
genotypes were identified:

1st: Rubi;
2nd: Grafite, CV-48, Pioneiro, Xamego, IAC UNA, Carioca 80, Roxão, Horizonte, Thayú,
Triunfo, Sangue de Boi, Gen99TG3450, Timbó, Talismã, Pontal, Pérola, Requinte,
Gen99TG50-47, Vereda, IAC Alvorada and Rosinha Brilhosa;
3rd: Pitanga, Roxo 90, FT Nobre, Ônix, LP01-38, Rudá, Colibri, Aporé, Tybatã,
Macanudo, Rosinha Cipó, Princesa, Xodó, Valente, Diamante Negro, CNF 10,
Roxinho, IAPAR 31, Rosado and Chapecó;
4th: Gen99TGR1-10, Ouro Negro, Rosinha, Safira, Rosinha G2, Jalo Precoce and Roxão
EEP; and
5th: IAC Boreal.

106
TABLE 1. MEAN VALUES OF SHOOT DRY MATTER (SDM), P CONCENTRATION (P CONC)
AND P UPTAKE IN SDM, SPECIFIC ACTIVITY (SA), L-VALUE AND L-VALUE
DISCOUNTING THE P FROM THE SEED (L-S VALUE) OF 50 COMMON BEAN GENOTYPES
AT THE FIRST SAMPLING

Genotype Sdm P conc P uptake SA L-value L-s value

(g pot-1) (g kg-1) (mg pot-1) (dpm µg-1 P) (mg kg-1 soil) (mg kg-1 soil)
Chapecó 5.33 1.42 7.57 56.06 11.81 11.00
IAPAR 31 5.56 1.51 8.38 65.40 10.10 9.38
Rosado 5.62 1.45 8.11 63.48 10.42 9.28
Roxinho 5.77 1.51 8.69 51.04 13.00 12.23
CNF 10 5.78 1.51 8.71 58.81 11.25 10.25
Rubi 5.80 1.71 9.92 49.47 13.43 12.43
Diamante Negro 5.82 1.52 8.82 65.87 10.03 9.29
Pitanga 6.00 1.33 7.97 51.16 12.98 11.83
Grafite 6.06 1.59 9.60 44.86 14.82 13.44
Valente 6.07 1.46 8.86 60.08 11.02 10.05
CV-48 6.20 1.56 9.65 53.00 12.51 11.05
Xodó 6.32 1.43 9.04 58.42 11.33 10.69
Princesa 6.43 1.43 9.20 57.95 11.43 10.61
Colibri 6.47 1.32 8.50 66.46 9.94 9.18
Rosinha de Cipó 6.47 1.43 9.26 53.42 12.41 11.58
FT Nobre 6.47 1.27 8.24 47.68 13.95 12.50
Aporé 6.49 1.34 8.65 57.00 11.62 10.28
Roxo 90 6.53 1.23 8.03 76.04 8.66 7.81
Xamego 6.53 1.52 9.93 56.38 11.75 11.08
IAC UNA 6.56 1.51 9.93 52.85 12.54 11.34
Horizonte 6.58 1.55 10.18 44.28 15.02 13.80
Macanudo 6.59 1.36 8.95 55.39 11.96 11.15
Pioneiro 6.60 1.49 9.82 43.33 15.37 14.33
Tybatã 6.60 1.34 8.82 50.49 13.15 11.77
Carioca 80 6.63 1.51 10.01 53.68 12.34 11.36
Vereda 6.66 1.43 9.51 51.21 12.99 11.96
Thayú 6.67 1.57 10.44 46.21 14.40 13.54
Roxão 6.68 1.51 10.10 52.91 12.53 10.85
Ônix 6.68 1.27 8.47 49.53 13.41 12.20
Rudá 6.79 1.30 8.83 52.53 12.62 11.75
Pérola 6.80 1.44 9.81 53.40 12.41 11.37
LP01-38 6.85 1.23 8.46 54.93 12.08 10.81
Gen99TG50-47 6.88 1.41 9.70 48.87 13.58 11.08
Requinte 6.91 1.42 9.78 59.88 11.05 10.31
Talismã 6.97 1.45 10.12 57.35 11.54 10.71
Rosinha Brilhosa 7.01 1.33 9.33 54.86 12.07 11.23
Pontal 7.02 1.42 9.96 55.58 11.92 10.99
IAC Alvorada 7.05 1.35 9.49 60.09 11.01 9.86
Timbó 7.20 1.39 10.03 49.39 13.44 12.63
Gen99TG34-50 7.22 1.38 9.97 46.67 14.23 12.51
Sangue de Boi 7.40 1.37 10.13 47.30 14.05 13.06
Triunfo 7.42 1.38 10.25 51.75 12.82 11.61
Safira 7.86 1.33 10.45 52.61 12.60 11.86
Ouro Negro 7.98 1.19 9.45 58.96 11.23 10.31
Rosinha G2 7.99 1.31 10.48 57.90 11.44 10.72
Jalo Precoce 8.01 1.31 10.51 39.72 16.77 15.14
Gen99TGR1-10 8.14 1.15 9.34 51.52 12.88 11.39
Mean 6.78 1.40 9.44 53.80 12.51 11.43
CV (%) 12.46 9.10 9.43 12.58 12.42 12.42

107
 Rubi
Grafite
CV-48
Pioneiro
Xamego
IAC UNA
Carioca 80
Roxão
Horizonte
Thayú
Triunfo
Sangue Boi
GEN99TG3450
Timbó
Talismã
Pontal
Pérola
Requinte
GEN99TG50-47
Vereda
IAC Alvorada
Ros Brilhosa
Pitanga
Roxo 90
FT Nobre
Ônix
LP01-38
Rudá
Colibri
Aporé
Tybatã
Macanudo
Rosinha Cipó
Princesa
Xodó
Valente
Diamante Neg
CNF 10
Roxinho
IAPAR 31
Rosado
Chapecó
GEN99TGR1-10
Ouro Negro
Rosinha
Safira
Rosinha G2
Jalo Precoce
Roxão EEP
IAC Boreal

0.0 0.5 1.0 1.5 2.0

Euclidean distance

FIG. 1. Dendrogram resulting from hierarchical cluster analysis of 50 common bean genotypes, based
on shoot dry matter (Sdm), P concentration and P uptake at the first sampling.

Of the five groups of common bean genotypes formed by hierarchical cluster analysis
(Fig. 1) and considering the values of Sdm and P uptake shown in Table 1, it was found that
the genotype IAC Boreal presented higher values of Sdm and P uptake, while the genotypes
in contrast, Pitanga, Roxo 90, FT Nobre, Ônix, LP01-38, Rudá, Colibri, Aporé, Tybatã,
Macanudo, Rosinha Cipó, Princesa, Xodó, Valente, Diamante Negro, CNF 10, Roxinho,
IAPAR 31, Rosado and Chapecó presented lower values of Sdm and P uptake. Jalo Precoce
was classified in the second group that presented higher values of Sdm and P uptake. Among
eight common bean genotypes grown under a low P rate (24 mg dm-3 P2O5) in the substrate
(pots with 16 kg of sand) and harvested at 45 days after germination, genotypes BAT 477,
Jalo Precoce and Roxo produced the most Sdm [4].

Rubi did not group with the other genotypes, because it had high P concentration and
P uptake, but low production of Sdm. This probably indicates that this genotype was not
efficient in utilizing P taken up to produce the Sdm, at the beginning of plant development.
We emphasize that genotypes more productive in terms of Sdm, are not necessarily more
efficient in P uptake under conditions of low P availability in the substrate; the P in the plant

108
derived from the seed, from which the plant originates, is an important source of P at the early
developmental stage, in evaluating the efficiency of P uptake by plants. Moreover, the
production of Sdm also involves the concept of efficiency of use, and genotypes more
efficient in P use are those that best convert the nutrient uptake into Sdm. It was observed that
the variables SA, L-value and L-s value of 50 common bean genotypes, harvested in the first
sampling, were the ones with the highest Pearson correlation coefficients between them. The
SA correlated significantly and negatively with the L-value (–0.982***) and the L-s value (–
0.960***), and the L-value correlated positively with the L-s value (0.974***).

By cluster analysis with the variables SA and L-value (Fig. 2) four groups of common
bean genotypes were identified:

1st: low efficiency in P uptake (Roxo 90);

2nd: moderately efficient in P uptake (Colibri, Diamante Negro, IAPAR 31, Rosado, IAC
Alvorada, Valente, Requinte, Ouro Negro, CNF 10, Xodó, Princesa, Rosinha G2,
Talismã, Aporé, Xamego, Chapecó, Pontal, Macanudo, LP01-38 and Rosinha
Brilhosa);
3rd: efficient in P uptake (Carioca 80, Rosinha Cipó, Pérola, CV-48, Roxão, IAC UNA,
Safira, Rudá, Triunfo, Gen99TGR1-10, Vereda, Pitanga, Roxinho, Tybatã, Ônix, Rubi,
Timbó, Roxão EEP, GenTG50-47, IAC Boreal, FT Nobre, Sangue de Boi, Rosinha,
Gen99TG3450 and Thayú);
4th: very efficient in P uptake (Grafite, Horizonte, Pioneiro and Jalo Precoce).

Grafite, Horizonte, Pioneiro and Jalo Precoce genotypes were those with lower values
for SA and higher L-values. Thus, these common bean genotypes were classified as more
efficient in P uptake. It is noteworthy that in this classification of genotypes for the P uptake
efficiency, P was not discounted in the plant from the seeds. In another study evaluating the
efficiency of P uptake by eight genotypes of common bean, without the use of the technique
with 32P, the Jalo Precoce was also classified as efficient in P uptake [4].

By cluster analysis with the variables SA and L-s value (Fig. 3), four groups of
common bean genotypes were identified:

1st: low efficiency in P uptake (Roxo 90);

2nd: moderately efficient (Colibri, Diamante Negro, IAPAR 31, Rosado, IAC Alvorada,
Valente, Requinte, Ouro Negro, CNF 10, Aporé, Talismã, Princesa, Rosinha G2 and
Xodó);
3rd: efficient in P uptake (Rosinha Brilhosa, Macanudo, Pontal, Chapecó, Xamego, LP01-
38, Roxão, CV-48, IAC UNA, Pérola, Carioca 80, Rosinha Cipó, Gen99TGR1-10,
Triunfo, Rudá, Safira, Vereda, Pitanga, Tybatã, Gen99TG50-47, IAC Boreal,
Roxinho, Ônix, Rubi, Roxão EEP, Timbó, FT Nobre and Gen99TG3450);
4th: very efficient in P uptake (Sangue de Boi, Rosinha, Thayú, Grafite, Horizonte,
Pioneiro and Jalo Precoce).

109
 Roxo 90
Colibri
Diamante Neg
IAPAR 31
Rosado
IAC Alvorada
Valente
Requinte
Ouro Negro
CNF 10
Xodó
Princesa
Rosinha G2
Talismã
Aporé
Xamego
Chapecó
Pontal
Macanudo
LP01-38
Ros Brilhosa
Carioca 80
Rosinha Cipó
Pérola
CV-48
Roxão
IAC UNA
Safira
Rudá
Triunfo
GEN99TGR1-10
Vereda
Pitanga
Roxinho
Tybatã
Ônix
Rubi
Timbó
Roxão EEP
GEN99TG50-47
IAC Boreal
FT Nobre
Sangue Boi
Rosinha
GEN99TG3450
Thayú
Grafite
Horizonte
Pioneiro
Jalo Precoce

0 1 2 3 4
Euclidean distance

FIG. 2. Dendrogram resulting from hierarchical cluster analysis of 50 common bean genotypes based
on specific activity (SA) and L-value at the first sampling.

Xamego, Chapecó, Pontal, Macanudo, LP01-38, Rosinha Brilhosa, Sangue de Boi,

Rosinha and Thayú genotypes were classified into different groups, when the hierarchical
cluster analysis was performed based on L-value discounting (Fig. 3) or not (Fig. 2), the P in
the plant derived from the seed. Therefore, the P present in the seed affected the assessment
and classification of common bean genotypes on P uptake efficiency.

Xamego, Chapecó, Pontal, Macanudo, LP01-38 and Rosinha Brilhosa genotypes were
placed in the third group, when the P in the plant derived from the seed was discounted in the
calculation of the L-value, and were classified as efficient in P uptake. Sangue de Boi,
Rosinha and Thayú genotypes were grouped in the fourth group, when discounting the seed P,
and were classified as very efficient in P uptake.

When plants are grown under P limiting condition, the roots become a strong drain of
carbohydrates and this causes major limitation to the growth of the shoot than the root [18].
Roots of bean plants grown under conditions of P deficiency had much higher concentrations
of sugars in the roots compared with plants with an adequate P supply, due to the increased
shoot translocation of photo-assimilates [19]. In another study it was observed that the

110
difference in the selection of bean genotypes and the P-use efficiency and dry matter
production was related to the translocation of P from roots to shoots [4]. The root architecture
is another factor that differentiates the P uptake among genotypes, and relates to the spatial
configuration of the root system, i.e. the geometry of the development of the root axes [20].

Roxo 90
Co lib ri
Diamante Neg
IAPAR 31
Rosado
IAC Alvorada
Valen te
Requinte
Ouro Negro
CNF 10
Ap oré
Talismã
Prin cesa
Rosinha G2
Xodó
Ros Brilhosa
Macanudo
Pontal
Chapecó
Xamego
LP01-38
Roxão
CV-48
IAC UNA
Pérola
Carioca 80
Rosinha Cipó
GEN99TGR1-10
Triu nfo
Rudá
Safira
Vereda
Pitanga
Tybatã
GEN9 9TG50-47
IAC Boreal
Roxinho
Ônix
Rubi
Roxão EEP
Timbó
FT Nobre
GEN99TG3450
Sangue Boi
Rosinha
Thay ú
Grafite
Horizonte
Pioneiro
Jalo Precoce

0 1 2 3 4
Euclidean distance

FIG. 3. Dendrogram resulting from hierarchical cluster analysis of 50 common bean genotypes, based
on specific activity (SA) and L-value, discounting the P in the plant derived from the seed at the first
sampling.

3.2. Second sampling - shoot

Mean values of shoot dry matter (Sdm), P concentration (P conc) and P uptake (P
uptake) in Sdm of 50 common bean genotypes AT the second sampling are given in Table 2.
The Sdm of common bean genotypes in the second sampling, correlated significantly and
positively with P uptake (0.587***), but not with P concentration.

111
TABLE 2. MEAN VALUES OF SHOOT DRY MATTER (SDM), P CONCENTRATION (P CONC)
AND P UPTAKE (P UPTAKE) IN SDM OF 50 COMMON BEAN GENOTYPES AT THE
SECOND SAMPLING

Genotype Sdm P conc P uptake

(g plant-1) (g kg-1) (mg plant-1)
CNF 10 6,12 0,68 4,19
Macanudo 6,25 0,56 3,48
Ouro Negro 6,35 0,76 4,80
Rosinha G2 6,38 0,84 5,34
Gen99TG50-47 6,41 1,00 6,38
Aporé 6,47 0,62 3,98
Carioca 80 6,71 0,62 4,14
Sangue de Boi 6,90 0,94 6,52
Safira 6,96 0,72 4,99
CV-48 6,99 0,76 5,30
Colibri 7,03 0,98 6,86
Gen99TG34-50 7,32 0,79 5,77
LP01-38 7,37 0,76 5,57
Xamego 7,37 0,80 5,87
Princesa 7,43 0,72 5,36
IAPAR 31 7,45 0,78 5,78
Pontal 7,52 0,83 6,23
Jalo Precoce 7,53 0,87 6,58
Horizonte 7,55 0,89 6,71
IAC Alvorada 7,59 0,64 4,86
Rubi 7,67 0,71 5,43
Xodó 7,69 0,74 5,70
Pioneiro 7,82 0,60 4,71
Valente 7,85 0,74 5,77
Triunfo 7,88 0,71 5,57
Roxão EEP 7,96 0,59 4,67
Talismã 7,99 0,79 6,32
Pérola 8,03 0,63 5,03
Pitanga 8,11 0,58 4,65
Rosinha 8,13 0,86 6,95
IAC UNA 8,20 0,65 5,31
Rosinha de Cipó 8,22 0,94 7,69
Chapecó 8,25 0,91 7,52
Diamante Negro 8,36 0,78 6,48
Rosinha Brilhosa 8,37 0,87 7,30
FT Nobre 8,39 0,67 5,62
Roxinho 8,43 0,73 6,18
Vereda 8,48 0,64 5,39
Rudá 8,63 0,82 7,08
Timbó 8,63 0,54 4,68
Rosado 8,64 0,57 4,94
Gen99TGR1-10 8,79 0,82 7,18
Roxo 90 8,89 0,75 6,64
Thayú 8,89 0,58 5,14
Requinte 8,90 0,85 7,55
Ônix 9,05 0,80 7,28
Roxão 9,26 1,15 10,68
Mean 7,91 0,76 5,99
CV (%) 13,10 14,26 15,12

112
 The dendogram obtained by grouping these two correlated variables is presented in
Fig. 4.

Roxão
IAC Boreal
Tybatã
Grafite
Roxo 90
Rudá
GEN99TGR1-10
Ônix
Requinte
Rosinha Cipó
Chapecó
Ros Brilhosa
Rosinha
Talismã
Diamante Neg
Roxinho
Timbó
Rosado
Thayú
FT Nobre
Vereda
IAC UNA
Pérola
Pitanga
Roxão EEP
Pioneiro
IAC Alvorada
Triunfo
Valente
Xodó
Rubi
Princesa
LP01-38
GEN99TG3450
Xamego
IAPAR 31
Pontal
Jalo Precoce
Horizonte
Colibri
Sangue Boi
GEN99TG50-47
Safira
CV-48
Rosinha G2
Ouro Negro
Carioca 80
Aporé
CNF 10
Macanudo

0 1 2 3
Euclidean distance

FIG. 4. Dendrogram resulting from hierarchical cluster analysis of 50 common bean genotypes, based
on shoot dry matter (Sdm) and P uptake in Sdm at the second sampling.

The four genotype groups formed were:

1st: Roxão (genotype with highest values of dry matter yield and P accumulation in
MSPA);
2nd: IAC Boreal, Tybatã and Grafite;
3rd: Roxo 90, Rudã, Gen99TGR1-10, Ônix, Requinte, Rosinha Cipó, Chapecó, Rosinha
Brilhosa, Rosinha, Talismã, Diamante Negro, Roxinho, Timbó, Rosado, Thayú, FT
Nobre, Vereda, IAC UNA, Pérola, Pitanga, Roxão EEP, Pioneiro, IAC Alvorada,
Triunfo, Valente, Xodó, Rubi, Princesa, LP01-38, Gen99TG3450, Xamego, IAPAR
31, Pontal, Jalo Precoce, Horizonte, Colibri, Sangue de Boi and Gen99TG50-47;
4th: Safira, CV-48, Rosinha G2, Ouro Negro, Carioca 80, Aporé, CNF 10 and Macanudo
(genotypes with lower Sdm and P uptake).

113
3.3. Second sampling - grain

Mean values of grain yield (Gw), P concentration (P conc) and P uptake (P uptake) in
Gw of 50 common bean genotypes at the second sampling are presented in Table 3. The grain
yield was significantly correlated with bean Sdm (Y = –2610.23 + 5.58 X – 0.0013 X2, R2 =
0.60 **), where Y = grain yield (kg ha-1) and X = Sdm (kg ha-1), and the maximum
productivity of approximately 3200 kg ha-1 of grain was obtained with the production of 2098
kg ha-1 of Sdm [5] . It was observed in this study that a quadratic model was used to explain
the relationship between grain yield and Sdm, which are two quantitative variables and
dependent. It is recommended to apply Pearson linear correlation analysis to these types of
variables.

Considering the 50 common bean genotypes evaluated in the second sampling, a

linear and positive correlation was observed between Sdm and Gdm, but with a low
correlation coefficient (0.292***). This indicates that the relationship between dry matter
yield and grain yield depends on the bean genotype, i.e. this effect does not seem to be
general with all common bean genotypes. For example, Roxão showed higher values of dry
matter yield and P uptake, but was ranked as moderately productive and moderately rich in
grain P (Fig. 5). Thus, in comparative studies of production between genotypes, even if
conducted in a greenhouse, the plants should be developed to the production of grain. There
was a significant and positive correlation between the P content in grain and bean grain
productivity [5]. Therefore, according to these authors, it is possible to increase the
productivity of common bean by increasing the absorption and accumulation of P in grains,
with the use of efficient genotypes.

The Gdm values correlated significantly and negatively with P concentrations in the
Gdm (–0.654***) and positively with P content in the Sdm (0.604***). The P content in the
Gdm was higher with increasing dry matter yield, but decreased P concentrations, so probably
there was a dilution on P content in Gdm. By cluster analysis with these three variables the
following five groups of common bean genotypes were identified (Fig. 5):

1st: little productive and high P in grain (Talismã);

2nd: little productive and low P in grain (Ouro Negro and Horizonte);
3rd: moderately productive and moderately rich in grain P (IAC Boreal, Roxão, Jalo
Precoce, Gen99TG50-47, Gen99TGR1-10, Rubi, Gen99TG3450, Rosinha G2,
Princesa, Safira, Rosinha Brilhosa, Macanudo, IAC Alvorada, Pitanga, CV-48,
Diamante Negro, Rosinha Cipó, LP01-38, Grafite, Colibri, Chapecó and Xamego);
4th: productive and rich in P in grain (Valente, Rosado, Ônix, Roxo 90, Tybatã, Roxinho,
IAC UNA, Thayú, Sangue de Boi, Rosinha, Vereda, FT Nobre, Xodó, Requinte and
Triunfo);
5th: very productive and very rich in grain P (Carioca 80, CNF 10, Pérola, IAPAR 31,
Roxão EEP, Aporé, Pioneiro, Pontal, Timbó and Rudá).

114
TABLE 3. MEAN VALUES OF GRAIN YIELD (GW), P CONCENTRATION (P CONC) AND P
UPTAKE (P UPTAKE) IN GW OF 50 COMMON BEAN GENOTYPES AT THE SECOND
SAMPLING

Genotype Gw P conc P uptake

(g plant-1) (g kg-1) (mg plant-1)
IAC Boreal 6.48 3.62 23.45
Talismã 6.74 4.51 30.30
Ouro Negro 6.86 3.95 26.99
Roxão 6.87 3.29 22.57
Horizonte 7.15 3.78 27.02
Jalo Precoce 7.26 3.24 23.54
Rosinha G2 7.30 3.44 25.09
Princesa 7.34 3.49 25.57
Gen99TGR1-10 7.46 3.30 24.65
Gen99TG50-47 7.49 3.20 23.93
Rubi 7.64 3.34 25.48
Gen99TG34-50 7.81 3.22 25.04
Safira 7.85 3.41 26.77
Xamego 8.04 3.06 24.56
IAPAR 31 8.11 3.83 31.09
Rosinha Brilhosa 8.18 3.28 26.84
CV-48 8.19 3.44 28.21
Chapecó 8.26 3.16 26.06
IAC Alvorada 8.39 3.31 27.75
CNF 10 8.46 3.62 30.66
Macanudo 8.46 3.27 27.60
Pitanga 8.47 3.32 28.09
Carioca 80 8.58 3.49 29.94
Pérola 8.64 3.62 31.21
Roxão EEP 8.69 3.82 33.10
Triunfo 8.72 3.33 29.05
Rosinha de Cipó 8.73 3.15 27.44
Grafite 8.74 2.92 25.55
Diamante Negro 8.76 3.18 27.82
LP01-38 8.84 3.11 27.44
Colibri 8.91 2.98 26.49
Requinte 9.14 3.23 29.50
Aporé 9.16 3.58 32.78
Xodó 9.30 3.19 29.62
Thayú 9.59 3.04 29.12
Roxinho 9.63 2.87 27.63
FT Nobre 9.64 3.17 30.50
Rudá 9.65 3.32 32.01
Pioneiro 9.67 3.41 32.94
Pontal 9.68 3.39 32.80
Rosinha 9.69 3.09 29.97
IAC UNA 9.70 2.85 27.57
Timbó 9.71 3.31 32.14
Sangue de Boi 9.76 3.03 29.56
Tybatã 9.77 2.92 28.47
Vereda 9.85 3.10 30.54
Roxo 90 9.95 2.95 29.17
Mean 8.64 3.29 28.20
CV (%) 12.68 10.56 9.70

115
 Talismã
Ouro Negro
Horizonte
IAC Boreal
Roxão
Jalo Precoce
GEN99TG50-47
GEN99TGR1-10
Rubi
GEN99TG3450
Rosinha G2
Princesa
Safira
Ros Brilhosa
Macanudo
IAC Alvorada
Pitanga
CV-48
Diamante Neg
Rosinha Cipó
LP01-38
Grafite
Colibri
Chapecó
Xamego
Valente
Rosado
Ônix
Roxo 90
Tybatã
Roxinho
IAC UNA
Thayú
Sangue Boi
Rosinha
Vereda
FT Nobre
Xodó
Requinte
Triunfo
Carioca 80
CNF 10
Pérola
IAPAR 31
Roxão EEP
Aporé
Pioneiro
Pontal
Timbó
Rudá

0 1 2 3
Euclidean distance

FIG. 5. Dendrogram resulting from hierarchical cluster analysis of 50 common bean genotypes, based
on grain dry matter yield (Gdm), P concentration and P content in Gdm at the second sampling.

Talismã did not form a group with other genotypes because it had low grain yield, but
high P concentration and P content of Gdm. Therefore, other groups were classified in four
groups in terms of grain yield and P in Gdm.

Carioca 80, CNF 10, Pérola, IAPAR 31, Roxão EEP, Aporé, Pioneiro, Pontal, Timbó
and Rudá were the most productive genotypes when grown under low P availability
conditions. The 10 bean genotypes most productive in terms of grain had higher P utilization
efficiency, defined as the ability to convert the element taken up by the plants into agricultural
product of commercial value (leaf, fruit, root and stem) [21]. The efficiency of utilization
(EU) is generally associated with productivity, that is, the greater the EU, the higher is the
grain yield [22]. Among these genotypes, only Pioneer was rated as efficient in P uptake (Fig.
3).

116
4. CONCLUSIONS

− The common bean genotypes Sangue de Boi, Rosinha, Thayú, Grafite, Horizonte,
Pioneiro and Jalo Precoce were the most efficient in P uptake;

− The common bean genotypes Carioca 80, CNF 10, Pérola, IAPAR 31, Roxão EEP,
Aporé, Pioneiro, Pontal, Timbó e Rudá were the most productive in grain under
conditions of low available soil P (genotype more efficient in P utilization);

− The seed derived P in the plant, when the 32P L-value technique is used, affects the
identification and classification of common bean genotypes.

ACKNOWLEDGMENTS

The work was supported by IAEA (International Atomic Energy Agency) research
contract 13779. Vinícius Ide Franzini acknowledges a graduate fellowship from CAPES
(Coordenação de Aperfeiçoamento de Pessoal de Nível Superior, Brazil).

REFERENCES

[1] BARBOSA FILHO, M.P., FAGERIA, N.K., SILVA, O.F., Aplicação de Nitrogênio
em Cobertura no Feijoeiro Irrigado, Circular Técnica 49, EMBRAPA–CNPAF, Santo
Antônio de Goiás (2001) 8 p.
[2] CARVALHO, A.M., et al., Resposta do feijoeiro à aplicação de fósforo em solos dos
Cerrados, Rev. Bras. Ciência Solo 19 (1995) 61–67.
[3] OLIVEIRA, I.P., et al., Avaliação de cultivares de feijão quanto à eficiência no uso de
fósforo, Pesq. Agropec. Bras. 22 (1987) 39–45.
[4] LANA, R.M.Q., et al., Variabilidade entre genótipos de feijoeiro na eficiência no uso
do fósforo, Ciência Rural 36 (2006) 778–784.
[5] FAGERIA, N.K., BARBOSA FILHO, M.P., STONE, L.F., “Nutrição de fósforo na
produção de feijoeiro”, Proc. Symp. Phosphorus in Brazilian Agriculture, Associação
Brasileira para Pesquisa da Potassa e do Fosfato, Piracicaba 17 (2004) 435–455.
[6] CISSE, L., AMAR, B., The importance of phosphate fertilizer for increased crop
production in developing countries, Proc. AFA 6 th Int. Annual Conf. Cairo, Egypt,
IMPHOS (2000) 1–17. http://www.imphos.org/download/ImphosPaper5.pdf
[7] YAN, X., BEEBE, S.E., LYNCH, J.P., Genetic variation for phosphorus efficiency of
common bean in contrasting soil types: II yield response, Crop Sci. 35 (1995) 1094–
1099.
[8] LYNCH, J.P., BROWN, K.M., Topsoil foraging-an architectural adaptation of plants
to low phosphorus availability, Plant Soil 237 (2001) 225–237.
[9] DOS SANTOS H.G., et al., Sistema Brasileiro de Classificação de Solos, 2nd Edn,
EMBRAPA-Solos, Rio de Janeiro (2006) 306 p.
[10] VAN RAIJ, B., et al., Análise Química para Avaliação da Fertilidade de Solos
Tropicais, Instituto Agronômico, Campinas (2001) 285 p.
[11] EMBRAPA-SOLOS, Manual de Métodos de Análises de Solos, 2nd Edn, Rio de
Janeiro, (1997) 212 p.
[12] VAN RAIJ, B., et al., Recomendação de adubação e calagem para o Estado de São
Paulo, 2nd Edn, Instituto Agronômico e Fundação IAC, Campinas (Boletim Técnico,
100) (1997) 285 p.

117
[13] MURAOKA, T., et al., “Comparison of the ability of different plant species and corn
hybrids to access poorly-available soil phosphorus in an Oxisol of the Cerrado region,
Brazil”, Management Practices for Improving Sustainable Crop Production in Tropical
Acid Soils, Proceedings Series, IAEA, Vienna (2006) 137–146.
[14] HOCKING, P.J., et al., “Comparation of the ability of different crop species to access
poorly-available soil phosphorus”, Plant Nutrition for Sustainable Food Production
and Environment, (ANDO, T., et al., Eds), Kluwer Academic Publishers (1997) 305–
308.
[15] BROOKES, P.C., Correction for seed-phosphorus effects in L-value determinations, J.
Sci. Food Agric. 33 (1982) 329–335.
[16] LARSEN, S., The use of 32P in studies of the uptake of phosphorus by plants, Plant
Soil 4 (1952) 1–10.
[17] SAS INSTITUTE, SAS User’s Guide: Statistics, Version 8.2, SAS Institute, Cary, NC
(2001).
[18] ARAUJO, A.P., MACHADO, C.T.T., “Fósforo”, Nutrição Mineral de Plantas,
(FERNANDES, M.S., Ed.), Universidade Federal de Viçosa, Viçosa, MG (2006)
253–280.
[19] WANKE, M., et al., Response to phosphate deficiency in bean (Phaseolus vulgaris L.)
roots. Respiratory metabolism, sugar localization and changes in ultrastructure of bean
root cells, Ann. Bot. 82 (1998) 809–819.
[20] LYNCH, J., Root architecture and plant productivity, Plant Physiol. 109 (1995) 7–13.
[21] MALAVOLTA, E., Manual de Nutrição Mineral de Plantas, Editora Agronômica
Ceres, São Paulo (2006) 638 p.
[22] MALAVOLTA, E., AMARAL, F.A.L., “Nutritional efficiency of 104 bean varieties
(Phaseolus vulgaris L.)”, Plant Nutrition 1978, Proc. 8th Int. Colloq. Plant Analysis
and Fertilizer Problems, (FERGUSON, A.R., BIELSKI R.L., FERGUSON, I.B., Eds),
NZ DSIR Information Series No. 34, Government Printer, Wellington (1978) 313–318.

118
 SELECTION OF COMMON BEAN LINES, RECOMBINANT INBRED LINES
AND COMMERCIAL GENOTYPES TOLERANT TO LOW PHOSPHORUS
AVAILABILITY IN AN ACRISOL SOIL ON THE BASIS OF ROOT TRAITS
AND GRAIN YIELD

A. GARCIA, L.A. GOMEZ, A. MORALES, M. MOSQUERA, J. PASCUAL, M.

HERNANDEZ, N. CHÁVEZ, J.M. DANTIN
Instituto de Suelos,
MINAG,

G. HERRERO
Instituto de Ecología y Sistemática,

S. CURBELO, J.L. REYES, J.A. FUSTES, E.E. RODRIGUEZ, S. GONZÁLEZ, R.

RAMOS
Estación Experimental Forestal Viñales,
MINAG,

Cuba

J.J. DREVON
Eco & Sols, INRA-IRD- SupAgro,
Montpellier,
France

Abstract

Common bean (Phaseolus vulgaris L.) is the most important food legume for human
consumption worldwide and especially in Latin America and Africa, but low soil phosphorus (P)
availability limits grain production in these areas. For these reason eighty five recombinant inbred
lines (RILs) of BAT 477 x DOR 364 and twenty commercial bean genotypes were sown in plots in an
Acrisol soil with low P availability to evaluate nine root traits and grain yield. The study was carried
out in Pinar del Rio province in Cuba between November 2006 and February 2009. The plots received
basal fertilization (N and K) and P fertilization between 15 and 90 kg P2O5 ha-1. Ten plants were
sampled from each plot at R6 pod fill to evaluate root traits and shoot biomass, and at R9 physiological
maturity to estimate grain yield. The 85 RILs showed great variability for root traits, grain yield and P
stress tolerance calculated as relative grain yield. The commercial bean lines also showed large
diversity in yield parameters. Principal Component Analysis showed that there were high and
significant correlations between root traits (basal root number, primary root depth, adventitious root
length and adventitious root number) and grain yield parameters (grain yield at 15 P level and relative
grain yields). Adventitious root traits showed the greatest correlation with yield under low P.
Promising RILs included 75.1.1, 60.1.1, 38.1.1, 14.1.1 and 38.1.1 and promising commercial bean
lines included ICA Pijao, BAT 482, ICA 23, BAT 24 and BAT 832.

1. INTRODUCTION

Common bean (Phaseolus vulgaris, L.) is the most important food legume for human
consumption worldwide, and especially in Latin America and Africa [1]. Low phosphorus (P)
availability is a primary constraint to common bean production, and P fertilization is
frequently not economically feasible for resource-limited farmers in the developing world,
and is often inefficient in tropical soils due to fixation of P in insoluble forms by iron and
aluminium oxides. Therefore fertilization is not an adequate solution in itself; rather, bean

119
genotypes that are capable of producing economic yields at low levels of native or added P
are desirable. Such genotypes may be superior due to a greater ability to recover P from
fertilizer and / or native fractions of soil P and / or due to greater productivity per unit of P
absorbed [2].

Substantial genotypic variation in adaptation of bean to low P availability has been

linked with root traits that enhance the efficiency of soil foraging [3]. For instance, it was
affirmed that root gravitropism determines the relative distribution of plant roots in different
soil layers [4], and therefore may influence the acquisition of shallow soil resources.
Adventitious roots play an important role in P acquisition, as they are localized near the soil
surface where P is relatively abundant [5]. The effect of low soil P availability on the increase
of adventitious rooting in common bean could be used as a strategy to increase P acquisition
for this legume [6].

Soil P deficiency is widespread as a limiting factor for food production in the tropics
and subtropics because most of the forms in which it occurs are poorly available to crops [7].
P fertilization with water soluble sources (e.g. single and triple super phosphate, DAP, MAP
in Cuba) is a solution to overcome this constraint, but third-world small farmers, including
small farmers in Cuba, generally lack the resources to invest in this input. In addition, the
efficiency of P fertilization is low in many tropical and subtropical soils because of fixation
into plant-unavailable forms.

At present Cuban agriculture faces a fertilizer deficit with negative consequences for
the quantity and quality of agricultural production. Soil nutrient deficiencies, mainly nitrogen
(N) and P, coupled with soil acidity occur in many of the small agricultural areas. The use of
plants with low adaptability to these constraining factors, as well as inadequate management
practices among others, limit agricultural production. At the same time there is no
improvement of soil fertility and consequently agriculture production. In global terms, 45% of
agricultural lands in Cuba have low fertility and 40% are affected by acidity.

Therefore, as a strategy to improve common bean production in tropical and

subtropical infertile soils, it is essential to evaluate plant root traits linked to high P
acquisition from low P bioavailability through the selection of more adapted germplasm to
soil constraints. This work was carried out to identify common bean lines (RILs and
commercial cultivars) for use in Cuba that were more tolerant of P deficiency in an Acrisol
soil on the basis of root traits and grain yield.

2. MATERIALS AND METHODS

Three experiments were carried out to characterize the root traits of 106 common bean
lines (85 recombinant inbred lines and 19 commercial lines) used in Cuba.

2.1. Laboratory screening

2.1.1. Experiment 1

Four plants of 15 recombinant inbred lines (RILs) i.e. (1.1.1, 2.1.1, 5.1.1, 6.1.1, 14.1.1,
16.1.1, 20.1.1, 22.1.1, 27.1.1, 30.1.1, 31.1.1, 33.1.1, 34.1.1, 36.1.1) and two commercial lines
(UPR 56 and UPR 70) were evaluated using Jonathan Lynch’s methodology “cigar roll
method” (personal communication). A rating scale of 1 to 9 was used to rank crown root
architecture (Table 1).

120
TABLE 1. ROOTS TRAITS AND RATING SCALE USED TO EVALUATE BEAN LINES

Root traits Rating scale

1. Adventitious root length (ARL) 1 = 1 cm to 9 = 15–20 cm
2. Adventitious root Number (ARN) Actual number plant-1
3. Adventitious root branching (ARB) 1 = no lateral branching to 9 = multiple lateral branches
with up to 4 orders of branching
4. Basal root length (BRL) 1 = absent to 9 = 20–30 cm (width of excavation)
5. Basal root number (BRN) Actual number of basal roots plant-1
6. Basal root branching (BRB) 1 = no lateral branching to 9 = multiple lateral branches
with up to 4 orders to branching
7. Basal root depth (BRD) 1 = horizontal to 9 = vertical
8. Primary root depth (PRD) 1 = no taproot left to 9 = 20 – 30 cm (depth of
excavation)
9. Primary root branching (PRB) 1 = no lateral branching to 9 = multiple lateral branches
with up to 4 orders of branching
10. Combined adventitious roots Sum of results of rating scale analysis of the 1, 2 and 3
root traits
11. Combined basal roots Sum of results of rating scale analysis of the 4, 5 and 6
root traits
12. Combined primary roots Sum of results of rating scale analysis of the 7, 8 and 9
root traits

2.1.2. Experiment 2

Seventeen commercial common bean lines in Cuba (BAT 24, BAT 58, BAT 93, BAT
304, BAT 482, BAT 832, Guama 23, Rosas M112, Bolita 42, Güira 89, Jamapa, CC – 259R,
ICA Pijao, Bonita 11, Red Kloud and Velazco Largo) were also evaluated using the cigar roll
method (Table 1).

2.2. Field evaluation – Experiment 3

Eighty five recombinant inbred lines (Table 2) as well as BAT 477 and DOR 364 were
evaluated between November 2006 and February 2007 for root traits.
TABLE 2. RECOMBINANT INBRED LINES EVALUATED IN EXPERIMENT 3

Recombinant inbred lines

1.1.1 22.1.1 45.1.1 65.1.1 83.1.1
2.1.1 25.1.1 46.1.1 66.1.1 84.1.1
3.1.1 26.1.1 47.1.1 67.1.1 85.1.1
4.1.1 27.1.1 48.1.1 68.1.1 88.1.1
6.1.1 28.1.1 50.1.1 69.1.1 89.1.1
7.1.1. 30.1.1 53.1.1 70.1.1 90.1.1
8.1.1 31.1.1 54.1.1 71.1.1 91.1.1
9.1.1 33.1.1 55.1.1 72.1.1 92.1.1
10.1.1 34.1.1 56.1.1 73.1.1 93.1.1
12.1.1 35.1.1 57.1.1 74.1.1 94.1.1
13.1.1 35.1.4 58.1.1 75.1.1 95.1.1
14.1.1 36.1.1 59.1.1 76.1.1 96.1.1
15.1.1 38.1.1 60.1.1 77.1.1 97.1.1
17.1.1 39.1.1 61.1.1 79.1.1 98.1.1
19.1.1 41.1.1 62.1.1 80.1.1 99.1.1
20.1.1 42.1.1 63.1.1 81.1.1 128.1.1
21.1.1 43.1.1 64.1.1 82.1.1

121
 The soil was an Acrisol from Viñales, Pinar del Rio province in the west of Cuba. The
physic-chemical characteristics were: pH (H2O), 4.8; pH KCl, 3.8; Organic matter (Walkley-
Black method), 3.02%; available pP (Bray-Kurtz I method, 4.4 mg P kg-1 ; exchangeable bases
(cmol kg-1); Ca, 0.75; Mg, 0.43; Na, 0.10; K, 0.05; N total (Kjeldhal method), 0.115%;
exchangeable acidity 0.380; exchangeable H, 0.013; exchangeable Al, 0.367; Al saturation,
23.1 %.

The plots size was 2.8 m2 (two rows at 0.70 m × 2 m long) and plots were surrounded
by 0.5 m of untreated soil. All plots received basal fertilization of 15 kg−N ha-1 as urea and 60
kg K2O ha-1 as KCl. The common bean seeds were inoculated with Rhizobium sp. strain 6bIII
from the Soil Institute Collection in Cuba.

Each RIL was evaluated at two P levels (15 and 90 kg P2O5 ha-1) using triple super
phosphate (TSP) as the P source. The experiment was arranged in a two factor randomized
block design with 87 genotypes × 2 P fertilization levels, where every row was considered to
be one replicate. Three plants of each plot were sampled at the R6 – R7 growth stage on
January 3, 2007 for root characteristics evaluation (Table 1). Ten plants of each treatment
were harvested at the R9 growth stage (85 to 95 days after sowing) to determine grain yield.

2.3. Statistical analysis

The statistical software MSTAT-C version 2.10 [8] was used to perform the analyses
of variance at P<0.05 for differences of means traits among genotypes, P treatment and
interactions, and to calculate the correlation between parameters. The root traits evaluated and
grain yield parameters were used to perform a Principal Component Analysis (PCA) using
software XLSTAT [9].

3. RESULTS

3.1. Root traits in bean lines evaluated under laboratory conditions

Great differences in eight roots traits were found in 17 bean lines evaluated in the first
experiment. The rating scale used during root evaluation showed that ARL ranged from 2 to
3, ARN from 1 to 5, ARB from 1 to2, BRL from 2 to 4, BRN from 1 to 5, BRB from 1 to 3,
PRD from 4 to 7 and PRB from 2 to 5 (Table 3).

TABLE 3. MINIMUM, MAXIMUM, AVERAGE AND STANDARD DEVIATION OF 8 ROOT

TRAITS IN 15 RILS AND 2 COMMERCIAL LINES OF COMMON BEAN IN THE FIRST
LABORATORY EXPERIMENT

Root traits Minimum Maximum Average Standard deviation

1. Adventitious root length (ARL) 2 3 2.84 0.43
2. Adventitious root number (ARN) 1 5 3.38 0.88
3. Adventitious root branching (ARB) 1 2 1.10 0.20
4. Basal root length (BRL) 2 4 3.66 0.62
5. Basal root number (BRN) 1 5 3.56 1.06
6. Basal root branching (BRB) 1 3 1.44 0.54
7. Primary root depth (PRD) 4 7 5.75 0.71
8. Primary root branching (PRB) 2 5 3.54 0.76

Principal component analysis (PCA) showed that the first fourth components
explained more than 80% the variance (Table 4). Correlated positively to F1 were
adventitious root length, and adventitious root number (Table 5). On the contrary, negatively

122
related were basal root number, primary root depth and primary root branching. Positively
related to F2 were adventitious root numbers, basal root length, primary root depth and
primary root branching but negatively correlated with basal root number. The F3 component
was only positively linked to two variables, adventitious root branching and basal root
branching (Table 5). Adventitious root branching, basal root length and basal root number
were correlated with F4.

TABLE 4. PRINCIPAL COMPONENT ANALYSIS OF 8 ROOT TRAITS IN 15 RILS AND 2

COMMERCIAL LINES OF COMMON BEAN IN THE FIRST LABORATORY EXPERIMENT

Parameter F1 F2 F3 F4
Characteristic value 2.20 1.94 1.43 0.92
Variability (%) 27.47 24.25 17.88 11.47
% accumulated 27.47 51.72 69.61 81.09

TABLE 5. CORRELATION COEFFICIENTS BETWEEN 8 ROOT TRAITS AND FACTORS

FROM THE PRINCIPAL COMPONENT ANALYSIS OF 15 RILS AND 2 COMMERCIAL
COMMON BEAN LINES IN THE FIRST LABORATORY EXPERIMENT

Root traits F1 F2 F3 F4
1. Adventitious root length (ARL) 0.80 0.29 0.21 0.13
2. Adventitious root number (ARN) 0.44 0.76 – 0.17 0.08
3. Adventitious root branching (ARB) 0.17 0 0.81 0.33
4. Basal root length (BRL) – 0.28 0.58 0.27 0.38
5. Basal root number (BRN) – 0.76 – 0.38 – 0.03 0.48
6. Basal root branching (BRB) – 0.17 – 0.18 0.79 –0.45
7. Primary root depth (PRD) – 0.61 0.67 0.03 0.06
8. Primary root branching (PRB) – 0.53 0.57 0.04 –0.45

The PCA analysis showed that the lines with the best adventitious root traits were
36.1.1 and 27.1.1 (Fig. 1). The lines with the highest primary root depth, primary root
branching and basal root lengths were 20.1.1, 30.1.1 and 31.1.1 and the lines with the most
basal root number and basal root branching were 16.1.1 and 20.1.1.

123
 F1 + F2 = 51,7 %
4

3 22.1.1 ARN
PRD 36.1.1
PRB BRL 27.1.1
2 30.1.1
31.1.1 ARL
1 38.1.1
34.1.1
F2 (24,25 %)

14.1.1 ARB
0 5.1.1
33.1.1
UPR70
-1 BRB 2.1.1.
16.1.1 20.1.1
BRN 1.1.1
-2

-3
UPR 56
-4
6.1.1
-5
-7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6

F1 (27,48 %)

FIG. 1. Coordinates of variables and factors in principal component analysis of 15RILs and 2
commercial lines (numbered) of common bean in the first laboratory experiment. Root trait
abbreviations are given in Table 3.

Dendrogram analysis permitted the identification of three clusters according to root

traits for 17 bean lines evaluated in the first laboratory experiment (Fig. 2, Table 6). Class 1
(C1) included 6 lines (1.1.1, 2.1.1, 6.1.1, 20.1.1, UPR 56 and UPR 70) having low production
of adventitious root number (ARN); Class 2 (C2) included nine lines (5.1.1, 14.1.1, 16.1.1,
22.1.1, 30.1.1, 31.1.1, 33.1.1, 34.1.1 and 38.1.1) having high primary root depth (PRD), basal
root length and basal root number and Class 3 (C3) included 2 lines (27.1.1 and 36.1.1)
having high adventitious root number (ARN) and adventitious root length (ARL).

18
17
16
16
14
15
12
14
Dissimilitude

10
13
8
12
6
11
4
10
2 9 2 6
9
0
UPR 56

UPR70

C2 C3 C1
2.1.1.
22.1.1
14.1.1
33.1.1
30.1.1
31.1.1
16.1.1
34.1.1
5.1.1
38.1.1
27.1.1
36.1.1
6.1.1
1.1.1
20.1.1

FIG. 2. Dendrogram of 8 root traits in 15 RILs and 2 commercial lines of common bean in the first
laboratory experiment.

124
TABLE 6. CLASS “CENTROID” OF DENDROGRAM ANALYSIS OF 8 ROOT TRAITS OF 15
RILS AND 2 COMMERCIAL LINES OF COMMON BEAN IN THE FIRST LABORATORY
EXPERIMENT

Class ARL ARN ARB BRL BRN BRB PRD PRB

1 2.88 3.00 1.13 3.13 3.33 1.50 5.08 3.04
2 2.67 3.28 1.11 4.00 4.14 1.50 6.22 3.92
3 3.50 5.00 1.00 3.75 1.63 1.00 5.63 3.38

PCA showed that commercial lines evaluated in the second experiment were more
variable in root traits than RILs (Table 7). ARL ranged from 2 to 5, ARN from 2 to 6, ARB
from 1 to 3, BRL from 1 to 6, BRN from 0 to 5, BRB from 0 to 4, BRD from 0 to 7, PRD
from 4 to 7 and PRB from 1 to 6. The first four components of PCA explained 82 % of the
variance (Table 8). Positive correlations with F1 were BRL, BRN, BRB and BRD, but
negatively ARL (Table 9). Positively correlated to F2 were ARL, ARN, BRL and PRD, but
negatively correlated with ARB and PRD. Correlated positively to F3 were ARL, BRB and
PRB, but negatively with PRD. F4 was negatively correlated with ARN and PRB.

Dendrogram analysis identified fives class for the 17 commercial lines evaluated
(Table 10, Fig. 3, Fig. 4). Class 1 included 3 lines (BAT 24, Guama 23 and Velazco Largo)
with high ARL, ARN, BRL, BRD and PRD. Class 2 included 7 lines (BAT 58, BAT 832,
Rosas, M 112, Bolita 42, Bonita 11 and Red Kloud) with medium to high BRD and PRD.
Class 3 included 1 line (BAT 93) with low production of BRN, BRB, BRD and PRD. Class 4
included 3 lines (BAT 304, BAT 482 and ICA Pijao) with high BRD and PRD, and Class 5
included 3 lines (Guira 89, Jamapa and CC 25 – 9R) with high BRD and ARL.

TABLE 7. MINIMUM, MAXIMUM, AVERAGE AND STANDARD DEVIATION OF 9 ROOT

TRAITS IN 17 COMMERCIAL COMMON BEAN LINES IN THE SECOND LABORATORY
EXPERIMENT

Root traits Minimum Maximum Average Standard deviation

1. Adventitious root length (ARL) 2 5 3.2 0.67
2. Adventitious root number (ARN) 2 6 3.9 1.06
3. Adventitious root branching (ARB) 1 3 1.2 0.43
4. Basal root length (BRL) 1 6 3.7 1.07
5. Basal root number (BRN) 0 5 3.5 1.26
6. Basal root branching (BRB) 0 4 1.5 0.75
7. Basal root depth (BRD) 0 7 5.6 1.58
8. Primary root depth (PRD) 4 7 5.5 0.66
9. Primary root branching (PRB) 1 6 3.2 1.04

TABLE 8. PRINCIPAL COMPONENT ANALYSIS OF 9 ROOT TRAITS IN 17 COMMERCIAL

COMMON BEAN LINES IN THE SECOND LABORATORY EXPERIMENT

Parameter F1 F2 F3 F4
Characteristic value 3.22 1.87 1.35 0.94
Variability (%) 35.88 20.77 15.02 10.44
% accumulated 35.88 56.65 71.67 82.12

TABLE 9. CORRELATION COEFFICIENTS BETWEEN 9 ROOT TRAITS AND FACTORS

FROM THE PRINCIPAL COMPONENT ANALYSIS OF 17 COMMERCIAL COMMON BEAN
LINES IN THE SECOND LABORATORY EXPERIMENT

125
Root traits F1 F2 F3 F4
1. Adventitious root length (ARL) –0.476 0.553 0.574 0.247
2. Adventitious root number (ARN) 0.014 0.831 0.288 –0.100
3. Adventitious root branching (ARB) 0.162 –0.454 0.211 0.795
4. Basal root length (BRL) 0.835 0.458 –0.218 0.008
5. Basal root number (BRN) 0.881 0.014 –0.089 0.150
6. Basal root branching (BRB) 0.704 0.138 0.521 0.009
7. Basal root depth (BRD) 0.883 0.107 –0.239 0.015
8. Primary root depth (PRD) –0.187 0.543 –0.324 0.412
9. Primary root branching (PRB) 0.438 –0.362 0.637 –0.210

TABLE 10. CLASS “CENTROID” OF DENDROGRAM ANALYSIS OF 8 ROOT TRAITS OF 17

COMMERCIAL BEAN LINES IN THE SECOND LABORATORY EXPERIMENT

Class ARL ARN ARB BRL BRN BRB BRD PRD

1 3.7 5.6 1.2 5.3 4.7 1.8 6.8 6.0
2 2.9 3.0 1.4 3.6 4.1 1.5 5.5 5.4
3 4.5 5.0 1.0 1.0 0.0 0.0 0.0 5.8
4 3.1 3.5 1.0 3.3 2.3 1.2 6.0 6.0
5 3.2 4.3 1.2 3.8 3.3 2.3 6.3 4.6

F1 + F2 = 56,6 %
6
Velasco L.
5

4 ARN BAT 24

3 ARL PRD
BRL
2
BAT 93
F2 (20,7 %)

CC 25-9 R BRD
BAT 482
1 Guamá 23BRB
BAT 304
0 Red Kloud
BRN

-1 Ica Pijao Jamapa

Bolita 42 BAT 58
Rosas Güira 89
-2 PRB Bonita 11
BAT 832 ARB M 112
-3

-4

-5
-12 -10 -8 -6 -4 -2 0 2 4 6 8 10
F1 (35,8 %)

FIG. 3. Coordinates of variables and factors in principal component analysis of 17 commercial lines
(numbered) of common bean in the second laboratory experiment. Root trait abbreviations are given
in Table 7.

126
 70

60

50

Dissimilitude
40

30

20

10

CC 25-9 R
Guamá 23

Ica Pijao
Jamapa
M 112

Red Kloud

Güira 89
Rosas
BAT 93

BAT 24

Bolita 42
BAT 58

BAT 832
Bonita 11
BAT 304
BAT 482
Velasco L.

67

57

47
Dissimilitude

37

27

17
7 3
1 3 3
7
C3 C1 C2 C4 C5

FIG. 4. Dendrogram of 9 root traits in 17 commercial lines of common bean in the second laboratory
experiment

127
 75.1.1
28.1.1 14.1.1
76.1.1 4.1.1
62.1.1 97.1.1
59.1.1 59.1.1
60.1.1 98.1.1
75. 1.1 60.1.1
65. 1.1 28.1.1
3.1.1 128.1.1
95. 1.1 45.1.1
97. 1.1 92.1.1
35. 1.1 67.1.1
68. 1.1 95.1.1
84. 1.1 62.1.1
94. 1.1 15 94.1.1
2.1.1 43.1.1
98.1.1 90 35.1.1
4.1.1 91.1.1
77.1.1 76.1.1
10.1.1 79.1.1
70.1.1 77.1.1
1.1.1 89.1.1
13.1.1 47.1.1
92.1.1 31.1.1
21.1.1 26.1.1
69.1.1 84.1.1
128 .1.1 96.1.1
91.1.1 88.1.1
6.1. 1 34.1.1
15.1.1 69.1.1
80.1.1 1.1.1
7.1. 1 19.1.1
67.1.1 90.1.1
8.1. 1 83.1.1
89.1.1 15.1.1
17.1.1 7.1.1
9.1. 1 65.1.1
45.1.1 33.1.1
61. 1.1 99.1.1
41. 1.1 48.1.1
14. 1.1 68.1.1
48. 1.1 70.1.1
25. 1.1 58.1.1
12. 1.1 80.1.1
79. 1.1 46.1.1
46. 1.1 85.1.1
82. 1.1 41.1.1
43. 1.1 82.1.1
83.1.1 2.1.1
53.1.1 3.1.1
96.1.1 74.1.1
26.1.1 35.1.4
99.1.1 12.1.1
34.1.1 10.1.1
72.1.1 8.1.1
33.1.1 38.1.1
58.1.1 72.1.1
30.1.1 56.1.1
31.1.1 6.1.1
88.1.1
38.1.1 93.1.1
47.1.1 21.1.1
90.1.1 64.1.1
22.1.1 61.1.1
55.1.1 13.1.1
85.1.1 9.1.1
93.1.1 53.1.1
56.1.1 36.1.1
36.1.1 66.1.1
66.1.1 22.1.1
73. 1.1 20.1.1
19. 1.1 50.1.1
27. 1.1 17.1.1
DOR 364 DOR364
35. 1.4 55.1.1
64. 1.1 25.1.1
BAT 477 BAT 477
81. 1.1 63.1.1
71. 1.1 71.1.1
50. 1.1 73.1.1
54.1.1 54.1.1
63.1.1 27.1.1
39.1.1 39.1.1
57.1.1 81.1.1
74.1.1 42.1.1
42.1.1 30.1.1
20.1.1 29.1.1
29.1.1 57.1.1

0 1 2 3 4 5 6 7 8 0 25 50 75 100 125 150 175 200 225 250

Grain Yield (g plant-1) Relative Grain Yield at Low P (%)

FIG. 5. Grain yield and relative grain yield of 85 RILs, BAT 477 and DOR 364 at two P levels in the
field experiment.

128
3.2. Root traits and grain yield in bean lines evaluated under field conditions

Grain yield (GY) and relative grain yield (RGY) at 15 and 90 P2O5 fertilization rates
varied among the RILs (Fig. 5). The GY at 15 P ranged from 1.77 to 6.43 g plant -1 while at 90
P from 2.36 to 7.84 g plant-1. The relative grain yield (RGY) ranged from 43 to 234 %.

Large differences in root traits were observed in 85 bean RILs evaluated in the field
experiment (Table 11). When the rating scaling was applied to ARL, ARN and ARB root
traits the values ranged from 2 to 8, 2 to 13 and 1 to 8, respectively. The basal root branching,
number and depth were fewer than adventitious traits, so BRB, BRN and BRD ranged from 2
to 7, 2 to 5 and 1 to 6, respectively (Table 11). Primary root traits depth and primary root
branching ranged from 1 to 4 and 1 to 3, respectively.

TABLE 11. MINIMUM, MAXIMUM, AVERAGE AND STANDARD DEVIATION OF 9 ROOT

TRAITS AND 2 YIELD PARAMETERS IN 85 RILS OF COMMON BEAN GROWN AT 2 P
LEVELS IN THE FIELD EXPERIMENT

Root traits and yield parameters Minimum Maximum Average Standard deviation
1. Adventitious root length (ARL) 2 8 4.11 1.32
2. Adventitious roots numbers (ARN) 2 13 6.17 2.44
3. Adventitious root branching (ARB) 1 8 2.46 0.90
4. Basal root length (BRL) 2 7 3.37 0.89
5. Basal root number (BRN) 2 5 3.20 0.79
6. Basal root branching (BRB) 1 6 2.67 1.14
7. Basal root depth (BRD) 1 7 4.28 1.63
8. Primary root depth (PRD) 1 4 2.17 0.98
9. Primary root branching (PRB) 1 3 1.75 0.65
10. Grain yield at 15P (g plant-1) 1.77 6.43 4.28 1.17
11. Relative grain yield (%) 42.78 233.69 98.88 33.08

PCA performed with nine root traits and two parameters of grain yield showed that the
first five components explained 77 % of the variance (Table 12).

TABLE 12. PRINCIPAL COMPONENT ANALYSIS OF 9 ROOT TRAITS AND 2 YIELD

PARAMETERS IN 85 RILS OF COMMON BEAN GROWN AT TWO P LEVELS IN THE FIELD
EXPERIMENT

Parameter F1 F2 F3 F4 F5
Characteristic value 3.32 2.59 1.85 1.75 1.30
Variability (%) 23.74 18.49 13.20 12.50 9.27
% accumulated 23.74 42.23 55.43 67.93 77.20

Thus PCA showed that the F1 component was correlated with all traits studied except
basal root length, grain yield at 15P and relative grain yield, which were negatively correlated
(Table 13). Correlated with F2 were adventitious root length, adventitious root number, grain
yield at 15P, and relative grain yield, but correlated negatively with primary root depth and
primary root branching. Correlated to F3 were basal root number, primary root depth, primary
root branching, grain yield at 15P and relative grain yield. Correlated with F4 and F5
components were basal root depth, basal root number, adventitious root number, primary root
depth and primary root branching (Table 13). Higher correlations were observed between the
yield parameters (GY at 15P and RGY) and F2 and F3 components, and also between these
components and adventitious root length, adventitious root number, basal root number and
primary root depth.

129
 Results of the PCA (Fig. 6, Table 14 and Fig. 7,) show the greatest ARL and ARN
were found in RILs 22.1, 36.1.1, 47.1.1, 97.1.1 and 99.1.1; the greatest BRD, BRN and ARB
were observed in RILs 77.1.1, 71.1.1 and 22.1.1; superior production of BRB was seen in
RILs 39.1.1, 96.1.1 and 30.1.1; greater PRD and PRB were observed in RILs 56.1.1, 72.1.1,
and 54.1.1 and lines BAT 477 and DOR 364; high BRL was observed in RILs 48.1.1 and
7.1.1; lines with high grain yield at 15P and relative grain yield were RILs 60.1.1, 75.1.1,
85.1.1, 95.1.1 and 99.1.1.

TABLE 13. CORRELATION COEFFICIENTS BETWEEN 9 ROOT TRAITS, 2 YIELD

PARAMETERS AND FACTORS FROM THE PRINCIPAL COMPONENT ANALYSIS OF 85
RILS OF COMMON BEAN GROWN AT 2 P LEVELS IN THE FIELD EXPERIMENT

Root traits and yield parameters F1 F2 F3 F4 F5

1. Adventitious root length (ARL) 0.559 0.635 0.236 –0.102 0.185
2. Adventitious roots numbers (ARN) 0.373 0.708 0.165 –0.073 0.402
3. Adventitious root branching (ARB) 0.695 0.068 0.037 –0.315 –0.323
4. Basal root length (BRL) –0.432 –0.149 –0.128 –0.246 0.389
5. Basal root number (BRN) 0.252 0.095 0.520 0.534 –0.350
6. Basal root branching (BRB) 0.609 –0.183 0.187 –0.454 –0.343
7. Basal root depth (BRD) 0.304 0.106 0.118 0.791 0.189
8. Primary root depth (PRD) 0.279 –0.432 0.544 –0.144 0.406
9. Primary root branching (PRB) 0.489 –0.366 0.373 –0.042 0.514
10. Grain yield at 15P (g plant-1) –0.456 0.396 0.638 –0.180 –0.039
11. Relative grain yield (%) –0.350 0.556 0.605 –0.221 0.013

F1 + F2 = 36,96 %)
6

5
75 47 97

4
95 94
99
ARN 36
3
61 90 28
85 RGY ARL 22
60 83 82
2 93
89 GY15
F2 (16,10 %)

91 14 98 77
76 65 4
92
19 BRD
1 128 4667 70 27
BRN 71
35
1 35,4 ARB
84
0 66
59 68
62 33 2
13
69
9 38
45 BRL3 81 96
48 88
-1 12 6 58 20 30BRB 39
7 41 26 31 10
8 17 42 PRB
53 64
-2 80 73 PRD577915 50 54
21 74 72
34 43 55 477
364
-3 25
29
63 56
-4
-8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8
F1 (20,86 %)

FIG. 6. Coordinates of variables and factors in principal component analysis of 85 RILs (numbered)
of common bean at two P levels in the field experiment. Root trait abbreviations are given in Table 11.

130
TABLE 14. CLASS “CENTROID” OF DENDROGRAM ANALYSIS ON 85 RILS, BAT 477 AND
DOR 364 FOR 9 ROOT TRAITS AND 2 GRAIN YIELD PARAMETERS AT 2 P LEVELS IN THE
FIELD EXPERIMENT

Class ARL ARN ARB BRL BRN BRB BRD PRD PRB GY 15 RGY
1 3.9 6.2 2.3 3.6 3.2 2.5 4.2 2.2 1.7 4.5 103.6
2 4.4 6.7 2.5 2.8 3.3 2.8 3.8 2.3 1.6 5.5 154.3
3 4.4 5.9 2.8 3.2 3.1 3.0 4.6 2.1 1.9 3.1 59.3

51747

46747

41747
Disimility

36747

31747

26747

21747
23 52 12
16747
C3 C1 C2

C3 C1 C1 C2
9.1.1 1.1.1 61.1.1 4.1.1
17.1.1 2.1.1 62.1.1 14.1.1
20.1.1 6.1.1 64.1.1 28.1.1
22.1.1 7.1.1 65.1.1 45.1.1
25.1.1 8.1.1 68.1.1 59.1.1
27.1.1 10.1.1 69.1.1 60.1.1
29.1.1 12.1.1 70.1.1 67.1.1
30.1.1 13.1.1 72.1.1 75.1.1
36.1.1 15.1.1 74.1.1 92.1.1
39.1.1 19.1.1 76.1.1 97.1.1
42.1.1 21.1.1 77.1.1 98.1.1
50.1.1 26.1.1 79.1.1 128.1.1
53.1.1 31.1.1 80.1.1
54.1.1 33.1.1 82.1.1
55.1.1 34.1.1 83.1.1
57.1.1 35.1.1 84.1.1
63.1.1 35.1.1 85.1.1
66.1.1 38.1.1 88.1.1
71.1.1 41.1.1 89.1.1
73.1.1 43.1.1 90.1.1
81.1.1 46.1.1 91.1.1
BAT 477 47.1.1 93.1.1
DOR 364 48.1.1 94.1.1
56.1.1 95.1.1
58.1.1 96.1.1

FIG. 7. Dendrogram of 85 RILs of common bean, and lines BAT 477 and DOR 364 for 9 root traits
and 2 yield parameters at 2 P levels in the field experiment.

131
4. DISCUSSION

In summary, there was great variability for root traits and grain yield in the 85 RILs
and 20 commercial lines of common bean evaluated. High and significant correlations were
found for grain yield or relative grain yield at 15P and basal root number, primary root depth,
adventitious root length and adventitious root numbers in the 85 RILs. Lines identified under
laboratory conditions, such as 22.1.1, with favourable primary and basal root traits were also
found to be promising in the field. This high diversity is desirable both to select contrasting
RILs and commercial common bean cultivars as potentially elite material for breeding
programmes to improve grain yield at low soil P, and also because of the interest in
introducing material with high adaptability to low P input systems (soil and P fertilizer) in
small farmer’s plots.

In addition, these results are similar to earlier studies which recognized that bean
adaptation to low P availability in soil is associated with root traits [3, 4], and which identified
adventitious roots as playing the principal role in P acquisition [5, 6]. It is likely that this trait
is probably associated with enhanced grain yield when common bean lines are grown in soil
with low P availability as was observed in the present study.

In Cuba, new common bean lines are often introduced for ‘on-farm’ trials, but the
identification of superior genotypes with greater productivity in low-nutrient environments
similar to those utilized by smallholder farmers in Cuba is rarely considered. In this work, the
grain yields at low P availability of RILs such as 2.1.1, 5.1.1, 14.1.1, 22.1.1, 34.1.1, 38.1.1,
60.1.1 and 75.1.1 were observed to be high and stable, and thus theses RILs are promising
material for further ‘on-farm’ studies.

The possibility of identifying promising material in early growth stages with the “cigar
roll” method will permit evaluation of a greater number of lines in a cost-effective way. It was
concluded that there is high variability associated with root traits and grain yield among the
102 bean lines evaluated. Correlations of grain yield parameter and root traits, such as basal
root number, primary root depth, adventitious root length and adventitious root numbers, were
observed.

REFERENCES

[1] CIAT (1993), Bean Program Annual Report, CIAT, Cali, Colombia.
[2] BEEBE, S., et al., A geographical approach to identify phosphorus-efficient genotypes
among landrances and wild ancestors of common bean, Euphytica 95 (1997) 325–326.
[3] BEBEE, S., et al., Quantitative trait loci for root architecture traits correlated with
phophorus acquisition in common bean, Crop Sci. 46 (2006) 413–423.
[4] LIAO, H., et al., Genetic mapping of basal root gravitropism and phosphorus
acqusition efficiency in common bean, Funct. Plant Biol. 31 (2004) 959–970.
[5] OCHOA, I.P.M.W., LYNCH, J.P., QTL analysis of adventitous root formation in
common bean (Phaseolus vulgaris L.) under contrasting phophorus availability, Crop
Sci. 46 (2006) 1609–1621.
[6] MILLER, C.R., et al., Genetic variation for adventitious rooting in reponse to low
phosphorus availability: potential utility for phosphorus axquisition from stratified
soils, Funct. Plant Biol. 30 (2003) 973–985.
[7] ZAPATA, F., “Evaluating agronomic effectiveness of phosphate rocks using nuclear
and related techniques: Results from an FAO/IAEA Co-ordinated Research Project”,
Management and Conservation of Tropical Acid Soils for Sustainable Crop Production,
IAEA-TECDOC-1159, Vienna (2000) 91–100.

132
[8] STATISTICAL SOFTWARE, MSTAT-C version 2.10 (RUSSELL, D.F., Ed.), Crop
and Soil Science Department, Michigan State University (1988).
[9] THE STATISTIC SYSTEM, XLSTAT version 2010, Addinsoft SARL, Paris (2010).

133
 PHOSPHORUS USE EFFICIENCY FOR SYMBIOTIC FIXATION
NITROGEN IN VOANDZOU (VIGNA SUBTERRANEA) USING ISOTOPIC
EXCHANGE METHOD IN RHIZOTRON

A. ANDRIAMANANJARA, L. RABEHARISOA
Laboratoire des Radio-isotopes,
Université d'Antananarivo,
Antananarivo,
Madagascar

D. MASSE
Institut de Recherche pour le Développement,
UMR Eco&Sols,
Montpellier,

L. AMENC, C. PERNOT, J.J. DREVON

Institut National de la Recherche Agronomique,
UMR Eco&Sols,
Montpellier,

C. MOREL
INRA-ENITA,
Villenave d'Ornon,

France

Abstract

Low bioavailability of nitrogen and phosphorus is one of the main constraints in the acid soils
with high P-fixing capacity. Plants adapt to low nutrient availability through various biological and
physico-chemical mechanisms. Since genetic variation of N2 fixation exists in numerous legume
species, optimization of symbiotic nitrogen fixation (SNF) under P deficiency could be a way to the
replenishment of soil fertility in tropical soils. As the genetic potential of crops like Vigna subterranea
(Bambara groundnut or voandzou) is little studied, although its agronomic potential is interesting for
the farmers of Africa, a physiological study through legume screening for N2 fixation was performed
with 54 cultivars from Madagascar, Niger and Mali, inoculated with the reference strain of
Bradyrhizobium sp. Vigna CB756 in hydroponic culture under P deficiency and sufficiency (30 and 75
µmol KH2PO4 plant-1 week-1, respectively), corresponding respectively to 28 and 70 mg P kg-1 of soil.
Large variability of nodulation and plant biomass was found among cultivars. These two parameters
were generally correlated and the slope of the plant biomass regression as a function of nodulation was
considered as an indicator of the efficiency in use of the rhizobial symbiosis. For the two cultivars
most tolerant to P deficiency, V1 and V4 from Madagascar, the increase in use efficiency of the
rhizobial symbiosis under P deficiency was linked with an increase in nodulated root O2 consumption
linked to N2 fixation, and in phytase gene expression observed on the nodule sections by in situ RT-
PCR. As the complexity of P compartments makes it difficult to assess the P bioavailability in the
plant rhizosphere, an isotopic 32P exchange method was carried out in a rhizotron in order to assess the
direct effect of the roots on P mobilization in rhizosphere soil, comparing V1 and V4 with 28 or 70 mg
P kg-1 of soil. Throughout this study, the various rhizospheric mechanisms involved in the P
mobilization for the plant nutrition were assessed by diffusive phosphate ion (Pd), soil acidification by
pH decrease, organic anion complexation inducing a low Al and Fe content, and mineralization of
organic P through phosphatase. The gross amount of diffusive Pi (Pd) was determined as a function of
Cp and time (t) by coupling sorption-desorption experiments with subsequent isotopic dilution kinetics

135
in soil suspensions at steady-state. The Pd vs. (Cp, t) relationships varied significantly between
treatments, indicating that roots modify soil properties and consequently re-distribute diffusive Pi
between the soil solution and constituents. The Pd values were greater for the rhizosphere soil
obtained with V1 after applying 28 mg P kg-1.This could be attributed to a strong re-supplying
capacity of the soil solution in Pd along the exchange time leading to a large P nutrition of voandzou.
It is concluded that genotypic variability exists among voandzou cultivars for internal adaptation to P
deficiency.

1. INTRODUCTION

Phosphorus deficiency remains the main constraint of agricultural production in

various soils particularly the soils rich in Al and Fe sesquioxides. The phosphorus nutrition of
plants is due exclusively to phosphate ions in the soil solution [1]. Phosphate ions in the soil
solution account for 5 to 10% of total P absorbed annually by crops [2, 3]. The remaining
quantities come from P linked to soil components, particularly sesquioxides and clays which
permanently supply the soil solution of phosphate ions [4]. The bioavailability concept
integrates the existence of a zone strongly influenced by plant roots in the soil. This soil-plant
interface called the rhizosphere is defined as the soil volume around the living roots which is
subject to their actions [5]. Thus, plants adapt to P availability through various biological and
physico-chemical mechanisms. The dominant mechanism in soils under long-term cultivation
and fertilization of field crops in temperate regions is the molecular diffusion of phosphate
ions from the solid phase towards the soil solution [6]. Other biochemical processes that also
contribute to the P bioavailability in soil are the excretion of such compounds as proton efflux
and organic acids, and exo-enzymes like phosphatases. These excreted products involve such
reactions as desorption, dissolution of mineral phosphate, cation complexation in interaction
with phosphate ion and mineralization of organic P [5, 7] leading to phosphate ion release.

Many authors developed some devices in order to study closely the rhizospheric
mechanisms contributing to P bioavailability. Development of the rhizobox or rhizotron [8–
10] made it possible to refine the studies of the rhizosphere by using a thin layer of soil in
direct contact with the plant roots, which is considered as the plant rhizosphere. The
rhizosphere is the volume of a thin layer of soil immediately surrounding plant roots in which
strong root activities exist due to different physico-chemical and biological mechanisms
related to nutrient transfer at the soil-solution interface [5, 11, 12].Thus, it distinguishes from
the ‘bulk” soil which is not influenced directly by growing roots. Various methods are used to
assess the P bioavailability of soil. The conventional methods using chemical extractants are
the most used of the current standard analyses because of their rapidity and cost effectiveness.
Nevertheless, these chemical extractants can dissolve the P forms, either available or
unavailable, for plants [4, 13]. The P extracted with anion exchange membrane or resin, most
closely corresponds to the P absorbed by plants until now. Yet it cannot simulate specifically
the sorption-desorption reaction of phosphate ions in soils because of the limited capacity of
the inorganic P sink, the ions on the membrane surface (Cl-, F-, OH-, CO-) and the effect on
the soil-solution pH.

The isotopic tracers are the tools of choice when the system to be analyzed is made up
of multiple compartments and when the transferred quantities are weak compared to the
quantities in the receiver compartment [14]. The method of isotopic exchange kinetics
consists in surveying the behavior of phosphate ions transfer between the solid and soil
solution phases by the marked ions 32P or 33P in the two phase system. The isotopic tracing
allows the flux of phosphate ions to be quantified, which is likely to be transferred from the
solid phase towards the soil solution and which is difficult to quantify by other methods to

136
characterize P bioavailability [15]. The results from the rhizosphere and non-rhizosphere soil
enable the effect of rhizosphere mechanisms in terms of the P bioavailability to be estimated.

Legumes have been considered for a long time as an alternative technology for soil
fertility restoration. In order to study the contribution of rhizosphere mechanisms which
intervene in P bioavailability, our study focused on the legume voandzou (Vigna subterranea)
inoculated with Bradyrhizobium sp. Vigna CB756. Following a previous selection of lines
contrasting in P use efficiency for their nitrogen symbiotic fixation, two contrasting lines were
grown in a rhizotron to compare rhizosphere and non-rhizosphere soils, with the isotopic
technique using 32P. The aim of this study was to assess the contribution of rhizosphere
mechanisms which are involved in the P bioavailability and to test the genotype variability of
voandzou.

2. MATERIALS AND METHODS

2.1. Experimental design

In order to study the voandzou effect on biochemical parameters at the rhizosphere

scale, an experimental design of rhizotron culture was set up under glasshouse conditions.
This device consists of separating a thin layer of soil from roots through use of a porous
membrane whose mesh allows solute exchange between soil and roots while keeping roots
from penetrating into the soil compartment [8–10]. Approximately 200 g of soil was put in a
polyamide bag (45 x 20 cm) with 30 µm of mesh (Nytrel 0.2SPN, Fyltis-U.G.B., Lyon,
France) to obtain a 1 mm thick soil layer to serve as the rhizosphere for the whole plant.
Indeed, previous studies have shown that rhizospheric modification appears within 1 mm
around the root area [16]. The plant roots were spread over the polyamide bag, thereafter
rolled up and inserted into a 50 cm long PVC column. A filter paper ensured the connection
between soil and the soil solution supplied at the basis of the column. Three tubes ensured soil
oxygenation with compressed air at 400 ml-1 min-1. The rhizotron columns were vertically
maintained during the duration of the experiment. This system allowed an easy access to the
collection of voandzou rhizosphere samples by avoiding root deterioration during soil
separation by agitation.

The studied soil was sampled at 0-20 cm depth in Casevieille, France. It is a

fersiallitic, chromic Cambisol according to the FAO-UNESCO classification [17]. This soil
was marked by its high total P content with a low available P content (Table 1).

2.2. Culture conditions

Two contrasting cultivars of voandzou, previously selected among six cultivars for
their SNF ability for the PUE, were chosen for the rhizotron culture. Phosphorus supplies for
each treatment were 30 and 75 µmol KH2PO4 plant-1 week-1 corresponding respectively to 28
and 70 mg P kg-1 of soil. The experiment was carried out under glasshouse conditions with 20
/ 33°C temperature during 16 / 8 h day / night cycle, with additional illumination of 400 µmol
photons m-² when needed, and 70% daily relative humidity.

The seeds were sterilized with 3% calcium hypochlorite for 20 min and rinsed by 5
washings with sterile distilled water. Thereafter, the sterilized seeds were transferred for
germination in an incubator at 28-30°C on humidified wrapped filter paper in a slightly tilted
vat in which humidity was controlled regularly.

137
 After germination, the seeds were inoculated by soaking the roots for 30 min in a
suspension of 100 ml of Bradyrhizobium sp. Vigna CB756 containing 109 bacteria ml-1. The
inoculum was prepared from a rhizobial culture preserved in tubes at 4°C on 120°C sterilized
agar YEM (Yeast Extract Mannitol) medium: 900 ml distilled water; 100 ml of Bergersen
concentrated solution containing 1 g KCl, 0.1 g FeCl3, 0.4 g CaCl2, 4.5 g Na2HPO4.12H2O, 1
g MgSO4.7H2O, 1 g Yeast extract, 10 g mannitol and 15 g agar [18]. Four days after sowing
the plantlets were transferred to a 40 l vat with 20 inoculated plants vat -1, the roots being
carefully passed through the hole of a rubber stopper with cotton wool fixed at the hypocotyl
level, and grown with the following nutrient solution: CaCl2, 1650 µM; MgSO4.7H2O,1000
µM; K2SO4, 700 µM; Fe EDDHA, 8.5 µM as sequestrene Fe; H3BO3, 4 µM; MnSO4.H2O, 6
µM; ZnSO4.7H2O, 1 µM; CuSO4.7H2O, 1 µM; Na2MoO4.7H2O, 0.1 µM. The solution was
changed every 2 weeks. Urea (2 mM) was supplied during the first two weeks. P was supplied
in the form of KH2PO4 with an exponential distribution during successive weeks.

Four weeks after transplanting, the plants were transferred in the columns for the
rhizotron culture. The two P level, 28 and 70 mg P kg-1 of soil, were supplied according to the
exponential distribution of hydroponic culture until the 6th week. Three soil thin layers
without plants were prepared with the same nutritive solution in order to be used as a control
or non-rhizosphere soil (“bulk soil”). The plants were harvested after 12 days of soil contact.
Shoot and root biomass were oven-dried at 60°C. Soil samples were preserved in the
refrigerator at 4°C in order to keep them fresh for the analyses.

TABLE 1. PHYSICO-CHEMICAL PROPERTIES OF THE SOIL

Parameter Value
Clay (%) 48,60
Fine silt (%) 21,80
Coarse silt (%) 17.80
Fine sand (%) 11.60
Coarse sand (%) 3.00
pH H2O 7.2
pH KCl 6.1
βs (µmol OH-1(q soil)-1 (pH unit)-1) 53.73
CaCO3 (g kg-1) 1.48
Organic matter (g kg-1) 53.70
Organic carbon (g kg-1) 31
CEC (cmol (+) kg-1) (Cobaltihexamine) 25
Ca2+ (cmol (+) kg-1) (Cobaltihexamine) 22.30
Na+ (cmol (+) kg-1) (Cobaltihexamine) 0.12
Mg2+ (cmol (+) kg-1) (Cobaltihexamine) 0.98
K+ (cmol (+) kg-1) (Cobaltihexamine) 0.21
P total (mg kg-1) 960
P ass (mg kg-1) (P Dyer) 7.90
P Olsen (mg kg-1) 5.31
P CaCl2 (mg kg-1) 0.06
N total (g kg-1) 2.88
C: N 10.80
Fe (cmol (+) kg-1) (Cobaltihexamine) <0.005
Fe (g kg-1) (Ammonium oxalate) 1,63
Mn (cmol (+) kg-1) (Cobaltihexamine) 0.013
Al (cmol (+) kg-1) (Cobaltihexamine) 0.042
Al (g kg-1) (Ammonium oxalate) 4.58

138
2.3. Soil analysis

Acid phosphatase activity, in particular phosphomonoesterase release by roots and soil

micro-organisms, was measured from para-nitrophenyl-phosphate (p-NPP) hydrolysis in the
soil according to method initially developed by [19] as modified by [20, 21] with aliquots of
fresh soil immersed in the buffer solution at pH 6.5 and incubated at 37°C during 1 h. After
filtration, absorbance was determined spectrophotometrically at 400 nm. Enzyme activities
were expressed as µmol p-NPP g−1 soil h−1. Measurement pH was in 1:5 ratio of soil: water
[22]. Soil solution concentration of PO4 ions was determined using the malachite green
colorimetric method at absorbance 610 nm [23]. Total P content of soil and plant, was
measured by colorimetry following drying, ashing and acidification.

2.4. Diffusive PO4 ions transfer from the soil solid phase to the solution phase

The dynamic of diffusive PO4 ions (Pd) was determined using the isotopic exchange
method by sorption-desorption and subsequently following the isotopic dilution [24–26]. The
experiment was carried out on a fresh soil suspension preserved at 4 °C at equilibrium. Three
series of five amounts of P between 0 and 150 µg P g-1 soil, depending on the treatment, were
added as KH2PO4 to the soil suspension to obtain a final concentration between 0.01 to 5 mg
P l-1 in the soil solution, and to create a range of PO4 ion concentrations in the soil solution
allowing the determination of the diffusive PO4 ions flux. Soil suspensions were shaken on a
shaking table for 40 h to reach a steady state for the subsequent hours during which isotopic
dilution analysis was carried out. To the soil suspension was added 0.1 ml of biocide (toluene)
to avoid microbial activity, and isotopically labeled by introducing 0.1 ml of 32P-labeled PO4
ions with radioactivity (R) between 0.1 – 106 Bq ml-1 at time zero, and then shaken at 4°C.
This tracer was uniformly and instantaneously dispersed. The amount R of carrier-free 32P
ions was negligible compared to total amount of P ions in solution, which was 105 to106 times
higher, and did not disturb the equilibrium state of the soil suspension. After 4, 40 and 400
min, 2.5 ml of soil suspension were sampled with a plastic syringe and immediately filtered
through a 0.2 µm membrane. The remaining radioactivity (r) of the soil suspension was
counted by a liquid scintillation cocktail (Insta-gel Plus Packard, PerkinElmer, Boston, MA)
using a liquid scintillation counter (Packard TR 1100, PerkinElmer, Boston, MA).

The amount of isotopically exchangeable P (E) includes both Pw as the amount of PO4
present in the soil solution, and Pd as the gross amount of diffusive PO4 ions transferred
between solid and liquid phases. Pw (mg kg-1) was calculated as follows:

Pw = (V/M)*Cp = 10*Cp (1)

with V, the volume of distillated water (ml); M, the weight of soil; Cp, solution PO4 ions
concentration. The gross amount of diffusive PO4 ions from the solid phase to the soil solution
(Pd) was determined through the isotopic dilution principle: the isotopic composition ratio
(IC) of PO4 in the soil solution (Pw) is equal to IC in the solid phase (Pd). The amount of
unlabeled soil PO4 newly transferred to the solution can be measured by determining the
amount of unlabelled PO4 ions in solution (Pw) and the isotopic composition (IC) ratio:

rt / Pw = (R – rt) / Pd = R / E (2)

with r, the radioactivity remaining in solution; R, the radioactivity introduced initially, r / Pw

and (R – r) / Pd, the IC ratio of Pw and Pd.

The Pd value can be deduced from Eq. 2, giving:

139
 Pd(t) = {Pw (rt / R)-1} – Pw Pd(t) = Pw {(1 / (rt/R)-1) – 1} (3)

Pd variation as a function of time and Cp was fitted using the kinetic Freundlich equation (Eq.
4)
Pd = νCpw tp with Pd < P mineral (4)

with ν, the Pd value for t = 1 min at 1 mg P l-1; w, the nonlinear increase in Pd values with Cp;
p, the nonlinear increase in Pr values with time.

2.5. Statistics

The cultivar and treatment effects on soil characteristics were analyzed by a two-way
analysis of variance (ANOVA). Estimates of v, w and p kinetic Freundlich parameters were
obtained using the nonlinear procedure: Proc Mixed of Statistical Analysis Software SAS
[27]. PCA analysis was performed on nutrient content of the soil aliquot, kinetic Freundlich,
soil and plant parameters in order to highlight the rhizosphere effect at the voandzou cultivars
under limited on no available P.

3. RESULTS

3.1. Plant biomass and P concentration

Biomass was slightly affected by P supply, in particular for cultivar 1. Shoot biomass
values varied from 1.34 to 1.54 g plant-1 and root biomass from 0.56 to 0.79 g plant-1 (Fig. 1).
Similar P concentrations were observed for root and shoot biomass, around 1600 mg P kg-1,
except for cultivar 4 with 28 mg P kg-1 soil (Fig.1). P supply significantly increased the plant
nodulation (P< 0.001; R² = 0.50) in particular for cultivar 1.

3.2. Solution PO4 ions concentration, pH, and enzymatic activity of soils

P supply led to a significant response of soil solution PO4 ions concentration (Cp) at
two cultivars (P<0.001; R² = 0.79). The Cp at 70 mg P kg-1 of soil was ten times higher
compared to 28 mg P kg-1 of soil at two cultivars (Fig. 2). In addition, a significant difference
was observed between rhizosphere and non rhizosphere soil at the two P supplies. A high
phosphatase activity was observed in “bulk soil” in comparison to rhizosphere soil. More
microbial activity was highlighted under minimal P supply as 28 mg P kg-1 soil, compared to
70 mg P kg-1 soil, particularly for cultivar 1 (Fig. 2). The pH decreases of rhizosphere and
bulk soil were from 7.3 to 7.1 according to P supply (Fig. 2) with the lowest value at 28 mg P
kg-1 soil.

3.3. Nutrient composition of the soil solution

The nutrient composition in the soil solution of dry soil revealed the superiority of low
added P (28 mg P kg-1 soil) in terms of NO3-N, total N, Ca, Mg, Na and Cl (Table 2).
Otherwise, the high P supplied as 70 mg P kg-1 soil was higher in terms of Al, Fe, K, Si and
SO42--S. Cultivar 1 was marked by the lowest values of some parameters, in particular pH,
organic C, total C, NH4-N, total N, Mg, K, Cl and SO42--S. Results from a Principal
Component Analysis of nutrient content in soil solution with plant and soil parameters
summarize these observed trends (Fig. 3).

140
3.4. Dynamic calibration of the diffusive PO4 ions (Pd) by the kinetic Freundlich
equation

The isotopically exchangeable P transfer between the solid and soil solution phases
were closely related to Cp and the isotopic exchange time (Fig. 4). The increase of P supply in
the rhizosphere and bulk soil suspensions increased Cp for the two cultivars. Increases in
diffusive PO4 ion transferred from solid towards soil solution (Pd) phases were observed
when soluble P at different rates was applied with exchange times of 4, 40 and 400 min. Pd
values increased from 4 to 400 min at the same soil solution PO4 ion concentration, while
increasing Cp produced a further increase in Pr after addition of soluble P.

28 70 mg P kg-1 soil

mg P kg-1 of plant Nodule number plant-1 g DW of shoot plant-1 g DW of shoot plant-1 Shoot biomass
2

0
Root V1
biomass V4
1,0 Cultivars

0,5

0,0
V1 number
Nodule V4
60 Cultivars

40

20

0
V1P content
Shoot V4
3000 Cultivars

2000

1000

0
V1P content
Root Cultivars V4
mg P kg -1 of plant

3000

2000

1000

0
V1 Cultivars V4

FIG. 1. Shoot and root biomass, nodule number and P concentrations of shoots and roots. Data
represent the mean of 4 replicates at 45 days after transplanting. Errors bars represent standard
errors.

141
 1,5 Cp To
28 mg P kg-1 of soil

µg P ml-1
1,0 70 mg P kg-1 of soil

0,5

0,0
V1
150 Phosphatase activity V4 Variétés

µg PN g-1 h-1
100

50

0
7,6 V1 V4 Variétés
pH
7,4
pH

7,2
7,0
6,8
6,6
V1 V4
Cultivars

FIG. 2. Biochemical parameters of rhizosphere and bulk soil in terms of pH, phosphatase activity and
Cp under two P supplies. The data represent the mean of 4 replications at 45 days after transplanting.
Error bars represent standard errors.

TABLE 2. NUTRIENT CONCENTRATIONS IN THE SOIL SOLUTION FOR TWO CULTIVARS

AT THREE LEVELS OF APPLIED P (0, 28 AND 70 MG P KG-1 SOIL)

Parameter† Cultivar 1 at mg P kg-1 soil Cultivar 4 at mg P kg-1 soil O mg P kg-1

28 70 28 70 soil
pH 7.6 7.5 7.8 7.7 7.8
Organic C 50.2 49.5 70.8 83.6 60.0
Total C 58.9 59.6 79.7 92.8 70.2
NH4+-N 2.1 1.7 2.5 2.4 2.5
NO3--N (1) 1.9 1.1 1.6 1.3 2.2
Total N 6.2 5.3 6.5 6.2 5.7
Al 0.26 0.63 0.10 0.41 0.37
Ca 29.5 21.9 30.5 21.9 25.6
Fe 0.09 0.19 0.03 0.15 0.14
Mg 1.3 0.9 1.5 1.0 1.4
K 2.8 4.5 3.0 4.6 2.3
Si 0.1 0.4 0.1 0.4 0.1
Na 6.2 4.2 6.0 4.0 4.3
Cl 34.5 22.5 36.2 25.0 18.9
NO3--N (2) 3.3 3.3 3.3 0.05 3.2
SO42--S 4.9 4.4 5.2 6.7 4.4
†Units are mg l-1; NO3--N (1), spectrocolorimetric method; NO3--N (2), ion exchange chromatography method

In order to assess the influence of rhizosphere activities on phosphate ion transfers

between solid and soil solution phases, variations of Pd, Cp and time values according to P
supply levels were fitted through a regression function called the kinetic Freundlich equation:
Pd = νCpwtp (Table 3). The gross amount of Pi that diffuse in 1 min when the Cp value is 1
mg P l-1 , the ν parameter, significantly decreased at the lowest added P (28 mg P kg-1 soil),

142
being 77% lower than at the highest added P (70 mg P kg -1 soil) for all cultivars. A significant
variation was observed for the ν parameter between rhizosphere and bulk soil, marked by a
significant reduction for cultivar 1 with 28 mg P kg-1 soil as P supplied (Table 3). No
significant difference was observed for the other parameters. However, w and p parameter
values at low P supply were higher than the high P supply especially for cultivar 1.

4. DISCUSSION

A genotypic variation between the voandzou cultivars was observed, marked by a

large P mobilization for cultivar 1 under the lowest rate of added P. This is in agreement with
previous results for symbiotic nitrogen fixation optimization, P nutrition and genetic
improvement among legume species through genotypes with effective rhizobial strains [28–
31]. This genotypic variation was highlighted at the rhizosphere scale. It is known that
growing roots absorb phosphate ions and diffusion is the main rhizospheric mechanism of Pi
that contributes to replenish Pi in solution [1]. Thus, the plant available P was assessed by
measuring both soil solution Pi concentration and the dynamics of diffusive Pi between solid
and soil solution phases. Soil solution Pi concentrations were highest for the highest added P
rate for all cultivars. By comparing the biomass and the uptake of P by plants, it seems that
legumes used more soil and fertilizer P for biomass production under a low P availability
compared to a high P availability.

143
FIG. 3. Principal Component Analysis of nutrient content in soil the solution. V1P28, cultivar 1 with
28 mg P kg-1 soil; V4P28, cultivar 4 with 28 mg P kg-1 soil; V1P70, cultivar 1 with 70 mg P kg-1 soil;
V4P70, cultivar 4 with 70 mg P kg-1 soil; BS, shoot biomass; BR, root biomass; P shoot, shoot P
content; P root, root P content.

144
FIG. 4. The gross amount of diffusive phosphate ions (Pd) in relation with the soil solution Pi
concentration (Cp) for the five rates of soluble P and the different periods of isotopic dilution ( 4, 40
and 400 min) for the different treatments in cultivar V1 and cultivar V4 under 28 and 70 mg P kg-1 of
soil . Y-axis (Pd) and X-axis (Cp) are represented in log-log scales. Regression lines represent the
three periods of isotopic dilution, 4, 40 and 400 min. P supply levels for the determination of diffusive
PO4 ions flux: 0, 10, 20, 50, 100 µg P g-1 soil for 70 mg P kg-1 soil; 0, 40, 50, 80, 130 µg P g-1 soil for
28 mg P kg-1 soil ; 0, 60, 70, 100, 150 µg P g-1 soil for 0 mg P kg-1 soil.

Modeling of diffusive PO4 ion dynamics at the soil-solution interface showed that ν
parameter had the lowest value at 28 mg P kg-1 soil, in particular for cultivar 1. The ν
parameter is the PO4 ions likely to supply P through diffusion to the soil solution after 1 min
of isotopic dilution at PO4 ions concentration of 1 mg P l-1. The decrease of the ν parameter
compared to “bulk” soil were 36% for cultivar 1 with 28 mg P kg -1 soil, 11% for cultivar 4
with 28 mg P kg-1 soil and 3% for cultivar 4 with 70 mg P kg-1 soil. These values were in
contrast with those found for Pisum sativum under rhizotron culture, being around 7% lower
than the bulk soil [32]. The comparison of this parameter allows an appreciation of the
rhizosphere effect on soil physico-chemical properties. The decrease of diffusive PO4 ions in
the short-term, the ν parameter, may be explained by P plant nutrition and efficiency of crops

145
[33], which is marked by a large shoot and root biomass observed for cultivar 1 under low P
supply (Fig. 1).

TABLE 3. MEAN KINETIC FREUNDLICH (PD = ΝCPWTP) PARAMETERS DESCRIBING THE

TIME (T, MIN) AND CP DEPENDENCIES ON THE AMOUNT (PD, MG P KG-1) OF PO4 IONS
TRANSFERRED BETWEEN SOIL AND LIQUID PHASES OF RHIZOSPHERE AND NON-
RHIZOSPHERE SOIL SUSPENSIONS

Treatments† ν‡ w‡ p‡
V4 P28 16.5 ± 1.16 0.44 ± 0.03 0.15 ± 0.01
V4 P70 18.0 ± 0.37 0.38 ± 0.01 0.14 ± 0.01
V1 P28 11.9 ± 0.40 0.46 ± 0.01 0.19 ± 0.01
V1 P70 19.1 ± 0.43 0.37 ± 0.01 0.14 ± 0.00
TO 18.6 ± 0.32 0.42 ± 0.01 0.14 ± 0.00
Analysis of variance#
P level (P) 0.038* 0.103ns 0.061ns
Cultivar (C) 0.282ns 0.303ns 0.239ns
(P) x (C) 0.049* 0.663ns 0.147ns
† TO, bulk soil; V1, cultivar 1; V4, cultivar 4; P28, 28 mg P kg-1 soil added; P70, 70 mg P kg-1 soil added
‡ ν, w, p, kinetic Freundlich parameters; ±, standard error of the mean
# *, P<0.05; ns, not significant

The Pd(Cp, t) values were calculated for each 1-day interval for one week and the
exchange rates (µg P g-1 day-1) were deduced thereafter (Table 4).

TABLE 4. PD(CP, T) VALUES AND THE KINETICS OF THE RATE EXCHANGE IN PD(CP, T)
DURING THE 1ST WEEK OF EXCHANGE IN RHIZOSPHERE AND “BULK” SOIL

Line P level Pd (µg P g-1) Exchange time (day)

(mg P kg-1 soil) Rate (µg P g-1 d-1) 1 2 3 4 5 6 7
V4 28 Pd (0.052. t) 13.37 14.82 15.73 16.42 16.97 17.43 17.84
V4 28 Rate 13.37 1.44 0.92 0.68 0.55 0.46 0.40
V4 70 Pd (0.73. t) 46.93 52.00 55.21 57.62 59.55 61.18 62.59
V4 70 Rate 46.93 5.07 3.22 2.40 1.93 1.63 1.41
V1 28 Pd (0.08. t) 14.68 16.75 18.09 19.11 19.94 20.64 21.25
V1 28 Rate 14.68 2.07 1.34 1.02 0.83 0.70 0.61
V1 70 Pd (0.77. t) 47.99 52.88 55.97 58.27 60.12 61.68 63.02
V1 70 Rate 47.99 4.89 3.09 2.30 1.85 1.55 1.35
T0 0 Pd (0.02. t) 9.44 10.37 10.96 11.39 11.74 12.04 12.29
T0 0 Rate 9.44 0.93 0.59 0.44 0.35 0.29 0.26

In order to analyze the potential release of Pi from the solid phase to the soil solution,
the variation effect of Cp change and immediate exchange capacity of PO4 ions were assessed
from diffusive PO4 ions values from the solid phase towards the soil solution (Pd) through a
regression function with the kinetic Freundlich equation during one week. The comparison of
the Pr daily increase rate (µg P g-1 day-1) during one week deduced from Table 4 allowed
close assessment of the effect of root activity on Pr while separating the effect of soluble P
added and uptake P on Cp. For all treatments, Pd values increased with time during one week
of exchange, while the Pr increase rate per day decreased with the exchange time. Compared
to the first day of exchange, the Pr increase rate per day decreased by 3% for all treatments
after one week of exchange except for cultivar 1 with 28 mg P kg-1 of soil (by 4%).

Similar results were reported during an experiment with maize and pea rhizosphere
soil [32]. These results allowed the highest potential release of Pi in the soil solution to be

146
deduced for cultivar 1 with 28 mg P kg-1 soil, according to the exchange time. This may be
explained by the rhizosphere mechanism suggesting that cultivar 1 was able to optimize the P
assimilation in order to maintain a significant biomass level under the lowest rate of added P.
Otherwise, the slight reduction of the ν parameter and the Pr variation could not only be
explained by PO4 ion diffusion but could be attributed to other mechanisms.

The Principal Component Analysis of microelements in rhizosphere and “bulk” soil

solution (Fig. 3) with plant measurements showed the grouping of a low P supply (28 mg P
kg-1 soil) and the bulk soil with biological indicators of rhizosphere effects, in particular NO3--
N (2), acid phosphatase, shoot and root P concentration, Ca, Na, Cl and Mg, whereas the
highest rate of added P (70 mg P kg-1 soil) was associated with soil indicators including those
related to the soil solid phase such as soil solution Pi concentration Cp, Al, Fe, Si, K, pH, v
parameter, nodulation and shoot and root biomass.

Indeed, plants adapt to their environments according to nutrient availability. With a

low P level in the soil solution, some plant roots could develop a competitive capacity with
the soil solid phase, in particular, by P assimilation at a very low PO4 ion concentration for
natural plant development [33]. Under 28 mg P kg-1 soil, plant P concentration and potential
release of Pi into the soil solution and shoot biomass were greater for cultivar 1. Plants can
mobilize a contrasting P level in biomass in particular for cultivar 1 (Fig. 1). This may be
because rhizosphere acidification occurred and might increase the bioavailability of P sorbed
onto metal oxides such as goethite. The release of H+ in the rhizosphere was related to
differential uptake of cations and anions by plant roots such as Na+, Ca2+, Mg2+ and to the root
excretion of organic acid [33–35]. The legume plants adapt to environmental constraints, such
as low PO4 ions concentration, by the more important root excretions modifying the different
biochemical properties of soil [36].

The increase of the organic anion exudation was shown by many authors with
different species in response to P deficiency. These organic anions such as malate and citrate
can be complexed with solid phase Al or Fe thereby reducing the availability of soluble Al
while increasing the P bioavailability [33, 37–39]. This explains the low Al and Fe values
observed at the lower P supply in comparison with the higher P supply.

As acid phosphatase was reported to indicate the organic P mineralization potential

and biological activity of soils to enhance P availability, the excretion of acid phosphatase
could also explain the plant adaptation. Under limited P availability, the increase of acid
phosphatase excreted by roots was observed, in particular at the control and the low P supply
(Fig. 2) [40]. Indeed, the level of soil orthophosphate ions could inhibit the
phosphomonoesterase activities such as organic P and phytate hydrolysis. P supply of 10
µmol orthophosphate g-1 soil inhibited from 21 to 42% of acid phosphate activity and from 17
to 51% of alkaline phosphatase activity [41, 42], while a decreased P supply from 10 to 1
µmol of PO43- g-1 of soil gave increased phosphatase activity resulting from kinetic
competitiveness of orthophosphate substrate on enzymatic hydrolysis [41, 42]. A high
enzymatic activity, important for organic P mineralization, in particular soil phytate, was a
part of the adaptation mechanism of plants in P mobilization in rhizosphere soil, specifically
for cultivar 1 with low P supply.

147
5. CONCLUSION

Plants differ from one cultivar to another with respect to their ability to mobilize, to
use and to promote more nutrient availability than others. Our result indicated that a slight
modification on the Pi exchangeability from solid to soil solution phases was observed at the
lowest added P of 28 mg P kg-1 soil, particularly for cultivar 1, which probably occurred by
local root-induced changes in pH and root secretion. High P efficiency was shown at the
lowest added P level by changing root physiological and biochemical mechanisms and taking
up greater amounts of P from the soil. Cultivar 1 was able to use efficiently more P under low
PO4 ions concentration in order to produce enough biomass compared with cultivar 4. The
rhizosphere effect, mainly for cultivar, influenced the diffusive PO4 ions dynamic between
solid and soil solution phases under limited P availability. The different rhizospheric
mechanisms contributing to the phosphate ions mobilization for the plant PO4 nutrition were
diffusion, acidification, complexation and mineralization. Otherwise, the potential release of
Pi between solid and soil solution phases can be assessed by this quantitative approach
coupling a sorption-desorption experiments with subsequent isotopic dilution kinetics.

This study showed that genotypic variation existed in the P use of voandzou, a
difference which is particularly highlighted under limited P availability.

REFERENCES

[1] BARBER, S.A., Soil Nutrient Bioavailability: A Mechanistic Approach, John Wiley,
New York, USA (1995).
[2] AIGNER, M., FARDEAU, J.C., ZAPATA, F., Does the Pi strip method allow
assessment of the available soil P? Comparison against the reference isotope method,
Nutr. Cycl. Agroecosyst. 63 (2002) 49–58.
[3] GRANT, C., et al., Soil and fertilizer phosphorus: Effects on plant P supply and
mycorrhizal development, Can. J. Plant Sci. 85 (2005) 3–14.
[4] FROSSARD, E., et al., Processes governing phosphorus availability in temperate soils,
J. Environ. Qual. 29 (2000) 15–23.
[5] HINSINGER, P., et al., Spéciation et biodisponibilité du phosphore dans la
rhizosphère, Océanis 33 (2007) 37–50.
[6] MOREL, C., Caractérisation de la phytodisponibilité du phosphore du sol par la
modélisation du transfert des ions phosphates entre le sol et la solution. Considérations
théoriques et analyses du cycle de P et du comportement des plantes dans les parcelles
de grandes cultures, Institut National Polytechnique de Lorraine (2002) 80 p.
[7] MOREL, C., Mobilités et biodisponibilité du phosphore dans les sols cultivés:
mécanismes, modélisation et diagnostic, Océanis 33 (2007) 51–74.
[8] HINSINGER, P., GILKES, R.J., Root-induced dissolution of phosphate rock in the
rhizosphere of lupines grown in alkaline soil, Aust. J. Soil Res. 33 (1995) 477–489.
[9] KUCHENBUCH, R., JUNGK, A., A method for determining concentration profiles at
the soil-root interface by thin slicing rhizospheric soil, Plant Soil 68 (1982) 391–394.
[10] YOUSSEF, R. A., CHINO, M., Development of a new rhizobox system to study the
nutrient status in the rhizosphere, Soil Sci. Plant Nutr. 34 (1988) 461–465.
[11] JONES, D.L., HINSINGER, P., The rhizosphere: complex by design, Plant Soil 312
(2008) 1–6.
[12] MA, C.H., MANUEL, C.P., “Fertility management of the soil-rhizosphere system for
efficient fertilizer use in vegetable production”, Int. Workshop on Sustained
Management of the Soil-Rhizosphere System for Efficient Crop Production and
Fertilizer Use, Bangkok, Thailand (2006) 1–14.

148
[13] FARDEAU, J.C., JAPPE, J., “Valeurs caractéristiques des cinétiques de dilution
isotopique des ions phosphate dans les systèmes sol-solution”, Phosphore et Potassium
dans les Relations Sol-plante: Conséquence sur la Fertilisation, (GACHON, L., Ed.),
INRA, Paris (1988) 78–99.
[14] FARDEAU, J.C., STENGEL, P., “Sol et échange avec le couvert végétal”, Sol:
Interface Fragile, (STENGEL, P., GÉLIN, S., Eds), INRA Editions, Paris (1998) 41–
63.
[15] RABEHARISOA, L.R., Gestion de la fertilité et de la fertilisation phosphatée des sols
férralitiques des hautes terres de Madagascar, Faculté des Sciences, Département de
Biologie et Ecologie Végétales. Université d'Antananarivo, Antananarivo (2004) 199 p.
[16] GAHOONIA, T.S., NIELSEN, N.E., A method to study rhizosphere processes in thin
soil layers of different proximity to roots, Plant Soil 135 (1991) 143–146.
[17] FAO-UNESCO, Soil Map of the World, Revised Legend, International Soil Reference
and Information Centre, Wageningen, The Netherlands (1989) 138 p.
[18] VINCENT, J.M., A Manual for the Practical Study of Root-nodule Bacteria, Blackwell
Scientific Publications, Oxford (1970) 164 p.
[19] TABATABAI, M.A., BREMNER, J.M., Use of p-nitrophenol phosphate in assay of
soil phosphatase activity, Soil Biol. Biochem. 1 (1969) 301–307.
[20] EIVAZI, F., TABATABAI, M.A., Phosphatases in soils, Soil Biol. Biochem. 9 (1977)
167–172.
[21] ALEF, K., et al., “Phosphatase activity”, Methods in Applied Soil Microbiology and
Biochemistry, (ALEF, K., NANNIPIERI, P., Eds), Academic Press, London (1995)
335–336.
[22] AFNOR., Qualité des Sols. Vol 1: Recueil de Normes, Association Française de
NORmalisation, Paris (1999).
[23] VAN VELDHOVEN, P.P., MANNAERTS, G.P., Inorganic and organic phosphate
measurements in the nanomolar range, Anal. Biochem. 161 (1987) 45–48.
[24] FARDEAU, J.C., Le phosphore assimilable des sols: sa représentation par un modèle
fonctionnel à plusieurs compartiments, Agronomie 13 (1993) 317–331.
[25] FARDEAU, J.C., Dynamics of phosphate in soils. An isotopic outlook, Fertil. Res. 45
(1996) 91–100.
[26] FROSSARD, E., SINAJ, S., The isotope exchange kinetic technique: a method to
describe the availability of inorganic nutrients. Applications to K, P, S and Zn, Isot.
Environ. Health Stud. 33 (1997) 61–77.
[27] SAS INSTITUTE, SAS ⁄ STAT software. Release 8.2, SAS Institute, Cary, NC (2001).
[28] KISHINEVSKY, B.D., et al., Variation in nitrogen fixation and yield in landraces of
voandzou (Vigna subterranea), Field Crops Res. 48 (1996) 57–64.
[29] KUMAGA, E.D., DANSO, S.K.A., ZAPATA, F., Time-course of nitrogen fixation in
two bambara groundnut (Vigna subterranea L. Verdc.) cultivars, Biol. Fertil. Soils 18
(1994) 231–236.
[30] MASSAWE, F.J., et al., Breeding in Bambara groundnut (Vigna subterranea L.
Verdc.): strategic considerations, Afr. J. Biotechnol. 4 (2005) 463–471.
[31] DREVON, J.J., et al., “Do nodule phosphatase and phytase link with the phosphorus
use efficiency for N2-dependent growth in Phaseolus vulgaris?”, Proc. 15th Int.
Nitrogen Fixation Congr. and 12th Int. Conf. Afr. Assoc. for Biological Nitrogen
Fixation, Vol. 42, (DAKORA, F.D.C., et al., Eds), Springer (2008) 93–94.
[32] MOREL, C., HINSINGER, P., Root-induced modifications of the exchange of
phosphate ion between soil solution and soil solid phase, Plant Soil 211 (1999) 103–
110.
[33] HINSINGER, P., Bio-availability of soil inorganic P in the rhizosphere as affected by
root-induced chemical changes: a review, Plant Soil 237 (2001) 173–195.

149
[34] JAILLARD, B., Flux de protons dans la rhizosphère et acidification des sols, Colloque
Acidification des Sols, INRA, Versaille (2001).
[35] TANG, C., et al., Proton release of two genotypes of bean (Phaseolus vulgaris L.) as
affected by N nutrition and P deficiency: New challenges for rhizosphere research at
the entrance of the 21st Century, Plant Soil 260 (2004) 59–68.
[36] WRAGE, N., CHAPUIS-LARDY, L., ISSELSTEIN, J., Phosphorus, plant
Biodiversity and climate change, Sociology, Organic Farming, Climate Change and
Soil Science, Sustain. Agric. Rev. 3 (2010) 147–169.
[37] GUPPY, C.N., et al., Competitive sorption reactions between phosphorus and organic
matter in soil: a review, Aust. J. Soil Res. 43 (2005) 189–202.
[38] LAMBERS, H., et al., Plant-microbe-soil interactions in the rhizosphere: an
evolutionary perspective, Plant Soil 321 (2009) 83–115.
[39] ZHOU, L.L., et al., Rhizosphere acidification of faba bean, soybean and maize, Sci.
Total Environ. 407 (2009) 4356–4362.
[40] LI, M.G., et al., Secretion of phytase from the roots of several plant species under
phosphorus conditions, Plant Soil 195 (1997) 161–169.
[41] JUMA, N.G., TABATABAI, M.A., Distribution of phosphomonoesterase in soils, Soil
Sci. 126 (1978) 101–108.
[42] VATS, P., BANERJEE, U.C., Production studies and catalytic properties of phytases
(myo-inositolhexakisphosphate phosphohydrolases): an overview, Enzy. Microb.
Technol. 35 (2004) 3–14.

150
 EVALUATION OF SOYBEAN AND COWPEA GENOTYPES FOR
PHOSPHORUS USE EFFICIENCY

F.K. KUMAGA, K. OFORI, S.K. ADIKU, Y.O. KUGBLENU, W. ASANTE, H.

SEIDU
College of Agriculture and Consumer Sciences,
University of Ghana,
Legon, Accra,
Ghana

J.J. ADU-GYAMFI
Soil and Water Management & Crop Nutrition Laboratory,
International Atomic Energy Agency,
Vienna,
Austria

Abstract

Initial screening of one hundred and fifty-two (152) and fifty (50) genotypes of soybean and
cowpea, respectively, were conducted at the early growth stage to evaluate root traits associated with
phosphorus (P) efficiency. Fifty soybean genotypes were subsequently selected and evaluated on a
tropical low P soil (Lixisol) for growth and yield under low and adequate P availability. Plants were
sampled at twelve and thirty days after sowing and at maturity. Six cowpea genotypes were also
selected and evaluated in pots filled with Alfisol under low, moderate and high P availability. Plants
were sampled at forty days and assessed for shoot yield and nodulation under low P availability. Using
Principal Component Analysis (PCA), Phosphorus Efficiency Index (PEI) was used to determine P
efficiency of soybean and cowpea genotypes. A wide variation in root traits for soybean and cowpea at
the early growth stage was found, and allometric analysis showed a significant correlation between the
root and shoot parameters at this stage. The study provided an opportunity to compare root traits of
newly developed cowpea genotypes (early maturing, medium maturing, dual purpose and Striga
resistant lines) with older released cultivars. There were significant differences in root length among
the groups. In general, dual purpose, Striga resistant and medium/early maturing genotypes showed
the longest roots while the older varieties showed the least total root length. Field and pot results also
showed differential growth of soybean and cowpea with low P availability. Further, PCA of the results
indicated that soybean genotypes could be grouped into three distinct P efficiency categories.
Retaining the PC and the relative weight for each genotype in combination with yield potential under
high P, four categories of responsiveness to P were obtained. Cowpea genotypes were grouped into
three P efficiency categories and two categories of responsiveness to P. The study also found genetic
differences in nodulation under conditions of P stress. There were large genotypic variations for P
uptake under high P levels but not under low P levels. The study showed that there was significant
genotypic variation for root traits during early growth and genotypic differences for soybean and
cowpea growth under low P.

1. INTRODUCTION

Tropical soils exposed to long periods of weathering have low organic matter, low
cation exchange capacity and overall low inherent fertility. At present, the main nutrient
limiting factors in sub-Saharan African soils are N and P. While N can be introduced to the
soil through various organic inputs, there is no comparable process to N2 fixation for the
introduction of P into farming systems [1].

Soybean and cowpea are important food legumes because they are relatively cheap
sources of protein in most developing countries. Soybean contains 30-50% protein while

151
cowpea contains between 21-23% protein [2]. Additionally, they fix atmospheric N2, a
fundamental process in soil nutrient management practices. In soybean it is estimated that
25–85% of its total shoot N content can come from fixation [3]. Despite their importance,
yields of both legumes are extremely low. For example, cowpea grain yields on average
farmers’ fields are only half compared to yields obtained on researcher-managed on-farm
fields [4]. Low soil availability of P accounts for low productivity in legumes [5].

Low P status in Ghanaian soils has been extensively documented [6–10]. Two main
reasons may explain this occurrence, firstly P concentrations in the soil solution are generally
low (<0.05 µg ml-1) compared to nitrate-N concentrations (100 µg ml-1) [11]. Anne and Lal
[12] reported that soil available P in soils of West African is well below the critical level of
10.8 mg kg-1 required for grain legume production. Secondly, soils, especially those with low
pH, have a high amount of soluble Fe and Al which react with P and render it insoluble and
unavailable for plant use [13]. This is further compounded by the fact that, unlike N which
comes to plants by mass flow (i.e., soil water moves towards the roots as the plant loses water
through transpiration), P is absorbed mainly by diffusion through gradients created by root
uptake, which means that plant roots have to grow to come into contact with new soil from
which they can extract P. Root characteristics are therefore an important factor in P
acquisition. A larger root system provides greater root-soil contact and thereby higher uptake
of soluble P especially under low P availability [11].

The application of P amendments to these soils can greatly improve production levels,
but non-availability and high costs limit their use practically by a majority of smallholder
farmers. Reports indicate that a great percentage of Ghanaian farmers do not use inorganic
fertilizers to increase the soil nutrient availability [14, 15].

In crops, roots are the vital structure for water and nutrient uptake, but data on root
traits are rarely included in most screening and final breeding programs. Several researchers
have shown that root traits are important in enhancing the root system and consequently
nutrient uptake. For instance, Vieira et al. [16] reported that root hairs are important in P
uptake, and lateral roots have been reported to enhance top soil foraging of P [17]. A better
understanding of root physiological traits that improve P acquisition especially under low P
stress would facilitate selection of more P efficient crop genotypes which would be important
in P deficient soils [18]. It would therefore be desirable to discover crop genotypes which can
access a greater proportion of the total soil nutrient that may otherwise be unavailable to them
[19].

The objectives of this study were therefore to:

a) Characterize root and other traits of cowpea and soybean which contribute to efficient
P uptake in low P soils using established analytical protocols at the seedling stage.
b) Evaluate the performance of genotypes with different root traits under limiting P
conditions.
c) Evaluate whether genotypes with different root traits result in increased biomass
production and grain yield under conditions of varying P availability

2. MATERIALS AND METHODS

In a study conducted at the Crop Science department, University of Ghana (UG),

Legon, one hundred and fifty-two (152) soybean genotypes were obtained from the
International Institute of Tropical Agriculture (IITA), Ibadan, Nigeria, and the Crop Research
Institute (CRI), Fumesua, Ghana. Fifty cowpea genotypes were also obtained from IITA and

152
the Crop Science Department, University of Ghana (UG), Legon. The 50 cowpea genotypes
included 18 early maturing, 6 medium maturing, 10 Striga resistant and 6 dual purpose
genotypes obtained from IITA, and 10 from the Crop Science Department, UG. The latter
from UG were improved varieties which have been released and grown extensively in most of
the agro-ecological zones of Ghana. Seeds of each genotype were surface sterilized in 0.5%
NaOCl for one minute rinsed and sown into PVC tubes of length 15.0 cm and 3.4 cm in
diameter. The tubes contained a mixture of sawdust and rice husk in the ratio of 1:1 by
volume. The tubes were placed in aluminium troughs containing 0.5mM CaSO4 solution.
Each genotype was replicated six times. The set-up was placed in a screen house and watered
daily with tap water. Five days after planting (DAP) the seedlings were removed from the
medium and shoots cut from roots. The roots were washed free of the substrate and growth
parameters including primary root length, number of lateral roots on the primary root and total
root length were evaluated. Root length was determined using a modified line-intersect
method [20]. Counts were made on the intercept of roots with 2 cm vertical and horizontal
grid lines with the aid of a tally counter. Counts were converted to length measurements using
the formula:

Root length(R) = number of intercepts × length conversion factor

where length conversion factor for the 2 cm grid square was 1.5714.

Data were also collected for shoot growth parameters and subjected to Analysis of
Variance (ANOVA) with the differences between means compared using the least significant
difference (LSD). Principal Component Analysis was performed to evaluate inter- and intra-
specific differences among genotypes.

Fifty soybean genotypes were selected and used in a field experiment conducted at the
UG farm, Legon. The soil at the experimental site was classified as Adenta series under the
Ghanaian classification, and Vertic Lixisol under the FAO classification. For characterization,
undisturbed soil samples were taken from 0-10 cm and 10-20 cm of the profile on April 30,
2008, and mixed to form a composite sample (about 5 kg). The soil was analysed for pH in
water (electrometric), total N [21], organic carbon [22] and available P (Bray I), K, Ca and
Mg were determined by extraction with NH4OAC. Seeds of all the 50 genotypes were sown
on May 8, 2008 in double rows one metre long. There were two levels of phosphorus, 0 and
30 kg P ha-1. Triple super phosphate was applied in furrows along the individual rows of
soybean, 5 DAP. At the same time, potassium was applied as muriate of potash at the rate of
30 kg ha-1 to all soybean rows. The combination of soybean genotypes and P treatments was
replicated three times. Twelve and thirty DAP, four plants from each genotype were harvested
for the determination of number of roots, root dry weight and shoot dry weight. Pod number
per plant, 100 seed weight and seed yield were measured on 10 plants at maturity, when pods
had dried and turned brown. Pods were removed from individual plants and counted. Pods
were shelled by hand and the weight of seed determined at 12% moisture.

Six cowpea selected genotypes (IT03K-351-1, IT00K-901-5, IT93K-452-1, IT98K-

1263, IT97K-819-118, Asontem) were evaluated in a pot experiment. The soil used for the
study was Toje series (Ghanaian classification), or Alfisol (FAO classification). A sample of
the soil was obtained from 0-15 cm depth from an uncultivated area at the UG, Legon, and
characterized for its physical and chemical properties. Six litre capacity pots were filled with
5 kg P deficient soil, which was previously sieved with a 2 mm sieve. Three P treatments, as
Triple Super Phosphate, were administered; P at 40 kg ha-1 (P40), P at 20 kg ha-1(P20) and no P
(P0). The experimental design was a factorial randomized complete block with four

153
replications. Because of the laborious nature of root harvest and analysis, the experiment was
divided into two plantings. A basal N application of ammonium sulphate at a rate of 50 kg−N
ha-1 was applied before planting. Hoagland’s plant nutrient solution, not containing P, was
prepared by dissolving 0.4925 g CaCl2.2H2O, 1.875 g K2SO4, 3.075 g MgSO4.7H2O, 0.8043 g
Fe (III) Na EDTA, 0.0155 g H3BO3, 0.009 g MnCl2.4H2O, 0.0008 g CuSO4.5H2O and 0.0002
g H2MoO4.H2O in 1 litre of distilled water. The solution was adjusted to pH 5.5-6.0 with
0.1N NaOH and 20 ml applied to each pot 8 DAP. This was repeated weekly together with
application of tap water three times a week.

Plants were harvested by cutting the shoot at the soil surface. Leaves were separated,
counted and leaf area measured using an AM 100 leaf area meter (Delta-T Devices Burwell,
Cambridge). Roots were retrieved by washing under running water. Nodules were counted
dried and weighed. All plant parts were dried at a temperature of 65˚C for 48 h and milled for
determination of N and P.

Phosphorus efficiency of soybean genotypes was determined by phosphorus efficiency

index (PEI) [23], and assessed using principal component analysis (PCA) of standardized
values of plant biomass and seed yield factors at low P, and relative values at low P to those
obtained under high P supply. In cowpea, phosphorus efficiency was also assessed for
nodulation under limiting P.

3. RESULTS

3.1. Soils used to evaluate soybean and cowpea genotypes for phosphorus efficiency

Characterization of the soil used for the soybean and cowpea genotypes is shown in
Table 1. In common with most soils found in Ghana, the level of available P the soils used
for the study were low. Similarly the total N was also low consistent with the relatively low
OM status of the soil.

TABLE 1. PHYSICAL AND CHEMICAL CHARACTERISTICS OF SOILS

Parameter Classification and properties

FAO Classification Lixisol Alfisol
Ghanaian Classification Adenta Series Toje Series
Sand : Silt: Clay (%) 64: 6: 30 70: 4: 26
Bulk density (Mg m-3) 1.16 1.22
pH (KCl: H2O) 5.3: 5.5 5.0
Organic C (g kg-1) 5.8 6.5
N (%) 0.08 0.08
Total P (mg kg-1) 95.5 201
Available P (mg kg-1) 5.8 7
Al Saturation (%) 18.5 16.0
Fe Saturation (%) 6.5 6
CEC (cmol kg-1) 2.9 2.8

154
3.2. Shoot and root characteristics of soybean and cowpea at early growth stage

There were significant differences among soybean and cowpea genotypes for the
following parameters measured at the early growth stage; total root length, root number, root
weight, shoot weight and root: shoot ratio within each group of soybean and cowpea
genotypes (Table 2). The number of lateral roots in soybean genotypes varied from 4 to 50
with a mean of 30 while in cowpea it varied from 9 to 66. Tap root length varied between 5.1
and 18 cm. The average tap root length of cowpea genotypes was 13.1 cm. The lateral root
density was obtained by dividing the tap root length by the number of lateral roots counted.
There were no significant differences found among the genotypes in lateral root density.

There were significant differences in root biomass among soybean genotypes.

However, cowpea showed no significant difference. Even though a high coefficient of
variation was observed in root parameters of cowpea genotypes compared to soybean
genotypes, cowpea genotypes obtained the highest number of lateral roots and total root
length. This indicates that cowpea had more extensive root growth compared to soybean.
Principal component analysis was conducted on soybean and cowpea genotypes using 5 plant
parameters measured. The first two PCs accounted for 83% of the variability observed (Table
3). PC1 represented 60% whiles PC2 contributed 23% of the observed variability.

TABLE 2. ROOT AND SHOOT CHARACTERISTICS OF SOYBEAN AND COWPEA

GENOTYPE 5 DAP

Trait Soybean Cowpea

Mean Range CV (%) Mean Range CV (%)
Tap root length (cm) - - - 13.1** 5.1-18.0 22.7
Number of lateral roots 30*** 4–50 22.3 41** 9–66 36.0
Total root length (cm) 57.1*** 24.9–103.1 24.9 103.6* 6.2–258.6 56.2
Lateral root length (cm) – – – 100.3* 2.0–240.5 56.4
Lateral root density – – – 3.3ns 1.3–5.3 40.2
(number of roots/cm)
Root fresh weight (g) 0.3*** 0.1–0.5 24.0 0.30ns 0.10– 0.50 56.1
Shoot fresh weight (g) 0.6*** 0.3–1.2 15.0 0.67** 0.2– 1.2 18.6
Root: shoot ratio 0.46*** 0.15–0.81 15.4 0.45ns 0.1–1.3 43.5
Total biomass (g) 0.92*** 0.33–1.64 15.2 0.96*** 0.52–1.36 23.9
*; P<0.05, **; P< 0.01, ***; P< 0.001

TABLE 3. EIGEN VALUES AND PROPORTIONS OF TOTAL VARIATION AMONG SOYBEAN

GENOTYPES (N = 152) AND COWPEA GENOTYPES (N = 50) AS EXPLAINED BY THE FIRST
TWO PRINCIPAL COMPONENTS

Parameter F1 F2
Eigenvalue 3.62 1.37
Variability (%) 60.37 22.86
Cumulative % 60.37 83.23

There was wider variation observed in cowpea compared with soybean genotypes
(Fig. 1) even though 50 genotypes in cowpea were studied compared with 152 in soybean.
86% of the cowpea genotypes had lateral root numbers higher than 30 which was the average
for soybean genotypes, while only 2% of soybean genotypes had lateral root number higher
than 41 which was the mean among cowpea genotypes.

155
 4.000

3.000 a
2.000

1.000

0.000
-8.000 -6.000 -4.000 -2.000 0.000 2.000 4.000 6.000 8.000
-1.000

-2.000

-3.000

-4.000

4.000

3.000 b
2.000

1.000

0.000
-8.000 -6.000 -4.000 -2.000 0.000 2.000 4.000 6.000 8.000
-1.000

-2.000

-3.000

-4.000

FIG. 1. Genotypic variation in roots and shoot parameters of soybean (a) and cowpea (b) as given by
the first two principal components.

To eliminate the effect of variation in plant size the data were log transformed to show
the allometric relationship between parameters more or less independent of the size of the
plants (Fig. 2 and Fig. 3). There was strong allometric relationship between shoot weight and
total root length and root weight in soybean and cowpea genotypes. Results show a similar
allometric partitioning coefficient among soybean and cowpea genotypes.

The study also provided an opportunity to compare newly developed genotypes of

cowpea [early maturing (EM), medium maturing (MM), Dual purpose (DP) and Striga
resistance (SR) genotypes] with older genotypes released previously (ER). There were
significant differences in root length among the genotypes. In general, EM genotypes had
greatest tap root length followed by the SR and the DP genotypes (Table 4). The ER
genotypes had the least taproot length.

156
 log shoot weight v total root length of soybean genotypes
0.3
log shoot 0.25 y = 0.1652x - 0.0771
R² = 0.3632
0.2
0.15
0.1
0.80 1.00 1.20 1.40 1.60 1.80 2.00
log total root lentgh

log shoot weight v log total root length of cowpea genotypes

log shoot

0.3
0.25 y = 0.1083x + 0.004
R² = 0.3345
0.2
0.15
0.1
0.05
0.80 1.30 1.80 2.30
log total root length

FIG. 2. Allometric relationship between total root length and shoot biomass (log scale base 10) of
soybean and cowpea genotypes.

0.35 log shoot weight v root weight of soybean genotypes

0.3
y = 1.2188x + 0.0793
log shoot weight

0.25 R² = 0.5078
0.2
0.15
0.1
0.05
0.05 0.07 0.09 0.11 0.13 0.15 0.17
log root weight

log shoot weight v root weight of cowpea genotypes

0.35
log shoot weight

0.3
0.25 y = 0.9957x + 0.107
0.2 R² = 0.4717
0.15
0.1
0.05
0.00 0.05 0.10 0.15 0.20
log root weight
FIG. 3. Allometric relationship between root biomass and shoot biomass (log scale) of soybean and
cowpea genotypes.

The EM genotypes, which had the greatest tap root length, also had the highest root
dry weight while DP genotypes had the least root dry weight (Table 4). The root dry weight

157
was however not significantly different among the genotypes and within genotypes only EM
genotypes showed significant differences. The DP genotypes had the greatest number of
lateral roots, followed by the SR genotypes while the ER genotypes had the least number of
lateral roots. Within genotypes there were no significant differences in lateral root number
observed. The DP and SR genotypes of cowpea, which had the greatest number of lateral
roots, also showed the highest lateral root densities of 3.8 and 3.5, respectively but differences
were not significant. However, there were significant differences in lateral root density within
ER genotypes.

TABLE 4. MEANS OF ROOT TRAITS OF DUAL PURPOSE, EARLY MATURING, LOCALLY

RELEASED, MEDIUM MATURING AND STRIGA RESISTANT GENOTYPES OF COWPEA

Cowpea group† Number of Root length (cm) Lateral root density Root weight
lateral roots Tap Total Lateral (roots cm-1 taproot) Fresh (g) Dry (mg)
Dual purpose 50a 13.6a 141.5a 127.9a 3.8a 0.34a 17.6a
Early maturing 44b 13.8a 122.9a 109.1a 3.2a 0.33a 22.6a
Local released 30c 11.5a 66.3b 55.3b 3.0a 0.20c 19.2a
Medium maturing 34c 12.1a 85.0b 72.9b 3.4a 0.27b 22.4a
Striga resistant 47ab 13.6a 131.7a 118.2a 3.5a 0.34a 19.0a
† Number of genotypes within each classification; Dual purpose, n = 6; Early maturing, n = 18; Locally released,
n = 10; Medium maturing, n = 6; Striga resistant, n = 10.
Means within a column followed by the same letter are not significantly different among the groups at P<0.05

The shoot dry weight among genotypes in ascending order was

ER<EM<SR<MM<DP (Table 5). Plant height varied significantly (P<0.01) among and
within cowpea genotypes. DP and ER genotypes showed significant differences within each
genotype. There were however no significant differences within EM, MM and SR genotypes.

TABLE 5. MEANS OF SHOOT TRAITS OF DUAL PURPOSE, EARLY MATURING, LOCALLY

RELEASED, MEDIUM MATURING AND STRIGA RESISTANT GENOTYPES OF COWPEA

Cowpea group† Shoot fresh Shoot dry Plant height Root: shoot Total
weight (g) weight (mg) (cm) ratio biomass (mg)
Dual purpose 0.77a 112.5a 6.2a 0.16b 130.1a
Early maturing 0.70b 98.8a 5.7b 0.25ab 121.4a
Locally released 0.43c 57.2b 5.0c 0.35a 74.1b
Medium maturing 0.73ab 112.0a 6.0ab 0.19ab 134.4a
Striga resistant 0.72b 103.1a 5.9ab 0.19ab 122.1a
† Footnote same as for Table 4

3.3. Phosphorus efficiency among soybean genotypes

3.3.1. Plant growth and yield

P application significantly increased the shoot and root dry weight of soybean
genotypes (Table 6). On average both shoot and root biomass increased by more than 70%
when P was applied. The significant increase in shoot biomass of more than 80% in soybean
genotypes with P application was expected considering that the soil used was extremely low
in available P. Under low P availability significant differences were observed among fifty
soybean genotypes (Table 6). Shoot biomass ranged from 1.13g to 4.67g while root biomass
ranged from 0.14g to 0.62 g. Relative shoot biomass, computed as shoot biomass of low P
plants as % of that of high P plants ranged from 20.3 to 213.2%, while relative root dry
weight ranged from 28 to 193.5%. A significant and positive correlation was observed

158
between the shoot biomass and root biomass under low and high P availability (Fig. 3) that
was independent of differences in plant size. Seed yield also increased by as much as 99%
among the 50 genotypes when P was applied. Seed yield varied significantly among soybean
genotypes under P limiting conditions varying from 1.2 to 9.2 g per plant. Although 100 seed
weight varied significantly among soybean genotypes, P application produced no significant
difference in seed size.

TABLE 6. MEAN RESPONSES OF GROWTH PARAMETERS OF 50 SOYBEAN GENOTYPES

GROWN IN A P-DEFICIENT SOIL UNDER LOW (LP) AND HIGH P (HP).

Growth parameters Means CV Probability

LP Range HP Range (%) P G
Total root length 12 DAP (cm plant-1) 41.53 26.97- 63.64 28.49 13.95-49.56 155 *** ***
Number of lateral roots 12 DAP 26 17- 41 23 13-37 115 *** ***
Root DW 12 DAP (g plant-1) 0.04 0.02-0.07 0.03 0.02-0.05 123 *** ***
Shoot DW 12DAP (g plant-1) 0.18 0.11-0.25 0.18 0.10-0.28 99 ns **
Root: shoot ratio 12 DAP 0.22 0.13-0.48 0.17 0.09-0.31 128 *** ***
Total root length 30 DAP (cm plant-1) 56.14 39.48-83.15 60.1 32.80-97.36 99 * **
Number of lateral roots 30 DAP 21 9-40 20 6-30 109 * ***
Root DW 30 DAP (g plant-1) 0.35 0.14-0.62 0.47 0.23-0.84 86 *** *
Shoot DW 30 DAP (g plant-1) 2.14 1.13-4.67 3.01 1.03-6.25 82 *** *
Root: shoot ratio 30 DAP 0.17 0.11-0.24 0.16 0.10-0.25 6.4 * *
Number of pods plant-1 24 9-42 35 16-70 76 *** *
Seed yield (g plant-1) 3.26 1.23-9.22 5.34 1.31-13.29 83 *** **
100 seed weight 12.32 7.76-19.90 12.63 7.20-17.35 99 ** ns
†P, phosphorus; G, genotype; P x G interactions were not significant; *, P<0.05; **, P<0.01; ***, P<0.001; ns,
not significant

3.3.2. Root traits

Total root length and number of roots was measured at 12 DAP and 30 DAP. At 12
DAP total length and number of roots varied significantly (P<0.001) among soybean
genotypes, and when P was applied (Table 6). In general, P application reduced total root
length of soybean genotypes with mean total root length being 20.7 and 18.3cm plant -1 for P0
and P30, respectively. Under P limiting conditions total root length varied from 9 to 32.3 cm
plant-1 and the number of roots showing the same trend. On average, P application reduced the
number of roots. At 30 DAP total root length and number of roots were significantly (P<0.05)
increased under P limiting conditions.

3.3.3. Phosphorus use efficiency and phosphorus responsiveness

Phosphorus efficiency was assessed using principal component analysis of 6 growth

parameters of the 50 genotypes. The 6 parameters at low P along with 6 indices at low P
relative to high P differed significantly among accessions. Three principal components (PCs)
of each genotype with Eigen values greater than one were retained whose cumulative
contribution was 45%.

The relative weight of each principal component was weighted by the corresponding
contribution rate accounting for variation of all growth traits. All genotypes were grouped into
3 clusters: cluster 1 (PEI <−0.01) as low efficient genotypes, cluster 2 (0.04 < PEI < 0.32) as
moderately efficient genotypes and cluster 3 (PEI > 0.45) as highly efficient genotypes (Table
7). The first cluster consisted of 31genotypes that showed low efficiency. The second cluster
was comprised of 13 moderately P efficient genotypes and third cluster consisted of 4 highly
P efficient genotypes. Using Phosphorus Efficiency Index (PEI) and standardized values of

159
shoot biomass and seed yield at high P, the P efficiency of soybean genotypes was determined
(Fig. 4).

TABLE 7. CATEGORISING 50 SOYBEAN GENOTYPES INTO LOW EFFICIENT (LE),

MODERATE EFFICIENT (ME) AND HIGH EFFICIENT (HE), ACCORDING TO P EFFICIENCY
INDEX (PEI)

ID Cluster PEI† ID Cluster PEI

9 LE 0.03 23 LE –0.24
5 LE 0.25 68 LE –0.12
17 LE 0.04 107 LE –0.27
59 LE 0.23 7 LE –0.04
143 LE 0.27 92 LE –0.11
48 LE 0.19 34 LE –0.21
14 LE 0.06 84 LE –0.20
65 LE 0.26 27 LE –0.15
38 LE 0.19 3 ME 0.11
90 LE 0.20 77 ME 0.23
101 LE 0.28 57 ME 0.04
86 LE 0.10 55 ME 0.16
138 LE 0.46 137 ME 0.29
134 LE 0.20 4 ME 0.04
26 LE 0.10 46 ME 0.26
69 LE 0.05 67 ME 0.07
60 LE 0.05 106 ME 0.32
28 LE 0.01 145 ME 0.24
22 LE 0.12 43 ME 0.04
121 LE 0.03 94 ME 0.15
113 LE 0.35 45 ME 0.31
16 LE 0.04 129 HE 0.75
87 LE 0.02 141 HE 0.45
66 LE 0.20 104 HE 1.03
103 LE 0.04 10 HE 0.60
†All values in the column are negative

Standardized values of both shoot yield and seed yield were used because even though
the main goal of breeding efficient genotypes is to increase yield, this trait however is
influenced by several environmental conditions compared with shoot parameters in the field.
Interestingly, the outcome was different for each trait among genotypes. Genotypes were
grouped into four distinct categories - (i) efficient and responsive (ER); (ii) non-efficient and
responsive (NER); (iii) non-efficient and non-responsive (NENR); and (iv) efficient and non-
responsive (ENR).

160
 4.00
Standardized shoot biomass at high P NE
107
143
3.00 ER
2.00 106
4 46
23 69
113 59 27 2843 145
1.00 92
26
10138 717 129
651346814 10
60 77 9
87 67
138 0.00
66 -1657 55
103 141 104
-1.10 -0.60 3448 22
84
-0.10 45
0.40 0.90
3
5-1.00 121 94 137
90 86
-2.00

-3.00 NR
NENR
-4.00

4.00
17
69
3.00 ER
NER
Standardized seed yield at high P

87
2.00
9027
113 134 2843 145 46
104 106
1.0067 77
107 7
149
103
566 141
101 38 68 0.00574 55
84
1435948 -0.10 45 10
-1.10 -0.60 16 3 0.40 0.90
138 6534 92-
23 22
2660
86
-1.00 137 129
121 94
-2.00
NENR -3.00 NR
-4.00
The scores of P efficiency index (PEI) among soybean

FI

FIG. 4. Classification of soybean genotypes into four distinct responsive groups efficient and
responsive (ER), non-efficient and responsive (NER), non-efficient and non-responsive (NENR) and
efficient and non-responsive (ENR) according to P Efficiency Index (PEI) and standardized values of
shoot biomass under high P. Standardized values of shoot dry weight (Xs) are estimated as the
following function: = ( − )/

3.3.4. Correlation between root traits, plant biomass and yield

There was no significant correlation between root traits, biomass and yield in this
study (Table 8). Shoot dry weight correlated positively with seed yield at 30 DAP under low
P availability. Root dry weight correlated significantly and positively with shoot dry weight at
30 DAP under low P availability.

161
TABLE 8. CORRELATION MATRIX AMONG ROOT TRAITS AND SHOOT BIOMASS: ROOT
NUMBER (RN), TOTAL ROOT LENGTH (TRL), SHOOT DRY WEIGHT (SDW), NUMBER OF
PODS AND SEED YIELD

Parameter 12 DAP 30 DAP SH Number Seed

TRL RN RDW TRL RN RDW DW of pods yield
TRL 12 DAP 1 0.59 0.15 0.19 0.03 0.18 0.17 0.04 0.26
R N 12 DAP 0.49 1 0.04 0.23 0.04 –0.01 –0.08 –0.08 –0.16
RDW 12 DAP –0.12 0.46 1 0.08 0.21 0.06 –0.06 –0.10 0.03
TRL 30 DAP 0.22 0.08 0.15 1 0.43 0.42 0.33 –0.08 0.26
RN 30 DAP 0.21 0.21 0.19 0.51 1 0.08 –0.02 0.04 0.19
RDW30 DAP –0.25 –0.25 0.01 0.23 –0.01 1 0.88 0.12 0.19
SDW 30 DAP –0.24 –0.03 –0.10 0.11 –0.14 0.87 1 0.12 0.17
Number of pods –0.20 –0.10 0.01 –0.19 –0.19 0.22 0.27 1 0.71
Seed yield –0.27 0.05 –0.12 –0.24 –0.33 0.27 0.32 0.64 1

3.3.5 Correlation between P efficiency and plant parameters

P efficiency was positively correlated to total root length at 30 DAP, shoot DW at 12

DAP, shoot and root DW at 30 DAP, number of pods plant -1, seed yield and 100 seed weight
at low P (Table 9).

TABLE 9. CORRELATIONS BETWEEN P EFFICIENCY AND PARAMETERS OF 50 SOYBEAN

GENOTYPES GROWN IN A P-DEFICIENT SOIL WITHOUT (LOW P) OR WITH P ADDITION
(HIGH P) AND LOW P AS A % OF HIGH P (LOW P/HIGH P)

Parameter Correlation coefficients (r)

Low P High P Low P/high P
-1
Total root length 12 DAP (cm plant ) 0.17 0.20 –0.04
Number of lateral roots 12 DAP –0.10 0.05 –0.19
Root DW 12 DAP (g plant-1) –0.06 –0.02 0.06
Shoot DW 12DAP ((g plant-1) 0.30* 0.09 0.14
Root: shoot ratio 12 DAP –0.17 0.05 –0.02
Total root length 30 DAP (cm plant-1) 0.29* –0.01 0.23
Number of lateral roots 30 DAP –0.06 –0.01 –0.13
Root DW 30 DAP (g plant-1) 0.80*** 0.09 0.66***
Shoot DW 30 DAP (g plant-1) 0.97*** –0.11 0.64***
Root: shoot ratio 30 DAP –0.37** –0.10 –0.24
Number of pods plant-1 0.24 0.14 0.10
Seed yield (g plant-1) 0.26 0.27 0.15
100 seed weight 0.29 0.23 0.04
*, P<0.05; **, P<0.01; ***, P<0.001

However, P efficiency was negatively correlated to number of lateral roots at 12 DAP,

root dry weight at 12 DAP, root: shoot ratio at 12 DAP and root: shoot ratio at 30 DAP (Table
9). Under low P, the highest correlation was observed between P efficiency and root and
shoot DW at 30 DAP. At high P, similar trends were observed with the exception of shoot
DW at 30 DAP, which was negatively correlated with P efficiency. Furthermore, P efficiency
showed the highest positive correlation with the values of root and shoot DW at 30 DAP at
low P as a percentage of high P (Table 9).

162
3.4. Phosphorus efficiency among cowpea genotypes

3.4.1. Shoot biomass

Shoot biomass increased significantly under moderate and high P limiting conditions.
Among the six genotypes there was an average increase of 55% of shoot biomass under
moderate P and 36 % increase of shoot biomass under high P conditions (Table 10).

Shoot biomass varied significantly (P<0.01) from 2.0 to 3.2 g. Genotype IT98K-1263
showed the highest shoot biomass while Soronko the least (Table 11). Genotypic variation in
shoot dry weight was related to differences in area per leaf and number of leaves (Table 10),
which accordingly affected total leaf area. In general, smaller leaves and lower leaf numbers
were observed in plants grown in low P treatments (P0, P20) as compared to those with high
P treatment (P40).

3.4.2. Nodulation, P uptake and P efficiency

Nodule number and nodule dry weight (NDW) increased significantly under moderate
and high P conditions (Table 10). Generally NDW increased 24% under moderate P and a
further rise of 11% under high P conditions. Significant variation was observed in nodulation
of cowpea genotypes under P limiting conditions.

TABLE 10. MEANS AND LEVEL OF SIGNIFICANCE FOR GROWTH PARAMETERS OF 6

COWPEA GENOTYPES WITH LOW P (0 KG HA-1) MODERATE P (20 KG HA-1) AND HIGH P
(40 KG HA-1) AVAILABILITY

Parameter P0 P20 P40 P0/P20 (%) P0/P40 (%) P G P×G

Leaf area(cm2 plant-1 ) 322 556 786 57.9 41.0 *** *** *
Leaf dry weight (g pot-1) 1.03 1.94 3.08 53.1 33.4 *** * ns
Leaf number 28 36 52 77.8 53.9 *** * ns
OLA 11.5 15.4 19.9 74.7 57.8 ** *** ns
Shoot dry weight (g pot-1) 2.41 4.33 6.39 55.7 37.7 *** ns ns
Root dry weight (g pot-1) 0.79 1.38 1.77 57.3 44.6 *** * ns
Root : shoot ratio 0.46 0.30 0.32 153.3 143.8 *** * ***
Nodule number 22 40 47 55.0 46.8 *** *** ns
Nodule dry weight (mg pot-1) 0.016 0.067 0.136 23.9 11.8 *** ** ns
P uptake (mg g-1) 10.9 23.7 42.8 46.9 25.5 *** ** ns
N uptake (mg g-1) 0.26 0.45 0.63 57.8 41.3 *** ns ns
P utilization efficiency 0.21 0.61 1.08 34.4 19.4 *** * ns
P uptake efficiency 14.30 9.70 9.28 147.4 154.1 *** * ns
*, P<0.05; **, P<0.01; ***, P<0.001; ns, not significant

Similarly, there were significant differences in root to shoot ratio, shoot P

concentration, root P concentration, ratio of shoot to root P concentration, shoot P content,
root P content, total P uptake, and P uptake efficiency amongst the genotypes and the P
treatments. However, with the exception of leaf area and root to shoot ratio no significant
interaction was observed (Table 10).

NDW varied from 3mg in IT93K-452-1 to 42 mg produced by IT98K-1263 under low

P conditions (Table 11). It is worth noting that genotype IT98K-1263 was consistently the
highest nodulating under all three P treatments.

163
164
TABLE 11. SHOOT, ROOT, NODULATION, P UPTAKE AND P UTILIZATION EFFICIENCY CHARACTERISTICS OF 6 COWPEA GENOTYPES
GROWN UNDER 0, 20 AND 40 KG P HA-1

Parameter† IT03K-351-1 IT00K-901-5 IT93K-452-1 IT98K-1263 IT97K-819-118 Soronko

0 20 40 0 20 40 0 20 40 0 20 40 0 20 40 0 20 40
Leaf area (cm2 pot-1) 201 446 649 319 492 624 371 526 822 433 565 864 295 362 467 189 859 905
Leaf number 27 37 59 32 41 53 34 31 70 27 32 37 36 43 63 17 34 29
Single leaf area (cm2) 7.51 11.85 12.92 11.31 13.17 14.76 11.12 17.22 13.07 15.88 18.45 24.82 9.56 9.27 8.51 15.19 25.72 42.58
Leaf DW (g pot-1) 0.83 1.95 3.24 0.88 1.91 2.18 1.05 1.74 3.57 1.55 2.11 4.19 1.08 1.53 2.47 0.78 2.39 2.83
Stem DW (g pot-1) 1.56 2.66 5.06 1.40 2.58 3.26 1.50 2.53 4.61 1.61 2.60 3.00 1.02 1.95 3.03 1.24 2.12 2.99
Shoot DW 2.39 4.61 8.30 2.28 4.49 5.44 2.55 4.27 8.18 3.16 4.71 7.19 2.10 3.48 5.50 2.02 4.51 5.82
Root DW 0.86 1.39 1.77 0.78 1.56 2.07 0.95 1.29 1.88 0.97 1.68 1.87 0.54 1.11 1.28 0.65 1.24 1.76
Root to shoot ratio 0.62 0.29 0.22 0.48 0.40 0.41 0.42 0.32 0.29 0.31 0.37 0.27 0.27 0.24 0.26 0.62 0.24 0.32
Nodule number 24 40 47 27 42 50 8 23 31 41 59 67 25 29 45 8 50 42
-1
Nodule DW (mg pot ) 25.82 79.7 156.8 19.53 73.1 159.7 3.07 43.3 89.7 42.41 100 183.7 19.24 43.8 110.5 9.65 64.4 115.3
Shoot N uptake (mg g-1) 8.58 28.69 74.38 7.85 36.24 61.58 13.95 30.43 59.90 15.69 38.91 63.95 4.95 22.53 43.26 5.60 26.53 58.52
Shoot P uptake (mg g-1) 0.84 1.90 4.46 0.80 2.12 3.48 0.97 2.03 5.18 1.11 1.84 3.45 0.63 1.25 2.41 0.57 1.65 2.57
P uptake efficiency 0.97 1.36 2.52 1.03 1.36 1.68 1.02 1.58 2.76 1.14 1.09 1.84 1.17 1.12 1.88 0.87 1.33 1.46
P utilization efficiency 0.21 0.71 1.42 0.18 0.62 0.76 0.23 0.51 1.26 0.37 0.80 1.31 0.18 0.42 0.72 0.17 0.59 0.98
†DW, dry weight
 The PEI value generated for the genotypes was positive for IT03K-351-1, IT00K-901-
5, IT93K-452-1and IT98K-1263 (Table 12). On the other hand PEI values generated were
negative for genotypes IT97K-819-118 and Soronko. Based on the PEI cowpea genotypes
were classified under three levels of P efficiency of low, moderate and high (Table 12).
Genotype Soronko was low, IT97K-819-118 moderate and IT03K-351-1, IT00K-901-5,
IT93K-452-1 and IT98K-1263 were highly efficient.

TABLE 12. CATEGORISING 6 COWPEA GENOTYPES INTO LOW EFFICIENCY (LE),

MODERATE EFFICIENCY (ME) AND HIGH EFFICIENCY (HE), ACCORDING TO P
EFFICIENCY INDEX (PEI) GENERATED FROM PCA USING PARAMETERS UNDER LOW P
(0 KG HA-1) AND HIGH P AVAILABILITY (40 KG HA-1)

ID Genotype Cluster PEI

8 IT03K-351-1 HE 0.059
11 IT00K-901-5 HE 0.064
14 IT93K-452-1 HE 0.121
22 IT98K-1263 HE 0.225
27 IT97K-819-118 ME –0.345
48 Soronko LE –0.094

Based on PEI generated from PCA and growth potential of shoot dry weight, nodule
dry weight and P uptake efficiency at high P, two categories were identified (Fig. 5).
Genotypes 11 (IT 00K-901-5, early maturing), 14 (IT93K-452-1, early maturing), 8 (IT03K-
351-1, early maturing) and 22 (IT98K-1263, medium maturing) were identified as efficient
and responders (ER). On the other hand genotypes 27 (IT97K-819-118, Striga resistant) and
48 (Soronko, released variety) were classified as inefficient and responders. Genotypes were
consistent in their classification by shoot dry weight, NDW and P uptake efficiency.
However genotype 22 (IT98K-1263, medium maturing) was highly efficient in NDW and P
uptake. Genotype 8 (IT03K-351-1, early maturing) was consistently highly efficient in shoot
biomass, NDW and P uptake.

3.4.3. Correlation analysis between P efficiency and plant growth parameters

P efficiency was positively correlated with leaf area, single leaf area, dry weights of
leaf stem, shoots and roots, root to shoot ratio, nodule number, NDW, shoot N and P uptake
and P utilization efficiency, and negatively correlated to leaf number and P utilization
efficiency at P0 (Table 13). At P40, P efficiency was positively correlated with all parameters
measured but negatively correlated with leaf number. Likewise, P efficiency was positively
correlated the relative indices at P0 as a % of those at P40, except leaf area, single leaf area,
leaf dry weight and P uptake efficiency.

165
 1.8 8
1.6 ER
14
NER 1.4
1.2
1 22
0.8
0.6
27 0.4
48 11
0.2
0
-0.400 -0.300 -0.200 -0.100 0.000 0.100 0.200 0.300

1.6 22
1.4 ER
Nodule Dry Weight

NER 11
1.2 8
1
0.8
48
27 0.6
14
0.4
0.2
0
-0.400 -0.300 -0.200 -0.100 0.000 0.100 0.200 0.300
1.2
22
Phosphorus uptake

1
NER
efficiency

48 0.8 11
8 14
0.6
27
0.4 ER

0.2

0
-0.400 -0.300 -0.200 -0.100 0.000 0.100 0.200 0.300

FIG. 5. Classification of cowpea genotypes into 2 distinct responsive groups according to P efficiency
index and standardized values of shoot biomass, nodule dry weight and P uptake efficiency under high
P availability.

166
TABLE 13. CORRELATIONS BETWEEN P EFFICIENCY AND PARAMETERS OF 6 COWPEA
GENOTYPES GROWN IN A P-DEFICIENT SOIL WITHOUT (P0) OR WITH P ADDITION (P40)
AND PARAMETERS AT P0 AS % OF THOSE AT P40 (P0/P40)

Parameter Correlation coefficients (r)

P0 P40 P0/P40
2 -1
Leaf area (cm pot ) 0.48 0.61 –0.21
Leaf number –0.16 –0.15 0.16
Single leaf area (cm2) 0.31 0.11 –0.43
Leaf dry weight (g pot-1) 0.37 0.67 –0.40
Stem dry weight (g pot-1) 0.96** 0.33 0.46
Shoot dry weight 0.77 0.60 0.09
Root dry weight 0.95** 0.85* 0.58
Root to shoot ratio 0.15 0.10 0.13
Nodule number 0.26 0.33 0.05
Nodule dry weight (mg pot-1) 0.33 0.52 0.03
Shoot N uptake (mg g-1) 0.84* 0.77 0.72
Shoot P uptake (mg g-1) 0.85* 0.66 0.07
P uptake efficiency –0.12 0.31 –0.33
P utilization efficiency 0.67 0.69 0.01
*, P<0.05; **, P<0.01

4. DISCUSSION

4.1. Shoot and root growth at early growth stage

The significant differences and correlation found among genotypes within soybean
and cowpea for total root length, root number, root weight, shoot weight and root: shoot ratio
were similar to the findings of Mia et al. [24] who found significant differences in several root
traits among legumes they studied during early growth. Interestingly, they reported that
cowpea showed the widest genotypic variability for root traits, and formed an extensive root
system by producing a large number of lateral roots, comparable to the findings in this study.
Extensive lateral rooting systems have been shown to be important adaptive traits that
enhance P uptake from low P soils [25].

Genotypic differences were observed in soybean and cowpea genotypes for lateral
rooting at 5 DAP. This has been reported in early growth stage in both chickpea and cowpea
[24]. The deployment of root architectural traits in plant breeding programs has great potential
to alleviate P deficiency, a primary constraint to crop production in world agriculture [18].
The significant differences found in soybean and cowpea at the seedling stage provides an
opportunity for screening a large number of genotypes for important root traits such as an
extensive lateral root system.

4.2. Phosphorus efficiency in soybean and cowpea

Soybean and cowpea, like most legumes grown in Ghana, are grown under little or no
nutrient application. In common with most soils found in Ghana, the levels of available P in
the soils used for the study were low. Similarly, the total N was also low, consistent with the
relatively low OM status of the soil. The most common stress that affects grain and forage
production in Ghana is nutrient stress, particularly low P. The soil would therefore need some
fertilization from an external source to sustain crop growth and permit continuous cultivation
[26]. However this is beyond the economical reach of most Ghanaian farmers and has

167
prompted genotype screening and selection for tolerance to low soil P conditions as an
important strategy to increase productivity [27].

We observed substantial variation among 50 soybean genotypes for growth in low P

soil and under high P conditions for most of the parameters studied. Genotypic evaluation of
crops should be under both low and high availability P since P efficiency is a complex
quantity trait involving growth parameters [23]. As in soybean, genotypic variation was
observed in 6 cowpea genotypes studied under P limiting conditions and high P conditions.
Genotypic variation in P efficiency has been identified in cowpea [19, 28, 29].

The method used for P efficiency analysis was the PCA which is relatively new
multiple-parameters screening method used in evaluating P efficiency. Due to the sensitive
nature of parameters such as shoot biomass and P uptake to P availability, multiple- parameter
screening methods are ideal because unlike single-parameter screening methods, they take
into account the relative contributions of all parameters measured to P efficiency [23].
Utilizing the same method, 50 soybean genotypes could be grouped into 3 categories of P
efficiency and 4 categories according to PEI, and in combination with P responsiveness, shoot
dry weight and seed yield at high P. Shoot biomass and yield are important parameters in
breeding, and shoot biomass is also an important determinant of seed yield [23]. Similar
results were obtained by Pan et al. [23] using the same method of analysis for soybean
genotypes.

The cowpea genotypes were also grouped under three categories of P efficiency.
Consistent with the findings of others [23, 30], as well as the findings from this study, PEI
generally places genotypes into three main categories of P efficiency of low, moderate and
high. The 6 cowpea genotypes were further classified into two P responsive groups using the
standardized values of shoot dry weight at high P and P efficiency. In contrast, the 50 soybean
genotypes were classified under four P responsive categories. The categories obtained for
cowpea were (i) efficient and responsive (ER); (ii) inefficient and responsive (NER). This
grouping indicates that even though all 6 genotypes were good responders to P application,
genotype was either efficient or inefficient.

IT03K-351-1, IT 00K-901-5, IT93K-452-1 and IT98K-1263 were more P efficient

than genotypes IT97K-819-118 and Soronko. According to Sanginga et al. [19], genotypes
with increased shoot dry weight with increasing levels of P, as observed in this study,
distinguishes all genotypes as P-responders. Genotype 48 (Soronko) was identified as low
efficiency under PEI, but was very responsive to P application. For instance, it had the highest
leaf area when P was applied at 40 kg ha-1 and produced the highest percentage increase when
P was applied. It is a typical cowpea variety which was produced during earlier breeding
programmes by IITA to respond well to P application. In earlier IITA breeding programmes,
yield parameters were predominantly targeted while factors such as P use and uptake
efficiency and root traits may not have been deliberately included.

When the combination of PEI with P responsiveness at high P is considered, soybean

genotypes 4, 43, 46, 55, 67, 77, 106 and 145 were consistently classified as efficient
responders for both shoot and seed yield, and therefore were the best genotypes, particularly
for the P-deficient soil used for the study. Genotypes categorized as the most P efficient are
most responsive to increased P availability [30]. Remarkably, the genotypes listed above were
classified as only moderately efficient using the PEI generated from the PC analysis. The
following genotypes could be classified as highly efficient: 10, 104, 129 and 141 showed
marked differences when classified according to shoot and seed yield potential at high P.
Genotype 129 was an efficient responder in terms of potential shoot biomass at high P, while

168
genotype 10 was an inefficient responder, and genotypes 104 and 141 were efficient non-
responders. In terms of potential yield at high P, genotypes 141 and 104 were efficient
responders, and genotypes 10 and 129 were efficient non-responders.

P application significantly increased leaf dry weight, stem dry weight, root dry weight,
leaf area, nodulation, P uptake and P uptake efficiency of all 6 cowpea genotypes. Increased
shoot growth of cowpea genotypes in response to P application in low P soils has also been
reported by several researchers [28, 29, 31–33]. The enhancing effects of P application on
nodulation have been associated with increased nodule mass and number in cowpea as shown
by this study and detailed by the other researches [19, 34, 35]. In contrast, Kolawole et al.
[29] found a decrease in nodule number when P was applied in cowpea production.

4.3. Correlation among parameters

With the exception of shoot biomass and root biomass no significant correlation was
found between PEI and most of the parameters studied in soybean genotypes. The study also
indicated little correlation between the plants harvested at 5 DAP and shoots and yield
parameters in the field. One possible explanation for the differences between the field and
greenhouse results is that the field environment may have presented other environmental
variables and stresses that could have difference observed in the plants. Lack of correlation
between root traits and yield per plant under low P in the field could have been caused by the
uncertainty associated with yield estimates from small plots in a single location and season.

5. CONCLUSIONS

The study concluded that there were significant genotypic variation for root traits at an
early growth stage and genotypic differences for soybean and cowpea growth under low P.
The PEI in combination with P responsiveness at high P showed that soybean genotypes 4,
43, 46, 55, 67, 77, 106 and 145 were efficient responders for both shoot and seed yield at high
P, and therefore were the best genotypes particularly for the P-deficient soil used for the
study. All 6 cowpea genotypes used in the P efficiency study, were good responders to P
application, but IT03K-351-1, IT 00K-901-5, IT93K-452-1 and IT98K-1263 were more P
efficient than IT97K-819-118 and Soronko.

ACKNOWLEDGEMENTS

The research work was financially supported by the International Atomic Energy
Agency (Vienna, Austria; Project No. 13800/RO) 2006–2008. The authors also thank
International Institute of Tropical Agriculture (IITA) and for the soybean and cowpea seed
used in the study.

REFERENCES

[1] VAN STRAATEN, P., Rocks for Crops: Agrominerals of Sub-Saharan Africa, ICRAF,
Nairobi, Kenya (2000) 338 p.
[2] KASSEM, A., The Effect of Rates of Phosphorus and Nitrogen on the Growth,
Nodulation and Yield of Soybean, M. Sc. Thesis, Department of Biological Sciences,
Ahmadu Bello University, Zaria, Nigeria (1991).
[3] VASILAS, B.I, et al., Relationship of nitrogen utilization patterns with soybean yield
and seed-fill period, Crop Sci. 35 (1995) 809–813.
[4] SARI (Savanna Agricultural Research Institute), Annual Report, SARI, Tamale,
Ghana (1999).

169
[5] SANCHEZ, P.A., et al., “Soil fertility replenishment in Africa: An investment in
natural resource capital”, Replenishing Soil Fertility in Africa, (BURESH, R.J.,
SANCHEZ, P.A., CALHOUN, F., Eds), Soil Science Society of America Special
Publication, Madison, Wisconsin, USA (1997) 1–46.
[6] OWUSU-BENNOAH, E., ACQUAYE, D.K., Phosphate sorption characteristics of
selected major Ghanaian soils, Soil Sci. 148 (1989) 114–123.
[7] ABEKOE, M.K., TIESSEN, H., Fertilizer P transformations and P availability in hill
slope soils of northern Ghana, Nutr. Cycl. Agroecosyst. 52 (1998) 45–54.
[8] OWUSU-BENNOAH, E., et al., Soil properties of a toposequence in the moist semi-
deciduous forest zone of Ghana, West Afr. J. Appl. Ecol. 1 (2000) 1–10.
[9] SENAYAH, J.K., KUFOGBE, S.K., DEDZOE, C.D., Land degradation in the Sudan
Savanna of Ghana: A case study in the Bawku Area, West Afr. J. Appl. Ecol. 8 (2005)
127–136.
[10] ASOMANING, S.K., ABEKOE, M.K., OWUSU-BENNOAH, E., Phosphate rock
dissolution and availability in some soils of semi-deciduous Rainforest Zone of Ghana,
West Afr. J. Appl. Ecol. 10 (2006) 154–164.
[11] ORTIZ-MONASTERIO, J.I., MANSKE, G.G.B., VAN GINKEL, M., “Nitrogen and
phosphorus use efficiency”, Application of Physiology in Wheat Breeding,
(REYNOLDS, M.P., ORTIZ-MONASTERIO, J.I., MCNAB, A., Eds), CIMMYT,
Mexico (2001) 200–207.
[12] ANNE, J.B., LAL, R., “The Tropical soil productivity calculator - a model for
assessing effects of soil management on productivity”, Soil management:
Experimental Basis for Sustainability and Environment Quality, Advances in Soil
Science (LAL, R., STEWART, B.A, Eds), Lewis Publishers, Boca Raton, USA (1995)
499–520.
[13] YERIMS, A., VAN RANST, E., Introduction to Soil Science; Soils of the Tropics,
Trafford publishing, UK (2005) 440 p.
[14] GGDP, Ghana Grains Development Project, 16th Annual Report (1994) 136 p.
[15] QUAYE, W., Food security situation in northern Ghana, coping strategies and related
constraints, Afr. J. Agric. Res. 3 (2008) 334–342.
[16] VIEIRA, R.F., JOCHUA, C.N., LYNCH, J.P., Method for evaluation of root hairs for
common bean genotype, Rev. Agropec. Bras. 42 (2007) 1365–1368.
[17] LYNCH, J.P., BROWN, K.M., Topsoil foraging – an architectural adaptation of plants
to low phosphorus availability, Plant Soil 237 (2001) 225–237.
[18] LYNCH, J.P., Turner Review No. 14. Roots of the second green revolution, Aust. J.
Bot. 55 (2007) 493–512.
[19] SANGINGA, N., LYASSE, P.O., SINGH, B.B., Phosphorus use efficiency and
nitrogen balance of cowpea breeding lines in low P soil of the derived savannah zone
in West Africa, Plant Soil 220 (2000) 119–128.
[20] TENNANT, D., A test of a modified line intersects method of estimating root length. J.
Ecol. 63 (1975) 995–1001.
[21] BREMNER, J.M., Determination of nitrogen in soil by the Kjeldahl method, J. Agric.
Sci. 55 (1960) 11–33.
[22] WALKLEY, A., BLACK, I.A., An examination of the Degtjareff method for
determining soils organic matter and a proposed modification of the chromic acid
titration method, Soil Sci. 37 (1934) 29–38.
[23] PAN, X., et al., Assessment of phosphorus efficiency characteristics of soybean
genotypes in phosphorus-deficient soils, Agric. Sci. Chin. 7 (2008) 958–969.
[24] MIA, W., YAMAUCHI, A., KONO, Y., Root system structure of six food legume
species: inter- and intra-specific variations, Jap. J. Crop Sci. 65 (1996) 131–140.

170
[25] ZHU, J., LYNCH, J.P., The contribution of lateral rooting to phosphorus acquisition
efficiency in maize (Zea mays) seedlings, Funct. Plant Biol. 31 (2004) 949–958.
[26] DARKO, A. D., Synchronizing Nitrogen Release from Plant Residues and Maize
Uptake; the Effect of Residue Type, Application Method and Soil Moisture, M. Phil.
Thesis, University of Ghana, Legon (2007) 150 p.
[27] ADU-GYAMFI, J.J., et al., “Genetic variability and physiological characteristics of
crop plants adapted to low-nutrient environments”, Food Security in Nutrient-Stressed
Environments: Exploiting Plants’ Genetic Capabilities, (ADU-GYAMFI, J.J, Ed.),
Kluwer Academic Publishers, Dordrecht, Netherlands (2002) 67–79.
[28] ANKOMAH, A.B., et al., Yield, nodulation, and N2 fixation by cowpea cultivars at
different phosphorus levels, Biol. Fertil. Soils 22 (1996) 10–15.
[29] KOLAWOLE, G.O., TIAN, G., SINGH, B.B., “Differential response of cowpea lines
to application of P fertilizer”, Challenges and Opportunities for Enhancing Sustainable
Cowpea Production, (FATOKUN, C.A., et al.,, Eds), Proc. World Cowpea Conf. III,
International Institute of Tropical Agriculture (IITA), Ibadan, Nigeria (2002) 319–328.
[30] BAYUELO-JIMÉNEZ, J.S., et al., Genotypic variation for root traits of maize (Zea
mays L.) from the Purhepecha Plateau under contrasting phosphorus availability, Field
Crops Res. 121 (2011) 350–362.
[31] BATIONO, A., et al., “Soil fertility management and cowpea production in the
semiarid tropics”, Challenges and Opportunities for Enhancing Sustainable Cowpea
Production, (FATOKUN, C.A., et al., Eds), Proc. World Cowpea Conf. III,
International Institute of Tropical Agriculture (IITA), Ibadan, Nigeria (2002) 301–318.
[32] OWOLADE, O.F., et al., Effects of application of phosphorus fertilizer on brown
blotch diseases of cowpea, Afr. J. Biotechnol. 5 (2006) 343–347.
[33] NWOKE, O.C., et al., Utilization of phosphorus from different sources by genotypes
of promiscuous soybean and cowpea in a low-phosphorus savanna soil, Afr. J. Agric.
Res. 2 (2007) 150–158.
[34] WAN OTHMAN, W.M., et al., Low level phosphorus supply affecting nodulation, N2
fixation and growth of cowpea (Vigna unguiculata L. Walp), Plant Soil 135 (1991)
67–74.
[35] KUMATA, A.S., ALIYU, B.S., SARATU, A.O., Influence of phosphorus fertilizer on
the development of nodules in cowpea (Vigna unguiculata L. Walp) and soybean
(Glycine max L. Merrill), Int. J. Pure Appl. Sci. 2 (2008) 27–31.

171
 GENOTYPIC VARIATION IN PHOSPHORUS USE EFFICIENCY FOR
SYMBIOTIC NITROGEN FIXATION IN VOANDZOU (VIGNA
SUBTERRANEA)

A. ANDRIAMANANJARA , L. RABEHARISOA
Laboratoire des Radio-isotopes,
Université d'Antananarivo,
Antananarivo,
Madagascar

M. MALAM ABDOU
Laboratoire Banques de gènes CERRA / KOLLO,
Institut National de Recherche Agronomique du Niger (INRAN),
Niamey,
Niger

D. MASSE
Institut de Recherche pour le Développement,
UMR Eco&Sols,
Montpellier,

L. AMENC, C. PERNOT, J.J. DREVON

Institut National de la Recherche Agronomique,
UMR Eco&Sols,
Montpellier,

France

Abstract

Vigna subterranea, known as voandzou or Bambara groundnut as an African indigenous crop

which is often neglected or under-used in African subsistence agriculture. Preliminary research and
country perceptions have shown its agronomic and nutritional properties, in particular under atypical
climates of arid and tropical areas, and in saline soils. There is a high potential to increase the
production by optimizing symbiotic nitrogen fixation (SNF) through effective inoculation even in
nitrate-rich environments. In this study, Vigna subterranea inoculated with the reference strain of
Bradyrhizobium sp. Vigna CB756 was studied in order to assess the symbiotic fixation potential of
different cultivars and landraces of Madagascar, Niger and Mali under low-P and sufficient-P
conditions. Six voandzou cultivars inoculated with Bradyrhizobium sp. Vigna CB756, were grown
under hydroaeroponic culture for 6 weeks supplied with four phosphorus levels of 15, 30, 75 and 250
µmol plant-1 week-1 in order to establish the response curve of voandzou to P supply, and to induce P
deficient and sufficient levels. In another experiment five tolerant cultivars with high SNF and five
sensitive cultivars with low SNF were chosen after a preliminary screening of 54 voandzou genotypes,
including 50 landraces from Madagascar, Niger and Mali supplied with 2 P levels as P deficient and P
sufficient (30 and 75 µmol plant-1 week-1 ) under hydroaeroponic conditions. Genotypic variation in
SFN for the high phosphorus use efficiency (PUE) was observed among the 54 cultivars and
landraces. Variability was especially related to the nodule and shoot biomass, nodule permeability,
nodule respiration and gene phytase expression. Contrasting cultivars and landraces in terms of PUE
for SNF were selected for further evaluation under field conditions.

173
1. INTRODUCTION

The improvement of symbiotic nitrogen fixation (SNF) in legumes makes it possible

to mitigate N deficiency of many soils, to promote the availability of some soil nutrients such
as P, and consequently to improve the crop’s production, especially in countries with low use
of inputs [1]. In addition, variability between landraces exists for certain traits related to N2
fixation [2]. In this way, researches for SNF optimization among legume species were
undertaken by numerous researchers [3–9]. A Thompson Boyce Institute report in 1978
mentioned that the genetic factors implied in nodulation should be used in the legume
selection in particular for the specific nodulation traits (weight, number), including the
inoculation of adequate rhizobia [10]. This concept was confirmed by the orientation of the
research carried out during recent years, which involves the integration of plant-soil-
microorganisms system. Indeed, the knowledge of the legume’s capacity to use nutrients
under sufficiency or deficiency of P would allow better adaptation of legumes in farming
systems for productivity improvement. Thus, one way is the study of legume genotype-
rhizobia combinations in N2 fixation improvement [11]. Indeed, genetic variability in SNF
under P sufficiency or deficiency was the subject of many studies for numerous species of
legumes including Phaseolus vulgaris [1, 12, 13], Vigna unguiculata [14], Glycine max [2]
and Vigna subterranea [3]. The increase of nodular respiration under P deficiency was
observed for Glycine max [15] and Phaseolus vulgaris [16].

Vigna subterranea, known as the vernacular name of voandzou or Bambara groundnut

as an African indigenous crop, often neglected or under-used in African subsistence
agriculture [17]. Preliminary research and country perceptions show the potentiality of its
agronomic and nutritional properties, in particular under atypical climates of arid and tropical
areas, and in saline soils. The genetic potential is little known, and some previous work is
available only in old publications or little-known languages [17, 18]. A genotype selection of
Vigna subterranea was initiated in multi-locational field trials in Africa, on the basis of their
vigor, resistances against fungal diseases and yield, in particular in Nigeria, Burkina Faso,
Camerounn and Botswana, showing a large special-temporal effect from one year to another
[19–21]. Great variation of yield from 200 to 3000 kg ha -1 among farming systems [18, 22]
and spatial variability for the same landraces makes it possible to formulate the hypothesis of
a differentiation in the symbiotic effectiveness of native Bradyrhizobia nodulating Vigna
subterranea. Moreover, the inoculation of Vigna subterranea with a Bradyrhizobium sp.
strain showed a significant increase of production in Togo and Senegal. These observations
indicate the potential to increase the production of Vigna subterranea via SNF optimization
[18]. The nodulation of Vigna subterranea was mainly depending on symbiotic nitrogen
fixation even in nitrate-rich environments [3].

In the present study, Vigna subterranea was inoculated with the reference strain of
Bradyrhizobium sp. Vigna CB756 in order to assess the symbiotic fixation potential of
different cultivars and landraces of Madagascar, Niger and Mali under P deficiency versus P
sufficiency. The results of this experimentation allow us to highlight and to select contrasting
lines of voandzou in P use efficiency (PUE) for SNF.

2. MATERIALS AND METHODS

2.1. Biological material and symbiosis culture

The experiment was carried out in hydroaeroponic culture under glasshouse conditions
during 2010 with 20/33°C temperature during 16 / 8 h day / night cycle coupled with an
intense ventilation of nodulated-roots, a complementary illumination of 400 µM photons m-2

174
when needed, and 70% daily relative humidity. These conditions allowed an optimal
expression of genetic potential of N2 fixation while controlling closely the mineral nutrition
mainly for N, P and for other essential elements.

The seeds were sterilized with 3% calcium hypochlorite for 20 min and rinsed by 5
washings with sterile distilled water. They were then transferred for germination on
humidified wrapped filter paper in a slightly tilted vat and placed in an incubator at 28-30°C.
After germination, inoculation was performed by soaking the roots of seedlings for 20 min in
a suspension of Bradyrhizobium sp. Vigna CB756 containing 109 bacteria ml-1. The inoculum
was prepared from a rhizobial culture preserved in tubes at 4°C on 120°C sterilized agar YEM
(Yeast Extract Mannitol) medium: 900 ml distilled water, 100 ml of Bergersen concentrated
solution (which is prepared with a mixture of 1 g KCl; 0.1 g FeCl3 ; 0.4 g CaCl2, 4.5 g
Na2HPO4.12H2O and 1 g MgSO4.7H2O, firstly in 100 ml of distilled water, then adjusted to 1
l), 1 g Yeast Extract, 10 g mannitol and 15 g agar [23]. From one of the preserved tubes, some
strains were taken and put on 100 ml of liquid YEM (without agar), and maintained at 28°C
for 24 h. The germinated seeds were carefully transferred for hydroaeroponic culture into
each 40 l vat with 20 inoculated plants per vat. The roots of each seedling were carefully
passed through the hole of a rubber stopper with cotton wool fixed at the hypocotyl level.
They received the following nutrient solution that was changed every 2 weeks: CaCl2, 1650
µM; MgSO4.7H2O, 1000 µM; K2SO4, 700 µM; Fe EDDHA, 8.5 µM as sequestrene Fe;
H3BO3, 4 µM; MnSO4.H2O, 6 µM; ZnSO4.7H2O, 1 µM; CuSO4.7H2O, 1 µM;
Na2MoO4.7H2O, 0.1 µM. 2000 µM of urea was supplied for all treatments during the first two
weeks’ culture. P was supplied in the form of KH2 PO4 with an exponential distribution for the
two levels: 30 µmol plant-1 week-1 and 75 µmol plant-1 week-1.

In the first experiment, six voandzou (Vigna subterranea) cultivars (kindly supplied
by FOFIFA, Madagascar), inoculated with Bradyrhizobium sp. Vigna CB756, were grown
under hydroaeroponic culture during six weeks in order to test the effectiveness of the
rhizobia-legume symbiosis. Four P levels, 15, 30, 75 and 250 µmol plant -1 week-1, were
supplied under KH2PO4 form in order to establish the response curve of voandzou to P supply
and to deduce P deficient and sufficient levels. Four replications were applied for each
treatment.

In the second experiment, a preliminary screening of 54 voandzou genotypes

including 14 landraces and 2 cultivars from FOFIFA in Madagascar, 36 landraces from Niger
and 2 cultivars from Mali, was performed with one replicate during six weeks in containers
under hydroaeroponic conditions. Two P supplies of P deficient as 30 µmol plant -1 week-1 and
P sufficient as 75 µmol plant-1 week-1 deduced from the previous experiment were applied in
order to screen the most tolerant and sensitive cultivars to P sufficiency versus P deficiency.
Five tolerant cultivars with high SNF and 5 sensitive cultivars with low SNF were chosen
from this prescreening on the basis of their nodule and shoot biomass under P deficiency.
These 10 most contrasting cultivars were grown under P deficiency and P sufficiency in a
randomized block design with 8 replicates. Three weeks after transplanting, the four most
contrasting cultivars, 2 tolerant and 2 sensitive, were transferred for oxymetric measurement
with three replications into 1 l bottles wrapped with aluminum foil to maintain darkness in the
rooting environment. These bottles contained the same nutrient solution as the containers
previously described.

175
2.2. Measurements of nodulated root O2 uptake

In order to assess the tolerance mechanisms in SNF under P deficiency, the nodulated
root gas exchange was performed with the four most contrasting cultivars and landraces. The
consumption of O2 by the nodulated roots (Conr) was measured in situ between 09.00 and
16.00 h with an oxymeter (Abiss, Verpillère, France) at the 6th week after transplanting. The
level of the nutrient solution in each bottle was reduced to one-third of the volume one day
before the Conr measurement for putting the whole nodule population in direct contact with
the gas phase. The measurement was performed with the oxymeter connected to a recorder
and a peristaltic pump to ensure the circulation and continuous homogenization of the gas
phase between the nodulated roots and the oxymeter, with 400 ml-1 of flow [24].

The O2 consumption was quantified according to a known volume of the bottle at

different O2 levels: 21, 25, 30, and 40 kPa O2. Conr was calculated as Conr = ∆pO2 (V / 24.2)
(60 / t), with ∆pO2 = pO2 initial – pO2 final in % of atmospheric pressure; V in l, volume of the
gas phase under the experimental conditions; t in min, time between the initial and final
measurement of O2 ; 24.2 in l, volume of 1 pure gas mol under the experimental conditions
[25]. Conr is expressed in µmol O2 consumed plant-1 h-1.

2.3. In situ RT-PCR

Nodules of 3 mm diameter were randomly harvested at early flowering from

replicated plants of each P treatment, thoroughly washed with DEPC- (diethyl pyrocarbonate)
treated water, then fixed in freshly prepared PFA (v/v) (2% paraformaldehyde, 45% ethanol
and 5% acetic acid) and stored overnight at 4°C. Fixed nodules were extensively rinsed with
four washings of DEPC-treated water over 30 min (2 x 5 min and 2 x 10 min) with agitation
to remove PFA. Thereafter, the nodules were included in low melting 9% (m/v) agarose
dissolved in filtered phosphate-buffered saline (PBS; 5 mM Na2HPO4, 300 mM NaCl, pH 7.5)
and cut into 50 µm thick sections using a microtome. The resulting sections were collected
into tubes containing 0.5 ml of DEPC-treated water and freed from residual agarose by three
washes with DEPC-treated water at 60°C. For reverse transcription, the fixed sections were
transferred to PCR tubes and incubated in 40 µl RT mix [RT 1X Reaction Buffer (50 mM
Tris-HCl, pH 8.3, 75 mM KCl, 3 mM MgCl2, 10 mM DTT) (Promega, Madison, WI, USA);
0.31 mM dNTP and 0.75 µM gene specific reverse primer (5’-TTC ACC TCT AGA ATC
CCA T-3’). The samples were then heated at 65°C for 5 min, transferred on ice for 2 min and
added with Moloney murine leukemia virus (M-MLV) reverse transcriptase H(−) (Promega)
to a final concentration of 5 U µl-1 followed by incubation at 42°C during 1 h. After reverse
transcription, the RT mix was removed, and the samples were washed three times with 100 µl
DEPC-treated water. After removing the last washing, they received 40 µl of of PCR mix (1×
PCR buffer, Invitrogen, Carlsbad, CA, USA), 1.5 mM MgCl2, 0.2 mM of each dNTP, 0.25
µM each of the gene-specific primer pair (Forward: 3’-GGA CAT GTT CAT GCC TAT
GAG-5’; Reverse: 5’-TTC ACC TCT AGA ATC CCA T-3’), 0.25 nM digoxigenin-11-2′-
deoxyuridine 5′-triphosphate (Dig-11-dUTP; Roche Diagnostics, Mannheim, Germany) and 1
U Taq DNA Polymerase (Invitrogen). Thermocycling was performed at 95°C for 3 min and
30 cycles (95°C for 30 s; 55°C for 30 s; 72°C for 45 s, 72°C for 2 min). Negative controls
(no-RT) were prepared by omitting the reverse transcription with the samples in 40 µl of
DEPC-treated water during the RT step and treating alike the other samples during the
following steps.

For the detection of the amplified cDNA, the PCR mix was removed after
amplification, and the samples were washed three times each for 10 min in 200 µl PBS under
gentle agitation, and then incubated in 100 µl blocking solution (2% BSA, in PBS, with 0.3%

176
triton) for 30 min under gentle agitation in darkness at 37°C. Then the blocking solution was
removed and replaced by 100 µl of alkaline phosphatase-conjugated anti-dioxygenin-Fab
fragment (Roche Diagnostics) diluted 1: 1000 in 2% BSA. The samples were incubated at
room temperature for 90 min and then washed three times for 10 min in PBS to remove
excess antibody. Detection of alkaline phosphatase was carried out using the ELF-97
(enzyme-labeled fluorescent) endogenous phosphatase detection kit (Molecular Probes,
Leiden, The Netherlands). The ELF substrate was diluted 1: 40 in the alkaline detection buffer
(Molecular Probes, Leiden, The Netherlands), vigorously shaken, and then filtered through a
0.22 µm filter (Millex®-GV, Millipore, Bedford, USA) to remove any aggregates of the
substrate that may have formed during storage. Samples were incubated in 20 µl ELF
substrate-buffer solution in the dark for 20 min, then washed 3 x 1 min with wash buffer (PBS
with 25 mM EDTA and 5 mM levamisole, pH 8.0) before the samples were mounted.
Observations were made using an Olympus BX61® microscope equipped with an
epifluorescence condenser, a Hoechst / DAPI filter set and color view camera.

2.4. Biomass, P content and statistical analysis

The plants were harvested at the end of the 6th week at the flowering stage. The shoot
was separated from the root at the cotyledonary node, and then weighted after drying for 48 h
at 80°C. Nodules were excised from roots, counted and weighted. The effect of P supply was
estimated from the response curves of the biomass and the critical P supply was deduced at
the maximum growth. In order to determine the contribution of P supply to plant growth, the
response curves were established from biomass values with the software Excel Microsoft
Office XP. The genotypes and P treatment effects were analyzed by two-way analysis of
variance (ANOVA). Correlation and regression analyses were performed to complete the
analysis.

3. RESULTS

3.1. Response curves of biomass to P supplies

A significant difference in the response to P supply was observed for the cultivars 1
and 3 in terms of nodulation and plant biomass compared to cultivars 4 and 6 (Fig. 1). The
maximum nodulation was 27 and 32 mg nodule DW plant -1 for cultivars 1 and 3, respectively,
versus around 7 mg nodule DW plant-1 for cultivars 4 and 6 (Fig. 1A). The critical P supply,
i.e. for which plants could express their genetic potentials in terms of nodulation, were around
125 µmol Pi plant-1 week-1 for cultivars 1, 4 and 6 versus between 150 and 175 µmol Pi plant-
1
week-1 for cultivar 3. The shape of the curves showed a decrease in the nodular biomass
after reaching the critical P supply. From these response curves, 30 versus 75 µmol Pi plant-1
week-1 were selected for P sufficiency and deficiency in subsequent experiments. Cultivar 1
was the most sensitive to P supply in terms of nodule, shoot and root biomass while cultivars
4 and 6 presented a low response slope for all biomass parameters. Cultivars 1 and 4, which
were respectively sensitive and tolerant to P supply, were selected as controls in the screening
experiments.

3.2. Screening of voandzou for nodulation under P sufficiency versus P deficiency

The first screening of 58 cultivars and landraces of voandzou without replication

enabled us to preselect contrasting landraces as tolerant and sensitive. A large variability
among cultivars and landraces of voandzou was evident (Fig. 2). The nodulation was higher
for Madagascar landraces than for Niger and Mali cultivars and landraces. The results
observed under P deficiency exceeded those under P sufficiency for Niger and Mali landraces.

177
The cultivars and landraces with high nodulation ability under P deficiency harboured more
than 0.2 g nodule DW plant-1 (Fig. 2). Under P deficiency, a significant nodulation,
corresponding to a high SNF, was recorded in landrace 7 followed by V1, V4, 8 and 1, 11
compared to low nodulation in cultivars 2, 10, N’jibawolo and 4.

FIG. 1. Response curves of growth of nodules (A), shoots (B) and roots (C) as a function of weekly P
supplies. Data are means of 4 replicates for 4 voandzou genotypes harvested at 45 DAS. Two P levels
were identified as P sufficient and P deficient: 30 versus 75 µmol Pi plant-1 week-1.

Thus, the first screening of Madagascar, Niger and Mali cultivars and landraces
highlighted 10 contrasting cultivars and landraces whose tolerant lines for shoot and nodule
biomass were the cultivar VMDV4 and the 4 landraces VMD4, VMD7, VMD8 and VMD10,
while the sensitive lines were the cultivar VMDV1 and the 4 landraces VMD1, VMD2,
VMD11 and N’jibawolo.

3.3. Efficiency in use of rhizobial symbiosis (EURS)

The SNF was closely related to biomass produced during the culture marked by a
significant relationship between nodule and shoot biomass (Fig. 3). Genotypic variability was
observed in some cultivars and landraces where high nodulation was associated with high
shoot biomass. The shoot biomass of landrace 7 and cultivar 1 were significantly the highest,
with more than 4 g DW plant-1, and nodulation between 0.2 g and 0.4 g nodule DW plant -1
(Fig. 3) compared with landraces 8, 11 and 1, and cultivar 4.

178
 The plants response to SNF was assessed by the regression line of shoot biomass as a
function of the nodule biomass. Thus, EURS between different cultivars and landraces was
calculated from the regression slope. The contrasting lines for EURS were the landraces 2, 7
and cultivar 1, with circa 14 g shoot DW g-1 DW nodule under P deficiency compared to the
others (Fig. 3). Interestingly, low nodulation with weak nodule biomass in landrace 2 was
compensated by a great EURS. The other landraces presented a lower EURS under P
deficiency than under P sufficiency. Such tolerant cultivars as V4 and 1 were in this group.

FIG. 2. Nodule growth of voandzou cultivars under P sufficiency (white bar) versus P deficiency
(black bar) under hydroponic conditions at 45 DAS. 58 cultivars were prescreened without
replication. The 10 most contrasting cultivars from prescreening based on nodule biomass were
replicated. Error bars represent standard deviations from 8 replicates. Difference in means at P<0.05
is indicated by different letters under P deficiency.

3.4. Nodulated-root respiration

The rhizospheric concentration of O2 (pO2) in contact with the nodulated-root was

significantly related to nodulated root O2 consumption (Conr) for landrace 11 (P = 0.038), for
cultivar 1 (P = 0.040); for landrace 7 (P = 0.012); and for cultivar 4 (P = 0.002) (Fig. 4).
Indeed, the increase of pO2 induced an increase of Conr to a maximum value corresponding to
the critical oxygen pressure (pO2 = 40%). The Conr at ambient rhizospheric pO2 (21% O2)
showed that roots and nodules respiration for the plants growth and maintenance under P

179
deficiency was 76% for cultivar 1, 30% for landrace 11 and 11% for landrace 7 of that under
P sufficiency. However, Conr values for the cultivar 4 under P deficiency was 182% of that
under P sufficiency (Fig. 4). The same observation was made on Conr at 40% O2 which
revealed the maximum respiration of the plants under exposure to a critical pO2.

FIG. 3. Efficiency in use of rhizobial symbiosis of 10 cultivars under P sufficiency (white circle) versus
P deficiency (black circle). Data are individual values for 10 voandzou cultivars and landraces
harvested at 6 weeks after transplanting. Symbols *, **, *** indicate P<0.05, P<0.01 and P<0.001,
respectively.

The general tendency of the regression curves reported the superiority of tolerant
landrace 7 and cultivar 4 under P deficiency compared to P sufficiency in contrast with the
sensitive landrace 11 and cultivar 1. Under P deficiency, the slope reflecting the nodule
permeability was higher for landrace 7 than for cultivar 4 (Fig. 4). Nodule permeability to O2
was calculated by dividing the slope of regression by nodule areas as previously described
[14, 16, 24, 26]. Nodule permeability under P deficiency was similar for cultivar 4 and
landrace 7, with 0.60 and 0.46 µm s-1, respectively, which were 4 times higher than under P
sufficiency.

3.5. Microscopic analysis of nodules

Under P deficiency, the cortical cells of “inner cortex” were larger in landrace 7 and
cultivar 4 with marked increase in cell size and intercellular space. The superiority of Conr

180
under P sufficiency for cultivar 1 could be explained by a greater elongation of intercellular
spaces in the “inner cortex” compared with P deficiency (Fig. 4).

FIG. 4. Nodulated root O2 consumption to increasing rhizospheric pO2 for 4 cultivars under P
sufficiency (white circle) versus P deficiency (black circle). Data are means of 3 replicates between
the 40th and 44th day after transplanting in serum bottles. Errors bars represents the standard error of
the mean. Significant correlation at P<0.05 and P<0.01 are indicated by symbols * and **,
respectively.

More expression of the phytase gene, as bright green points, especially around the
vascular trace, the “inner cortex” and the infected zone was observed in particular for landrace
7 under P deficiency (Fig. 5). An increase in phytase gene expression was also observed on
intercellular spaces for cultivar 4 under P deficiency. The expression of the phytase gene in
the nodule cross-section could explain the cultivar tolerance to P deficiency through the
biomass results.

4. DISCUSSION

This study highlights contrasting lines of voandzou for SNF and N2-dependent growth
under P deficiency, in agreement with Greder et al. who concluded that genotypic selection on
the basis of nodule weight is justified [27]. Thus, the selection of 58 cultivars and landraces of
Madagascar, Niger and Mali allowed the prescreening of cultivars and landraces with
nodulation above 0.2 g nodule DW plant-1 and shoot biomass above 2 g DW shoot plant-1

181
under P deficiency. The lower nodules and shoot biomass of cultivars and landraces of Niger
and Mali compared with those of Madagascar (Fig. 2) suggest that the genetic potentiality of
the different voandzou selected in terms of biomass differ according to the origin of seeds in
different agro-ecological regions, whose development is influenced by environmental factors,
in particular temperature and photoperiod [28]. It confirms indeed that a great inter-landrace
variation exists in the genetic resources of voandzou [29]. Numerous authors emphasized the
variability of legume SNF experimental conditions, which reflect the genetic ability of the
different legume species for N2 fixation and yield under environmental constraints [10]. Thus,
the critical P requirement of voandzou (Fig. 1) is lower than that of other legume species like
Glycine max [15] and Phaseolus vulgaris [26, 30], but higher than that of Vigna unguiculata
[14] and Acacia mangium [31].

FIG. 5. Cross-section of a voandzou nodule. MC: Middle cortex, IC: Inner cortex, VT: Vascular trace,
Py: Gene phytase expression, IZ: Infected zone. Scale bar: 500µm. Under P deficiency, the tolerant
cultivars (7 and V4) were marked by an increase of cell size and intercellular space coupled with
more gene phytase expression. 7+P: Landrace 7 under P sufficiency; 7-P: Landrace 7 under P
deficiency; V4+P: Cultivar 4 under P sufficiency; V4-P: Cultivar 4 under P deficiency.

The inoculation of voandzou with Bradyrhizobium sp. Vigna CB756, initially

developed for Vigna unguiculata cultivars, showed a great expression of genetic potential in
the P use for nodulation and biomass, in particular cultivar 1 with a great response of P
compared with cultivar 6. This genotypic variation could be attributed to the effectiveness of
the rhizobia-legume symbiosis where some genotypes have selectivity for a highly effective
strain selected for high nodulation and consequently for high productivity [10]. Kishinevsky
et al. reported from their experiments with 20 landraces of voandzou, that the plant genotype
and the Bradyrhizobium strain are important factors to be considered in a breeding program
for a high N2 fixation rate [5].

As a result of this pre-screening, ten contrasting lines were identified, including five
tolerant lines in terms of shoot, root, and nodule biomass under P deficiency, and five
sensitive in terms of the productivity decrease by P deficiency (Fig. 2). The most contrasting
lines confirmed their high SNF potential by high SNF in terms of nodules and shoot biomass,
in particular landrace 7, 11, cultivars V4 and V1. This great potential was highlighted by a
better EURS under P deficiency for such lines as landrace 7, 2 and cultivar V1 (Fig. 3).

182
Furthermore, EURS under P deficiency was much higher than under P sufficiency for
landraces 7 and 2, illustrating the tolerance of these lines to P deficiency. These results are in
agreement with those obtained on contrasting lines of bean [14] and Vigna unguiculata [26].

The increase of O2 consumption of nodulated root (Conr) in response to the variation

of rhizospheric O2 concentration for all the lines reveals an important energy respiratory
requirement for SNF. Under ambient rhizospheric condition, in particular under pO2 of 21%,
the root and nodule respiration for plant growth and maintenance is limited by P deficiency in
comparison with P sufficient, specifically for the sensitive cultivars V1 and 11. In addition, a
4 fold increase in nodule respiration, with a value 4 times higher in nodule permeability under
P deficiency compared with P sufficiency, is in agreement with previous results for soybean,
common-bean and [14–16]. Nodule permeability controls nodule respiration which provides
ATP for the N2 reduction in the infected zones of the nodule [13, 16, 26]. An alternative
respiratory coupled with a large O2 consumption by ATP produced under P deficiency can
explain this increase in nodule permeability [15, 16, 24]. Thus, limitation of nodulated root
respiration under ambient pO2 was higher for the sensitive cultivars V1 and 11 than for the
tolerant ones (Fig. 4).

The image analysis of histological observations on nodule cross-sections under P

deficiency (Fig. 4) confirms that the cortical cells of the “inner cortex” increase in size when
nodule permeability is increased, in agreement with previous observations of increases in cell
size and intercellular space of the inner cortex in soybean, cowpea, and alfafa, and the nodule
permeability to O2 diffusion in soybean [15, 32, 33]. The orthophosphate ions could be
implicated in the osmotic pressure regulation in the “inner cortex” cells [15]. In situ RT-PCR
image analysis of the nodule cross-section showed a larger increase in the expression of the
phytase gene in nodules under P deficiency, and specifically around the vascular trace, in the
“inner cortex” and the infected zone for landrace 7. The nodulated root phytase activity in
terms of the expression of acid phosphatase is in accordance with previous findings [4] for the
lines of bean contrasting in their PUE for SNF, where the expression of acid phosphatase
varied among the tissues, and was higher in the cells of the “inner cortex”. The phytase gene
expression is linked with the change of nodule permeability associated with the adaptation to
P deficiency, specifically for Landrace 7 and cultivar 4, where the nodule permeability to O2
was increased 4 times under P deficiency, suggesting a link, either direct or indirect, between
phytase activity and the regulation of the respiration linked to SNF by nodule permeability.

5. CONCLUSION

Genotypic variation of the SNF for the PUE was observed and highlighted in the 54
cultivars and landraces from Madagascar, Niger and Mali. Variability was especially related
to the nodule and shoot biomass, nodule permeability, nodule respiration and gene phytase
expression. These experiments made it possible to select the more contrasting cultivars and
landraces in terms of PUE for SNF, in particular the most tolerant ones such as landrace 7,
cultivar 4, and the most sensitive ones such as cultivar 1, to be further compared under field
conditions.

REFERENCES

[1] BARRON, E.J., et al., Response to selection for seed yield and nitrogen fixation in
common bean (Phaseolus vulgaris L.), Field Crops Res. 62 (1999) 119–128.
[2] GRAHAM, P.H., HUNGRIA, M., TLUSTY, B., Breeding for better nitrogen fixation
in grain legumes: Where do the rhizobia fit in?, Crop Manage. (2004)
doi:10.1094/CM-2004-0301-02-RV

183
[3] DAKORA, F.D., Nodule function in symbiotic Bambara groundnut (Vigna
subterranea L.) and Kersting's bean (Macrotyloma geocarpum L.) is tolerant of nitrate
in the root medium, Ann. Bot. 82 (1998) 687–690.
[4] DREVON, J.J., et al., "Do nodule phosphatase and phytase link with the phosphorus
use efficiency for N2-dependent growth in Phaseolus vulgaris?", Proc. 15th Int.
Nitrogen Fixation Congr. and the 12th Int. Conf. of the African Association for
Biological Nitrogen Fixation, (DAKORA, F.D.C., et al., Eds), Springer (2008) 93–94.
[5] KISHINEVSKY, B.D., et al., Variation in nitrogen fixation and yield in landraces of
voandzou (Vigna subterranea), Field Crops Res. 48 (1996) 57–64.
[6] KUMAGA, E.D., DANSO, S.K.A., ZAPATA, F., Time-course of nitrogen fixation in
two bambara groundnut (Vigna subterranea L. Verdc.) cultivars, Biol. Fertil. Soils 18
(1994) 231–236.
[7] LINDEMANN, W.C., GLOVER, C.R., Nitrogen Fixation by Legumes, Guide A-129,
Co-operative Extension Service, College of Agriculture and Home Economics, New
Mexico State University (2003) 1–4.
[8] MASSAWE, F.J., et al., Breeding in Bambara groundnut (Vigna subterranea L.)
Verdc.): strategic considerations, Afr. J. Biotechnol. 4 (2005) 463–471.
[9] RICHARDSON, A.E., et al., Acquisition of phosphorus and nitrogen in the
rhizosphere and plant growth promotion by microorganisms, Plant Soil 321 (2009)
305–339.
[10] HERRIDGE, D., ROSE, I., Breeding for enhanced nitrogen fixation in crop legumes,
Field Crops Res. 65 (2000) 229–248.
[11] BOURION, V., et al., Genetic variability in nodulation and root growth affects
nitrogen fixation and accumulation in pea, Ann. Bot. 100 (2007) 589–598.
[12] CHRISTIANSEN, I., GRAHAM, P.H., Variation in di-nitrogen fixation among
Andean bean (Phaseolus vulgaris L.) genotypes grown at low and high levels of
phosphorus supply, Field Cops Res. 73 (2002) 133–142.
[13] VADEZ, V., et al., Variability of N2 fixation in common bean (Phaseolus vulgaris L.)
under P deficiency is related to P use efficiency, Euphytica 106 (1999) 231–242.
[14] ALKAMA, N., et al., Genotypic variability in P use efficiency for symbiotic nitrogen
fixation is associated with variation of proton efflux in cowpea rhizosphere, Soil Biol.
Biochem. 41 (2009) 1814–1823.
[15] RIBET, J., DREVON, J.J. (1995) Increase in permeability to oxygen diffusion and
oxygen uptake of soybean nodules under limiting phosphorus nutrition, Physiol. Plant.
94 (1995) 298–304.
[16] VADEZ, V., et al., Nodule permeability to O2 and nitrogenase-linked respiration in
bean genotypes varying in the tolerance of N2 fixation to P deficiency, Plant Physiol.
Biochem. 34 (1996) 871–878.
[17] HELLER, J., BEGEMANN, F., MUSHONGA, J., Eds., Bambara Groundnut (Vigna
subterranea (L.) Verdc.), Promoting the Conservation and Use of Underutilized and
Neglected crops 9, (HELLER, J., BEGEMANN, F., MUSHONGA, J., Eds),
International Plant Genetic Resources Institute, Rome, Italy (1997) 166 p.
[18] DAKORA, F.D., MUOFHE, L.M., "Nitrogen fixation and nitrogen nutrition in
symbiotic voandzou (Vigna suterranea L. Verdc.) and Kersting’s bean (Macrotyloma
geocarpum (Harms) Maréch. Et Baud)", Bambara Groundnut (Vigna subterranea (L.)
Verdc.), Promoting the Conservation and Use of Underutilized and Neglected crops 9,
(HELLER, J., BEGEMANN, F., MUSHONGA, J., Eds), International Plant Genetic
Resources Institute, Rome, Italy (1997) 72–77.
[19] DRABO, I., SÉRÉMÉ, P., DABIRE, C., "Country reports, Institut d’Etudes et de
Recherches Agricoles (INERA)", Bambara Groundnut (Vigna subterranea (L.)
Verdc.), Promoting the Conservation and Use of Underutilized and Neglected crops 9,

184
 (HELLER, J., BEGEMANN, F., MUSHONGA, J., Eds), International Plant Genetic
Resources Institute, Rome, Italy (1997) 19–26.
[20] GOLI, A.E., "Bibliographical review", Bambara Groundnut (Vigna subterranea (L.)
Verdc.), Promoting the Conservation and Use of Underutilized and Neglected crops 9,
(HELLER, J., BEGEMANN, F., MUSHONGA, J., Eds), International Plant Genetic
Resources Institute, Rome, Italy (1997) 4–10.
[21] NGUY-NTAMAG, F.C., "Country reports", Bambara Groundnut (Vigna subterranea
(L.) Verdc.), Promoting the Conservation and Use of Underutilized and Neglected
crops 9, (HELLER, J., BEGEMANN, F., MUSHONGA, J., Eds), International Plant
Genetic Resources Institute, Rome, Italy (1997) 27–29.
[22] KARIKARI, S.K., et al., "Country reports", Bambara Groundnut (Vigna subterranea
(L.) Verdc.), Promoting the Conservation and Use of Underutilized and Neglected
crops 9, (HELLER, J., BEGEMANN, F., MUSHONGA, J., Eds), International Plant
Genetic Resources Institute, Rome, Italy (1997) 11–18.
[23] VINCENT, J.M., A Manual for the Practical Study of Root-nodule Bacteria,
Blackwell Scientific Publications, Oxford (1970) 164 p.
[24] JEBARA, M., DREVON, J.J., Genotypic variation in nodule conductance to the
oxygen diffusion in common bean (Phaseolus vulgaris), Agronomie 21 (2001) 667–
674.
[25] SCHULZE, J., DREVON, J.J., P-deficiency increases the O2 uptake per N2 reduced in
alfalfa, J. Exp. Bot. 56 (2005) 1779–1784.
[26] KOUAS, S., et al., Proton release by nodulated roots varies among common bean
genotypes (Phaseolus vulgaris) under phosphorus deficiency, J. Plant Nutr. Soil Sci.
171 (2008) 242–248.
[27] GREDER, R.R., ORF, J.H., LAMBERT, J.W., Heritabilities and associations of nodule
mass and recovery of Bradyrhizobium japonicum serogroup USDA 110 in soybean,
Crop Sci. 26 (1986) 33–37.
[28] BRINK, M., Rates of progress towards flowering and podding in Bambara groundnut
(Vigna subterranea) as a function of temperature and photoperiod, Ann. Bot. 80 (1997)
505–513.
[29] JORGENSEN, S.T., et al., "Identification of photoperiod neutral lines of bambara
groundnut (Vigna subterranea) from Tanzania", Climate Change: Global Risks,
Challenges and Decisions, IOP Publishing (2009) .
[30] TANG, L., et al., Effect of phosphorus deficiency on the growth, symbiotic N2 fixation
and proton release by two bean (Phaseolus vulgaris) genotypes, Agronomie 21 (2001)
683–689.
[31] RIBET, J., DREVON, J.J., The phosphorus requirement of N2-fixation and urea-fed
Acacia mangium, New Phytol. 132 (1996) 383–390.
[32] FLEURAT-LESSARD, P., et al., The distribution of aquaporin subtypes (PIP1, PIP2
and g-TIP) is tissue dependent in soybean (Glycine max) root nodules, Ann. Bot. 96
(2005) 457–460.
[33] SERRAJ, R., et al., Structural changes in the inner-cortex cells of soybean root
nodules are induced by short-term exposure to high salt or oxygen concentrations,
Plant Cell and Environ. 18 (1995) 455–462.

185
 GENOTYPIC VARIATION IN PHOSPHORUS USE EFFICIENCY FOR
SYMBIOTIC NITROGEN FIXATION IN COWPEA (VIGNA UNGUICULATA)

A. ANDRIAMANANJARA
LRI-SRA, Laboratoire des Radio-isotopes,
Université d'Antananarivo,
Antananarivo,
Madagascar

M. MALAM ABDOU
Laboratoire Banques de gènes CERRA / KOLLO,
Institut National de Recherche Agronomique du Niger (INRAN),
Niamey,
Niger

C. PERNOT, J.J. DREVON

Institut National de la Recherche Agronomique,
UMR Eco&Sols,
Montpellier,
France.

Abstract

Cowpea (Vigna unguiculata L. Walp) is an important food legume. In Africa, it is mostly

cultivated under such environmental constraints as drought and pest, and nutrient deficiency. In
particular low soil phosphorus strongly limits crop production for the poor farmers with limited access
to P fertilizers. Therefore breeding cowpea for the tolerance to P deficiency is considered as an
alternative to increase the productivity of traditional cowpea-cereal cropping systems in soils with low
P availability. This paper reports cowpea genotypic-variation in P use efficiency for symbiotic
nitrogen fixation as a contribution to select tolerant cowpea lines under P deficiency. Eighty cowpea
cultivars inoculated with the reference strain of Bradyrhizobium sp. Vigna CB756 were pre-screened
as a single replicate under hydroaeroponic culture for 6 weeks under P deficiency versus P sufficiency,
namely 15 vs 30 µmol plant-1 week-1. Large variability in nodule number per plant, and in shoot
growth as a function of nodule mass, was observed among the diversity of cowpea lines. From this
pre-screening experiment, the 40 cowpea lines showing the highest SNF-potential, i.e. high nodulation
linked with high N2-dependent growth under P sufficiency, and the most contrasting tolerance to P
deficiency, i.e. highest vs lowest N2-dependent growth under P deficiency, were grown again in
glasshouse hydroaeroponics with 6 replicates. As an illustration of the most contrasting lines, the
nodulation was decreased under P deficiency by less than 20% for IT82E-18 whereas by more than 80%
for IT95K-1105-5 or SUVITA 2. The variations in nodulation were correlated with variations in
growth with mean value of additional growth per unit increase in nodule biomass of 23 g shoot DW g-1
nodule DW under P sufficiency, showing 3 lines showing exceptionally high potential for symbiotic
nitrogen fixation, versus 28 g shoot DW g-1 nodule DW showing large variation among lines. Most of
the tolerance to P deficiency was due to increase in the nodule function rather than conservation of
nodule mass. In conclusion, the screening in glasshouse hydroaeroponics made it possible to sort
cowpea lines in SNF potential to fix N2, and in tolerance to P deficiency.

1. INTRODUCTION

Cowpea (Vigna unguiculata L. Walp) is an important food legume and mostly

cultivated in tropical savannas of African semi-arid and arid areas [1]. It is an essential
component of cropping systems covering over 9.3 million ha making this legume an
important source of quality nourishment to the urban and rural populations in various

187
developing country [2]. However, most African soils have low inherent levels of available P.
Soil-P availability during plant development is an important determinant of growth, symbiotic
nitrogen fixation (SNF) and grain yield of legumes [3]. Plants adapt to this environmental
constraint through various morphological and physiological mechanisms as root hair
development, higher root surface to increase P acquisition efficiency, and exudation of
phosphatase and organic acids to increase P availability [4, 5].

Current research on cowpea breeding has concentrated on development of improved

cowpea breeding lines, combining disease and insect resistance with quantitative traits such as
higher grain yield, and with tolerance to extreme temperature and adaptation to drought [6, 7].
Nevertheless, the contribution of cowpea to soil fertility through SNF under such an abiotic
constraint as P deficiency prompted variety screening under P deficiency for SNF [7, 8].
Contrasting cowpea genotypes for tolerance to P deficiency was considered as an alternative
for the soil with low P availability, which strongly limits crop production for the poor farmers
with limited accessibility to P fertilizers [8, 9]. Indeed, growth, nodulation and N2-fixation of
cowpea could be inhibited by P deficiency, whereas tolerant genotypes can access P which is
not usually available to the average genotypes under low P conditions. However, few
publications describe the physiological mechanisms by which cowpea adapts to P deficiency.
The optimization of SNF for P use efficiency (PUE) in legume species were already
undertaken for genotypic variation by numerous researchers [1, 2, 8]. Genotypic variability of
cowpea in PUE for symbiotic nitrogen fixation has been found to be associated with such
rhizospheric functions as nodulated-root proton efflux and respiration under low-P soils [1].
The contrasting genotypes increased the soil P availability by about 50% after a cultural cycle.
Thus, cowpea breeding could significantly affect the productivity of traditional cowpea-cereal
cropping systems. The present study aimed at screening a diversity of cowpea genotypes in
PUE for SNF under hydroaeroponic culture.

2. MATERIALS AND METHODS

2.1. Biological materials

The seeds of cowpea 24-125B-1, 524B, 58-53, 58-57, APAGBAALA, BAMBEY 21,
CB27, CB46, CC-27, UCR CC-36, DANILA, IAR7/8-5-4-1, IFE BROWN, IRON CLAY,
IT82E-18, IT84S-2049, IT84S-2246, IT85F-3139, IT85F-867-5, IT90K-284-2, IT93K-2046-
1, IT93K-503-1, IT93K-693-2, IT97K-437-1, IT95K-1105-5, IT95K-1479, IT95K-1491,
IT95M-190, IT95M-303, IT96D-602, IT96D-610, IT97K-207-15, IT97K-461-4-1, IT97K-
499-35-1, IT97K-499-39, IT97K-556-6, IT97K-569-9, IT97K-819-132, IT98D-1399, IT98K-
128-2, IT98K-205-8, IT98K-205-8, IT98K-498-1, IT98K-555-1, IT98K-558-1, IT98K-589-2,
IT99K-407-8, IT89KD-288, KVx-396-4-5-1, KVx403, KVx421-25, KVx525, KVx544-6-
151-3, KVx61-1, MELAKH, Mounge, MOURIDE, NDIAMBOUR, PRIMA, Sasaque,
SUVITA 2, TVu-7778, UCR Sh50-17-9-1, UCRP-24, UCR 779, YACINE, were kindly
supplied by Jeff Ehlers from the University of California (USA) where the lines were selected
for adaptation to water deficiency for the Sahelo-Sudanian pedo-climatic conditions. The
additional lines DD 07, DMI 07, IT90K-372-1-2, KA85-19, KB85-23, KVx30-309-6G,
TN121-80, TN256-87, TN259-87, TN27-80, TN3-78, TN5-78, TN88-63, and BAMBEY,
Mouride, were supplied by INRAN (Niger) and ISRA (Senegal).

188
2.2. Symbiosis culture

The seeds were sterilized with 3% calcium hypochlorite for 20 min and rinsed by 5
washings with sterile distilled water. They were then transferred for germination on
humidified wrapped filter paper in a slightly tilted vat and placed in an incubator at 28-30°C.

After germination, the inoculation was performed by soaking the roots of seedlings for
20 min in a suspension of Bradyrhizobium sp. Vigna CB756 containing 109 bacteria ml-1. The
inoculum was prepared from a rhizobia culture preserved in tubes at 4°C on 120°C-sterilized
agar YEM (Yeast Extract Mannitol) medium containing 900 ml distilled water, 100 ml of
Bergersen concentrated solution consisting of a mixture of 1 g KCl, 0.1 g FeCl3, 0.4 g CaCl2,
4.5 g Na2HPO4.12H2O and 1 g MgSO4.7H2O, 1 g Yeast extract, 10 g mannitol and 15 g agar
[10]. From one of the preserved tubes, some strains were taken and transferred into 100 ml of
liquid YEM, and maintained under agitation at 28°C for 24 h.

The germinated seeds were carefully transferred for hydroaeroponic culture into each
40 l vat with 20 inoculated plants per vat. The roots of each seed were carefully passed
through the hole of a rubber stopper with cotton wool fixed at the hypocotyl level. They
received the following nutrient solution that was changed every 2 weeks: CaCl2, 1650 µM;
MgSO4.7H2O, 1000 µM; K2SO4, 700 µM; Fe EDDHA, 8.5 µM as sequestrene Fe; H3BO3, 4
µM; MnSO4.H2O, 6 µM; ZnSO4.7H2O, 1 µM; CuSO4.7H2O, 1 µM; Na2MoO4.7H2O, 0.1 µM.
2000 µM of urea was supplied for all treatments during the first two weeks’ culture. Each P
was supplied in the form of KH2PO4 with an exponential distribution for the two levels: 15
and 30 µmol plant-1 week-1 as P deficiency and P sufficiency, respectively.

The experiment was carried out in a glasshouse at 20/33°C temperature during 16/8 h
day/night cycle coupled with an intense ventilation of nodulated-roots, a complementary
illumination of 400 µmol photons m-2 s-1 when needed, and 70% daily relative humidity.
These conditions allowed an optimal expression of genetic potential of nitrogen fixation while
controlling closely the mineral nutrition mainly for N, P and for the other elements.

2.3. Biomass measurements and statistical analysis

The plants were harvested at the end of the 6th week at the flowering stage. The shoot was
separated from the root at the cotyledonary node, and then weighted after 48 h at 80°C.
Nodules were excised from roots, counted and weighted. The effect of P supply was estimated
from the response curves of the biomass, and the critical P supply was deduced at the
maximum growth. In order to determine the contribution of P supply to plant growth, the
response curves were established from biomass values with the software Excel Microsoft
Office XP. The genotype and P treatment effects were analyzed by a two-way analysis of
variance (ANOVA). Correlation and regression analysis were performed to complete the
analysis.

3. RESULTS

3.1. Pre-screening for nodulation under P sufficiency versus P deficiency

Large variability in nodule number per plant was observed among the diversity of
cowpea line used as a single replicate in the pre-screening experiment (Fig. 1). Under P
sufficiency the highest number of nodules was more than 700 plant -1, e.g. IT90K-284-2,
whereas the lowest nodulation was 10 plant -1, e.g. Mouride, 58-53, IT97K569-9, IT97K-461-
4-1 and KA85-19. The mean nodulation was 300 plant -1 for one third of the lines, and below

189
100 plant-1 for another third of the lines. Under P deficiency, the highest number of nodules
was between 300 and 500 plant-1 for YACINE, NDIAMBOU, IRON CLAY, IFE BROWN,
IAR7/8-5-4-1, IT99K-407-8, IT98K-555-1 and IT97K-819-132, whereas more than two thirds
of the lines had less than 200 nod plant -1.

FIG. 1. Variation in nodule number per plant among cowpea lines under P sufficiency (■) versus P
deficiency (■). Data are single values per line.

190
 The nodule mass of the lines as a function of nodule number is shown in Fig. 2. The
significant relation between both parameters established that the mean individual-nodule mass
was 0.3 mg nod-1 under P sufficiency and 0.6 mg nod-1 under P deficiency. However the
biggest nodule reached almost 2 mg nod-1 whatever the P supply, although bigger nodules
were more frequent among lines with lower nodule number.

FIG. 2. Relation of nodule growth as a function of nodule number under P sufficiency (A) versus P
deficiency (B) for cowpea lines shown in Fig. 1. Data are single values per line.

The shoot growth of the lines as a function of nodule mass is shown in Fig. 3. The
significant relation between both parameters established that the mean growth per nodule
mass was around 5 g shoot DW g-1 nod DW without any significant difference between P
supply, although the growth without nodulation, i.e. dependent on seed N and starter N, was
twice as high under P sufficiency (0.6 g DW shoot plant -1) than under P deficiency (0.3 g DW
shoot plant -1). However, the largest N2-dependent growth was close to 10 and 15 g shoot DW
g-1 nod DW, under P sufficiency and P deficiency, respectively, among lines harboring more
than 0.2 g DW nod plant-1. The shoot growth was also significantly correlated with root
growth with a mean shoot: root ratio of 2.2, whatever the P supply. By contrast, there was no
significant correlation between root growth and nodule mass or number (data not shown). The
distribution of biomass among the three organs for the diversity of the lines tested in the pre-
screening is shown in Fig. 4.

191
3.2. Relation between root nodulation and plant growth among most contrasting lines

From the data in Figs. 2 and 3 lines were chosen having the maximal SNF-potential,
i.e. high nodulation linked with high N2-dependent growth under P sufficiency, and the most
contrasting tolerance to P deficiency, i.e. highest or lowest N2-dependent growth under P
deficiency. They were grown again in glasshouse hydroaeroponics with 6 replicates.

FIG. 3. Relation of shoot growth as a function of nodule number for cowpea lines under P sufficiency
(A) versus P deficiency (B). Data are single values per line.

Lines having more than 400 mg DW nod plant-1 under P sufficiency are shown in Fig.
5. Among these lines, IT82E-18 show the highest nodulation under P deficiency, by contrast
with such lines as IT95K-1105-5 or SUVITA 2 where nodulation decreased by more than 80%
under P deficiency. The variation in nodulation, as shown in Fig. 5, were correlated with
variation in growth, as illustrated in Fig. 6, showing that total biomass in shoots varied from
more than 12 g DW plant-1 to almost 2 g DW plant-1 under P sufficiency, and from more than

192
8 g DW plant-1 to less than 2 g DW plant-1 under P deficiency. Three lines were found to have
high nodulation with an exceptionally high potential for symbiotic nitrogen fixation,
supporting a growth of more than 10 g DW shoot plant -1.

FIG. 4. Variation in total growth and distribution in shoot (■), root (■) and nodule (■) per plant
among cowpea lines under P sufficiency (left) versus P deficiency (right). Data are single values per
line.

Under P sufficiency, the relation between both parameters was linear between 100 and
330 mg DW nod plant-1. The mean value of additional growth per unit increase in nodule,
namely the efficiency in use of the rhizobial symbiosis (EURS) was 23 g DW shoot g-1 DW

193
nod. Under P deficiency, the mean EURS was 28 g DW shoot g-1 DW nod, i.e. slightly and
significantly higher than under P sufficiency. In addition, there was more variation in EURS
for the lines under P deficiency than under P sufficiency. The two lines showing the highest
growth under P deficiency were TN259-87, IT82E-18 having higher EURS than the other
lines. Three other lines tolerant to P deficiency, TN5-78, UCR Sh50-17-9-1 and IRON CLAY,
also showed higher EURS than the rest of the lines. Thus most of the tolerance was due to
increase in the nodule function rather than conservation of nodule mass under P deficiency.
The two most sensitive lines to P deficiency were IT84S-2049 and IT95K-1105-5, combining
low nodulation with low EUSR, whereas the sensitivity of three other lines, IT97K-499-35-1,
NDIAMBOUR and SUVITA 2, was due to low EURS.

4. DISCUSSION

Eighty-one cowpea lines were screened under hydroaeroponic culture for adaptation to
P deficiency and phosphorus use efficiency (PUE) for symbiotic nitrogen efficiency (SNF).
The pre-screening measurement of shoot, root and nodule biomass under P sufficiency versus
P deficiency (Fig. 4) indicated the existence of useful genetic variation among cultivars for
PUE: SNF as confirmed by the subsequent experiment with 6 replicates for the most
contrasting genotypes (Fig. 5). Nodulation measurements revealed 21 cowpea lines with high
nodulation under P sufficiency, namely ITKOK-284-2, IAR7/8-5-4-1, CB46, IT85F-867-5,
IFE BROWN, IT97K-437-1, IT84S-2246, IRON CLAY, IT85F-3-139, YACINE, UCR 779,
TN256-87, UCR Sh50-17-9-1, IT97K-556-6, KVx544-6151-3, IT82E-18, MOURIDE,
IT93K-503-1, IT98K-555-1, IT95K-1-105-5, IT99K-407-8, while 14 lines were highlighted
under P deficiency with more than 150 mg dry weight nodule per plant, namely ITKOK-284-
2, IAR7/8-5-4-1, IFE BROWN, IRON CLAY, IT85F-3-139, YACINE, IT97K-556-6,
KVx544-6151-3, IT98K-555-1, IT99K-407-8, IT97K-819-132, IT97K-207-15, MELAKH
and NDIAMBOUR (Fig. 2). These nodulation results were higher than those found bunder P
deficiency [1], but lower than those found under P sufficiency [1, 11]. These discrepancies
may be explained by the temperature and the natural illumination of the glasshouse that might
have been lower in this work than during the previous experiments.

194
FIG. 5. Variation in nodule DW per plant under P sufficiency (■) versus P deficiency (■) among the
most contrasting cowpea lines for tolerance to P deficiency (Fig. 1). Data are means and bars
represent standard deviations of 6 replicates for each line.

195
FIG. 6. Relation of shoot growth as a function of nodule mass for cowpea lines under P sufficiency (A)
versus P deficiency (B) shown in Fig. 5.

Furthermore, plant biomass is an important indicator used in many allometric

relationships. Thus variations in biomass production are used as a selection criterion for
genotype assessment of nutrient efficiency at the seedling stage [12, 13]. Shoot biomass data
under P deficiency (Figs. 4 and 6) are in agreement with that observed previously [1], but the
values under P sufficiency were mostly higher than those found previously on cowpea lines [1,
14]. Previous studies reported that the low efficiency, shown by such lines as Melakh or
Bambey-21, could be explained by early maturity with more effects of environmental
constraints during vegetative- than reproductive- stages [15]. Gerloff reported that the
intolerance to P deficiency is often due to inefficient P-acquisition mechanism [16]. Studies

196
on pigeonpea reported that genotypic difference in response to P deficiency may be related to
the magnitude of resistance of the cultivar to a decrease in shoot growth, as the P deficiency
affects leaf initiation and photosynthate per unit leaf area [17]. Tolerant hybrid-cultivars with
low restriction of leaf number and leaf area by P deficiency, could partition more assimilates
for shoot growth than sensitive cultivars where partitioning to roots was important for
survival [17]. Efficient lines such as TN259-87, IT82E-18, TN5-78, UCR Sh50-17-9-1, IRON
CLAY, UCR 779, IT97K-819-132, PRIMA, 524B, IT98K-498-1, IT98K-555-1 and DANILA
with more than 150 mg nodule dry weight per plant, and with more than 4 g dry weight of
shoot per plant (Figs. 5 and 6), were less inhibited by P deficiency. Furthermore, responsive
lines which gave a higher plant biomass than other lines under high P sufficiency were also
highlighted as UCR Sh50-17-9-1, IT95K-1105-5, IT82E-18, IRON CLAY, UCR 779, IT90K-
284-2, IT97K-819-132, BAMBEY-21, SUVITA 2, IT84S-2246 and TN 256-87, with more
than 225 mg nodule DW plant-1, and with more than 5 g shoot DW plant -1. These
physiological features were considered as the most appropriate for an effective screening
procedure to detect intra-specific differences [16].

The shoot biomass correlation with the nodule mass (Figs. 3 and 6) made it possible to
define the efficiency in use of the rhizobial symbiosis (EURS), as estimated by the slope of
regression, as in previous work with other legume species [1]. The high EURS under P
deficiency (Fig. 6) confirms the results from pre-screening (Fig. 3). It suggests that internal
remobilization of acquired P through SNF ability may help the tolerant cultivars to establish
higher shoot biomass under P-deficiency. It highlights tolerant lines by high individual nodule
mass that combines with high EUSR under P deficiency compared to that under P sufficiency.
This agrees with the tolerance of nodulation to P deficiency that was attributed to a lower
immobilization of P in nodule structure under P deficiency than under P sufficiency [18, 19],
although greater P concentration was found in nodules than other tissues such as shoot and
root.

Under P-limiting conditions, N2-fixing legumes adopt some strategies to enhance N

and P acquisition, involving: (i) root morphology by decreased growth rate and increased
growth per unit of P uptake, and also by root development; (ii) remobilization of internal
inorganic P at the cellular level, (iii) alternative respiratory pathways, (iv) secretion of acid
phosphatase and organic acids [20–23]. Efficient N2-fixing legumes are capable of
maintaining a high metabolic activity in their root nodules under P deficiency as a mechanism
of tolerance [24]. Increase in the acid phosphatase and phytase activities in nodules of
Phaseolus vulgaris under P deficiency suggested the role of these enzymes in adaptive
mechanisms for N2-dependent legumes [25]. The up-regulation of acid phosphatases is mainly
the means for releasing and recycling inorganic P from both internal and external resources as
phytate [26]. Bio-physiological mechanisms to adjust nodule metabolism as a response to P
deficiency are developed by legumes such as increased nodule O2 permeability [19, 27] and
plasma membrane intrinsic proteins and tonoplast intrisinc protein aquaporins in the inner
cortex cells [28], involving changes in shape and volume of inner-cortex cells in response to
stimuli affecting nodule permeability [29].

Also, such nodule enzyme activities as phosphoenolpyruvate carboxylase and malate

dehydrogenase, involved in C cycling and energy substrates for N2 fixation, and phytase,
presumably involved in remobilizing P, were suggested as indicators of genotypic variation in
legume species such as P. vulgaris [30], Medicago sativa [19] and Vigna subterranea [31].
Many studies reported that species and cultivars differ in their ability to take up specific
elements, including P, due to several morphological, physiological, and biochemical
mechanisms [32, 33]. Advanced studies of Arabidopsis, common bean (P. vulgaris), and

197
white lupin revealed genetic and biochemical factors as a result of gene expression changes
that mediate plant adaptations to P deficiency [22]. Lastly, P nutrition can also affect legume
SNF by specific effects on rhizobial growth and on nodule formation and function [34].

5. CONCLUSION

The screening in glasshouse hydroaeroponic culture made it possible to sort cowpea

lines with high SNF potential versus significantly much lower potential to fix N2, and with
tolerance vs sensitivity to P deficiency. Thus contrasting lines of cowpea in PUE for SNF
such as UCR Sh50-17-9-1, IT82E-18, IRON CLAY, UCR 779 and IT97K-819-132, are now
available to test in farmers' fields whether they can be proposed as an alternative technology
in low P soils.

REFERENCES

[1] ALKAMA, N., et al., Genotypic variability in P use efficiency for symbiotic nitrogen
fixation is associated with variation of proton efflux in cowpea rhizosphere, Soil Biol.
Biochem. 41 (2009) 1814–1823.
[2] KOLAWOLE, G.O., TIAN, G., SINGH, B.B., "Differential response of cowpea lines
to application of P fertilizer", Challenges and Opportunities for Enhancing Sustainable
Cowpea Production, (FATOKUN, C.A., et al., Eds) International Institute of Tropical
Agriculture (IITA), Nigeria (2002) 319–328.
[3] VANCE, C.P., Symbiotic nitrogen fixation and phosphorus acquisition. Plant nutrition
in a world of declining renewable resources, Plant Physiol. 127 (2001) 390–397.
[4] GAHOONIA, T.S., NIELSEN, N.E., Root traits as tools for creating phosphorus
efficient crop varieties, Plant Soil 260 (2004) 47–57.
[5] JEMO, M., et al., Aluminum resistance of cowpea as affected by phosphorus-
deficiency stress, J. Plant Physiol. 164 (2007) 442–451.
[6] OLAKOJO, S.A., AYANWOLE, J.A., OBASEMOLA, V.I., Laboratory screening of
seeds of some cowpea cultivars (Vigna unguiculata) for tolerance to cowpea beetles
(Callosobruchus maculatus) in a hot humid environment, American-Eurasian J. Agric.
Environ. Sci. 2 (2007) 528–533.
[7] FATOKUN, C.A., et al., Eds., Challenges and Opportunities for Enhancing
Sustainable Cowpea Production, Proc. World Cowpea Conference III, Ibadan, 2000,
International Institute of Tropical Agriculture (IITA), Nigeria (2002).
[8] SAIDOU, A.K., et al., Genotypic variation in cowpea (Vigna unguiculata L.) for rock
phosphate use in low phosphorus soils of dry Sudan and Sahel Savannas of Niger and
Nigeria, West Africa, American-Eurasian J. Agric. Environ. Sci. 11 (2011) 62–70.
[9] SAIDOU, A.K., et al., Response of cowpea lines to low phosphorus tolerance and
response to external application of P, Afr. J. Microbiol. Res. 6 (2012) 5479–5485.
[10] VINCENT, J.M., A Manual for the Practical Study of Root-nodule Bacteria,
Blackwell Scientific Publications, Oxford (1970) 164 p.
[11] DREVON, J.J., et al., Influence of the Bradyrhizobium japonicum hydrogenase on the
growth of Glycine and Vigna species, Appl. Environ. Microbiol. 53 (1987) 610–612.
[12] AKHTAR, M.S., OKI, Y., ADACHI, T., Genetic variability in phosphorus acquisition
and utilization efficiency from sparingly soluble P-sources by Brassica cultivars under
P-stress, J. Agron. Crop Sci. 194 (2008) 380–392.
[13] GILL, M.A., AHMAD, Z., Inter-varietal differences of absorbed-phosphorus
utilization in cotton exposed to P-free nutrition: Part II. P-absorption and
remobilization in plant, Pak. J. Sci. Res. 55 (2003) 10–14.

198
[14] SUNGTHONGWISES, K., POSS, R., DREVON, J.J., Relations among growth,
nodulation, P efficiency and proton efflux for annual legumes, Asian J. Plant Sci. 8
(2009) 335–343.
[15] OGBONNAYA, C.I., et al., Selection of cowpea genotypes in hydroponics, pots, and
field for drought tolerance, Crop Sci. 43 (2003) 1114–1120.
[16] GERLOFF, G.C., Intact-plant screening for tolerance of nutrient-deficiency stress.
Plant Soil 99 (1987) 3–6.
[17] FUJITA, K., et al., Genotypic variability of pigeonpea in distribution of photosynthetic
carbon at low phosphorus level, Plant Sci. 166 (2004) 641–649.
[18] VADEZ, V., et al., Variability of N2 fixation in common bean (Phaseolus vulgaris L.)
under P deficiency is related to P use efficiency, Euphytica 106 (1999) 231–242.
[19] SCHULZE, J., DREVON, J.J., P-deficiency increases the O2 uptake per N2 reduced in
alfalfa, J. Exp. Bot. 56 (2005) 1779–1784.
[20] RICHARDSON, A.E., et al., Acquisition of phosphorus and nitrogen in the
rhizosphere and plant growth promotion by microorganisms, Plant Soil 321 (2009)
305–339.
[21] SCHACHTMAN, D.P., REID, R.J., AYLING, S.M., Phosphorus uptake by plants:
From soil to cell, Plant Physiol. 116 (1998) 447–453.
[22] VANCE, C.P., UHDE-STONE, C., ALLAN, D.L. Phosphorus acquisition and use:
critical adaptations by plants for securing a nonrenewable resource, New Phytol. 157
(2003) 427–447.
[23] BARGAZ, A., et al., A phosphoenol pyruvate phosphatase transcript is induced in the
root nodule cortex of Phaseolus vulgaris under conditions of phosphorus deficiency. J.
Exp. Bot. 63 (2012) 4723–4730.
[24] HØGH-JENSEN, H., SCHJOERRING, J.K., SOUSSANA, J.-F., The influence of
phosphorus deficiency on growth and nitrogen fixation of white clover plants, Ann.
Bot. 90 (2002) 745–753.
[25] ARAÚJO, A.P., PLASSARD, C., DREVON, J.J., Phosphatase and phytase activities
in nodules of common bean genotypes at different levels of phosphorus supply, Plant
Soil 312 (2008) 129–138.
[26] LOHRASEBI, T., et al., Differential expression of Arabidopsis thaliana acid
phosphatases in response to abiotic stresses. Iran J. Biotechnol. 5 (2007) 130–139.

[27] DREVON, J.J., et al., "Do nodule phosphatase and phytase link with the phosphorus
use efficiency for N2-dependent growth in Phaseolus Vulgaris?", Proc. 15th Int.
Nitrogen Fixation Congress and the 12th Int. Conf. of the African Association for
Biological Nitrogen Fixation, (DAKORA, F.D., et al., Eds), Springer (2008) 93–94.
[28] FLEURAT-LESSARD, P., et al., The distribution of aquaporin subtypes (PIP1, PIP2
and g-TIP) is tissue dependent in soybean (Glycine max) root nodules, Ann. Bot. 96
(2005) 457–460.
[29] SCHULZE, J., et al., Nitrogen fixation by white lupin under phosphorus deficiency,
Ann. Bot. 98 (2006) 731–740.
[30] JEBARA, M., DREVON, J.J., Genotypic variation in nodule conductance to the
oxygen diffusion in common bean (Phaseolus vulgaris), Agronomie 21 (2001) 667–
674.
[31] ANDRIAMANANJARA, A., et al., "Genotypic variation for N2-fixation in voandzou
(Vigna subterranea) under P deficiency and P sufficiency", Proc. 17th Int. Congr. on
Nitrogen fixation, Fremantle, Australia (2011) (abstract).
[32] AKHTAR, M.S., et al., Relative phosphorus utilization efficiency, growth response,
and phosphorus uptake kinetics of Brassica cultivars under a phosphorus stress
environment, Commun. Soil Sci. Plant Anal. 38 (2007) 1061–1085.

199
[33] MAHAMANE, S., Evaluation of Cowpea (Vigna unguiculata L. Walp) Genotypes for
Adaptation to Low Soil Phosphorus Conditions and to Rock Phosphate Application,
Ph. D Thesis, Texas A&M University, College Station (2008).
[34] TSVETKOVA, G.E., GEORGIEV, G.I., Effect of phosphorus nutrition on the
nodulation, nitrogen fixation and nutrient use efficiency of Bradyrhizobium
japonicum-soybean (Glycine max L. Merr.) symbiosis, Bulg. J. Plant Physiol. Special
Issue (2003) 331–335.

200
 SELECTION OF GREEN MANURE SPECIES FOR EFFICIENT
ABSORBTION OF POORLY-AVAILABLE FORMS OF SOIL PHOSPHORUS

V.I. FRANZINI, F.L. MENDES

Brazilian Agricultural Research Corporation,
EMBRAPA-Amazonia Oriental,
Belém, PA,

T. MURAOKA, E.C. DA SILVA

Center for Nuclear Energy in Agriculture,
University of São Paulo,
Piracicaba, SP,

Brazil

J.J. ADU-GYAMFI
Soil and Water Management & Crop Nutrition Laboratory,
International Atomic Energy Agency,
Vienna,
Austria

Abstract

Green manuring is an agronomic practice in which plants or their residues are added to the
soil, improving of the soil physical, chemical and biological attributes, and increasing organic matter
and fertility levels through nutrient cycling. It is estimated that green manures can increase P
bioavailability. The integration of plant species in crop rotations to immobilize P is one of the most
promising agronomic measures to improve the availability of P for the main crop. This study aimed to
assess 21 species of green manure and a standard plant species (Lupinus albus) on their ability to
absorb the available forms of P by the 32P isotopic dilution technique. It also aimed to determine if the
isotopically exchangeable P, the L-values, differed when calculated with or without taking seed N into
account. The results were statistically correlated and analyzed by hierarchical clustering (HCA) in
order to group similar plant species. Jack bean was the most efficient species in P utilization while the
Stylosanthes spp. were the most efficient in P uptake. The seed-derived P affected the P
uptake efficiency evaluated by L-value technique.

1. INTRODUCTION

Green manuring is an agronomic practice, in which plant material or crop residues are
added to soil, promoting the improvement of the soil physical, chemical and biological
attributes, and increase the soil organic matter and its fertility level [1] due to the retention of
nutrients [2]. Conditioner plants that have a greater ability to recycle phosphorus can recover
the organic and inorganic fractions of low availability and reduce the high amounts of
phosphate applied to soil [3], especially when grown in soils with low fertility.

Legumes, when used as green manures, have the ability to utilize insoluble phosphates [4–6] and to
release P in available form after mineralization. Thus, in cultivated soil with green manure, the reservoir of
available P is significantly enriched [7, 8]. Grasses have also been used, mainly in the Cerrado region
of Brazil, due to higher resistance to drought, higher biomass production and lower seed cost.
Moreover, the high temperature associated with high humidity in summer, promotes rapid
decomposition of plant residues with low C: N ratios [9].

201
 Isotopic labeling of plant tissue is widely used in studies of green manure nutrient
dynamics, especially for N, S [10] and P [6, 7, 11], using 15N, 35S, ³²P and ³³P, respectively.
This technique has the great advantage of being able to separate the nutrient derived from the
soil from that derived from the applied green manure, making it possible to determine the
exact green manure nutrient utilization by the crop. It also allows direct measurement of the
differences between plants in their absorption capability of less available forms of soil P [12].

The aim of this study was to use the ³²P isotopic technique (L-value), to compare the P
uptake efficiency of a range of green manure species, and to verify if the P content in the plant
derived from the seeds of the green manure affects the L-value.

2. MATERIALS AND METHODS

2.1. Experimental

The experiment was conducted in the greenhouse at the Center for Nuclear Energy in
Agriculture (CENA/USP), Piracicaba (22º42'30'' S, 47º38'01'' W), São Paulo State, Southeast
Brazil. The dystrophic Typic Haplustox soil [13] used in this study was collected from the
0.0–0.2 m depth, dried, sieved in a 2 mm mesh sieve and homogenized. The soil had clay, silt
and sand contents, respectively, of 280, 70 and 650 g kg-1, and the following chemical
characteristics: pH (0.01 mol l-1 CaCl2), 4.5; organic matter, 22.5 g kg-1; P extracted by resin,
6.25 kg-1; K, 0.75 mmolc kg-1; Ca, 14.4 mmolc kg-1; Mg, 6.5 mmolc kg-1 ; H + Al, 44.2 mmolc
kg-1 ; CEC, 65.9 mmolc kg-1 ; sum of bases, 21.6 mmolc kg-1; base saturation, 32.8%, according
to methodology described by [14]; and P by Mehlich-1, 3.7 mg kg -1 [15].The soil was limed
according to [14], to reach 70% base saturation, and incubated for 30 days prior to the
beginning of the experiment, maintaining the moisture content at approximately 70% of
water holding capacity.

The study was conducted in 3.0 l plastic pots, lined with polyethylene bags, containing
2.5 kg of air-dried soil. A ³²P radioisotope solution (74 MBq pot -1, carrier free) diluted in 200
ml l-1 of distilled water was used to label the soil. After applying this solution, the soil was
incubated for 20 days to reach isotopic equilibrium (³²P/³¹P).

The experimental design was complete randomized with three replications. The
treatment consisted of 21 green manure species: Brachiaria brizantha cv. Marandú, B.
brizantha cv. Xaraés, B. ruziziensis, Calopogonium muconoides, sunhemp (Crotalaria
juncea), and four other species of Crotalaria (C. breviflora, C. mucronata, C. ochroleuca and
C. spectabilis), Stylosanthes guianensis cv. Campo Grande, S. guianensis cv. Mineirão, jack
bean (Canavalia ensiformis), pigeon pea cv. Fava larga (Cajanus cajan), dwarf pigeon pea
(Cajanus cajan), sunflower cv. IAC-Iarama (Helianthus annuus), sunflower cv. IAC-Uruguai
(Helianthus annuus), lab lab (Dolichos lablab), millet (Pennisetum glaucum), dwarf velvet
bean (Mucuna deeringiana), grey velvet bean (Mucuna cinereum) and black velvet bean
(Mucuna aterrima). White lupin (Lupinus albus) was used as the standard species, described
in the literature as efficient in P uptake [5, 12, 16, 17].

Six seeds of the bigger sized species (jack bean, pigeon pea, sunflowers, lab lab,
velvet bean and white lupin) and twelve seeds of the others were sown in each pot and thinned
to two plants pot-1 seven days after emergence (DAE). N and K were applied at rates of 100
mg N kg-1 as urea and 100 mg K kg-1 as KCl. The pots were watered daily with deionized
water and were maintained at 70% of water retention capacity.

202
 The plant shoots were harvested 50 DAE, oven-dried at 60°C for 72 hours, weighed
and ground to pass a 20 mesh sieve. The plant samples were digested with a mixture of nitric
and perchloric acids and analyzed for ³²P activity in a Liquid Scintillation Counter by the
Cerenkov effect [18], and the total P concentration was determined by the method described
by [19].

2.2. L-values

With plant accumulated P values and ³²P / ³¹P specific activities (SA), the L-value and
L-value discounting the P in the plant derived from seed [12, 20] were calculated from the
following equations [21]:

SA = 32P / 31P; where SA is the specific activity (disintegration per minute, dpm µg
P ); P is radioisotope activity in the plant (dpm); 31P is P content in the plant (µg P pot-1);
-1 32

L-value = (A0 / SAf – 1); where L-value (mg P kg-1 soil); A0 is ³²P activity of the
applied carrier-free solution (dpm); SAf is the specific activity of the plant (dpm µgP-1).

L-s value = A0 / SAf; where L-s value is the L-value calculated considering the P in the
plant subtracting the P derived from the seed (mg of P kg-1 soil); Ao is carrier free 32P activity
of the applied solution (dpm); SAfs = 32P/31P, where 32P is the plant 32P activity and 31P is the
P content of the plant. The plant P content subtracting the seed derived P was calculated,
assuming that 60% of P contained in the seed is used in plant development [22].

The L-value is by definition the soil P availability obtained through the isotopic
dilution technique. Under similar soil P conditions, the higher the L-value the more efficient
is the plant species in obtaining P.

2.3. Statistical analysis

Data of shoot dry matter (SDM), P concentration and P uptake in SDM, L-value and
L-s value of the green manure species were arranged in a matrix, which was statistically
analyzed using Pearson linear correlation and hierarchical cluster analysis (HCA). The
statistical method of HCA was used in order to verify the similarities among the species of
green manure, by calculating the Euclidean distance among the samples.

The SAS 9.1 - "Statistical Analysis System” [23] and the SYSTAT version 10.2
software programs, using the UPGMA (un-weighted pair group arithmetic average
clustering), were used to perform binary grouping to define groups according to the degree of
similarity between the species [24]. The cluster analysis was preceded to the standardization
of data before the Euclidian distances calculation, as the variables presented different scales.
After standardization, all variables will be equally important in the determination of these
distances. Final results of the groups were presented as dendrograms. The P uptake efficiency
by plants is inversely proportional to SA and directly proportional to L-value and L-value
discounting the P in the plant derived from the seed.

3. RESULTS AND DISCUSSION

The results obtained with the white lupin as the standard species were: SDM = 1.77 g
pot-1, P in SDM = 1.66 mg pot-1, SA = 9.05 dpm g-1, L-value = 43.20 mg P kg-1 soil and L-s
value = 5.93 mg P kg-1 soil. White lupin was more efficient in absorbing P, resulting in the

203
lowest SA, and the highest L and L-s values among the green manure species (Tables 1 and
2).

TABLE 1. MEANS AND STANDARD DEVIATIONS OF SHOOT DRY MATTER (SDM), P

CONCENTRATION IN SDM (P CONC), P UPTAKE AND SEED P UPTAKE (P SEED)
OF 21 SPECIES OF GREEN MANURE

Species SDMM P conc P uptake P seed

(g) (g kg-1) (mg) (mg)
Brachiaria ruziziensis 3.37 ± 0.05 1.06 ± 0.05 3.57 ± 0.12 0.042
Brachiaria marandú 3.08 ± 0.23 1.05 ± 0.11 3.19 ± 0.22 0.042
Brachiaria xaraés 2.43 ± 0.33 1.18 ± 0.09 2.88 ± 0.52 0.042
Calopogonium 3.50 ± 0.17 1.27 ± 0.06 4.41 ± 0.03 0.102
Crotalaria breviflora 2.20 ± 0.13 1.61 ± 0.12 3.51 ± 0.07 0.132
Sunhemp 6.26 ± 0.11 0.80 ± 0.04 4.99 ± 0.31 0.372
Smooth rattlepod 2.10 ± 0.05 1.22 ± 0.04 2.57 ± 0.10 0.048
Crotalaria ochroleuca 3.43 ± 0.18 1.15 ± 0.07 3.95 ± 0.17 0.054
Showy rattlepod 3.55 ± 0.11 1.17 ± 0.04 4.16 ± 0.28 0.150
Stylosanthes cv. Campo Grande 0.99 ± 0.16 1.85 ± 0.04 1.82 ± 0.28 0.042
Stylosanthes cv. Mineirão 0.60 ± 0.13 1.74 ± 0.29 1.09 ± 0.33 0.042
Jack bean 16.10 ± 0.47 0.71 ± 0.02 11.46 ± 0.66 7.872
Pigeon pea 4.73 ± 0.13 1.03 ± 0.04 4.87 ± 0.11 0.804
Dwarf pigeon pea 4.73 ± 0.28 1.08 ± 0.04 5.07 ± 0.11 0.456
Sunflower cv. Iarama 5.17 ± 0.16 1.10 ± 0.03 5.70 ± 0.35 0.786
Sunflower cv. Uruguai 5.30 ± 0.12 1.06 ± 0.05 5.61 ± 0.32 0.576
Lab lab 5.24 ± 0.28 1.14 ± 0.05 5.94 ± 0.04 1.494
Millet 3.79 ± 0.24 0.87 ± 0.02 3.32 ± 0.28 0.042
Dwarf velvet bean 7.37 ± 0.61 0.86 ± 0.02 6.30 ± 0.37 3.726
Grey velvet bean 8.30 ± 0.59 0.86 ± 0.02 7.12 ± 0.41 5.634
Black velvet bean 10.08 ± 0.56 0.84 ± 0.01 8.45 ± 0.45 6.204

The values of plant SDM of 21 green manure species correlated significantly and
negatively with P concentrations in SDM (–0.733***) and positively with P uptake in SDM
(0.975***). The increase of SDM reduced the P concentration in SDM, while the total P
uptake was higher.

The cluster analysis identified the following six groups of green manure species (Fig.
1):

1st: jack bean;

2nd: black velvet bean,
3rd: grey velvet bean and dwarf velvet bean;
4th: Stylosanthes cv. Minerão and Stylosanthes cv. Campo Grande;
5th: lab lab, sunflower cv. Uruguai, sunflower cv. Iarama, dwarf pigeon pea, pigeon pea
and sunhemp;
6th: Crotalaria breviflora, Crotalaria macronata , Brachiaria brizantha cv. Xaraés,
Crotalaria spectabilis, Crotalaria ochroleuca, Brachiaria brizantha cv. Marandú and
Brachiaria ruziziensis.

Among the five groups of green manure species (Fig. 1) and the values presented in
Table 1, jack bean presented superior values of SDM and higher values of P uptake. The
variables that showed higher Pearson correlations coefficients were: SA and L-value (–
0.949***), SA and L-s value (–0.095ns) and L-value and L-s value (–0.0012 ns).

204
TABLE 2. MEANS AND STANDARD DEVIATIONS OF SPECIFIC ACTIVITY (SA), L-
VALUE AND L-VALUE DISCOUNTING THE P IN PLANT DERIVED FROM SEED (L-
S) AND VARIATION (∆) BETWEEN L-VALUE AND L-S VALUE OF 21 SPECIES OF
GREEN MANURE

Species SA L-value L-s value ∆

(dpm µg-1 P) (mg P kg-1 soil) (mg P kg-1 soil) (%)
Brachiaria ruziziensis 138.33 ± 14.83 2.79 ± 0.56 2.76 ± 0.55 1.18
Brachiaria marandú 172.52 ± 19.69 2.44 ± 0.16 2.41 ± 0.16 1.33
Brachiaria xaraés 241.22 ± 38.32 2.27 ± 0.21 2.24 ± 0.22 1.52
Calopogonium 140.23 ± 3.56 2.60 ± 0.16 2.54 ± 0.16 2.31
Crotalaria breviflora 192.37 ± 27.13 3.12 ± 0.68 3.00 ± 0.66 3.79
Sunhemp 80.56 ± 5.61 2.54 ± 0.22 2.35 ± 0.23 7.46
Smooth rattlepod 184.08 ± 6.96 3.29 ± 0.13 3.23 ± 0.12 1.88
Crotalaria ochroleuca 140.07 ± 7.35 2.65 ± 0.09 2.62 ± 0.08 1.38
Showy rattlepod 135.92 ±1.30 2.63 ± 0.14 2.54 ± 0.13 3.65
Stylosanthes cv. Campo Grande 300.54 ± 44.88 4.55 ± 0.75 4.45 ± 0.73 2.04
Stylosanthes cv. Mineirão 1047.05 ± 34.79 4.77 ± 0.79 4.61 ± 0.80 3.44
Jack bean 10.43 ± 0.55 7.59 ± 0.29 2.35 ± 0.62 68.98
Pigeon pea 88.11 ± 1.53 3.05 ± 0.23 2.55 ± 0.20 16.53
Dwarf pigeon pea 75.53 ± 3.79 3.60 ± 0.53 3.28 ± 0.48 9.03
Sunflower cv. Iarama 93.14 ± 6.45 2.65 ± 0.20 2.29 ± 0.21 13.82
Sunflower cv. Uruguai 84.44 ± 2.89 2.84 ± 0.17 2.55 ± 0.12 10.37
Lab lab 84.34 ± 4.34 2.89 ± 0.16 2.16 ± 0.12 25.17
Millet 150.92 ± 9.08 2.24 ± 0.18 2.21 ± 0.17 1.29
Dwarf velvet bean 46.36 ± 2.26 3.76 ± 0.40 1.51 ± 0.13 59.85
Grey velvet bean 31.38 ± 2.87 4.94 ± 0.15 1.01 ± 0.44 79.61
Black velvet bean 19.65 ± 1.44 6.47 ± 0.24 1.70 ± 0.48 73.68

Green manure species were classified with respect to P uptake efficiency in the
following five groups by the hierarchical cluster analysis with both variables SA and L-value
(Fig. 2).

1st: highly efficient (black velvet bean and jack bean);

2nd: very efficient (grey velvet bean, Stylosanthes cv. Minerão and Stylosanthes cv. Campo
Grande);
3rd: efficient (dwarf velvet bean, dwarf pigeon pea and Crotalaria macronata);
4th: moderately efficient (millet and Brachiaria brizantha cv. Xaraés);
5th: less efficient (lab lab, sunflower cv. Uruguai, pigeon pea, Crotalaria breviflora,
sunhemp, Brachiaria brizantha cv. Marandú, Crotalaria spectablis, sunflower cv.
Iarama, Crotalaria ochroleuca, Calopogonium and Brachiaria ruziziensis).

205
 St Mineirão

St CampoGr

Breviflora

Calopogônio

Showy rattl

Ochroleuca

Ruziziensis

Marandú

Xaraés

Smooth ratt

Millet

Pigeon pea

Dwarf PP

Sun Uruguai

Sun Iarana

Lab lab

Sunn hemp

Dwarf VB

Gray VB

Black VB

Jack bean

0 1 2 3
Euclidean distance

FIG. 1. Dendrogram obtained by hierarchical method of nearest neighbor based on distance

of 0.5 for the green manure species. Hierarchical Cluster Analysis (HCA) - Average Method -
for shoot dry matter (SDM), concentration and accumulation of P in SDM.

The cluster analysis with L-value discounting the P in plant derived from seed
classified green manure species in five groups (Fig. 3).

1st: highly P uptake efficient (Stylosanthes cv. Minerão and Stylosanthes cv. Campo
Grande);
2nd: very efficient (grey velvet bean);
3rd: efficient (black velvet bean and dwarf velvet bean);
4th: moderately efficient (dwarf pigeon pea, Crotalaria macronata and Crotalaria
breviflora);
5th: less efficient (lab lab, millet, Brachiaria brizantha cv. Xaraés, sunflower cv. Iarama,
jack bean, sunhemp, Brachiaria brizantha cv. Marandú, Crotalaria ochroleuca,
sunflower cv. Uruguai, pigeon pea, Crotalaria spectabilis, Calopogonium mucunoides
and Brachiaria ruziziensis).

206
 Millet

Xaraés

Marandú

Sunn hemp

Calopogônio

Showy rattl

Sun Iarana

Ochroleuca

Ruziziensis

Sun Uruguai

Lab lab

Pigeon pea

Breviflora

Smooth ratt

Dwarf PP

Dwarf VB

St CampoGr

St Mineirão

Gray VB

Black VB

Jack bean

0 1 2 3
Euclidean distance

FIG. 2. Dendrogram obtained by hierarchical method of nearest neighbor based on distance of 0.5 for
the green manure species. Hierarchical Cluster Analysis (HCA) - Average Method - for P specific
activity (SA) and L-value.

The green manure plants may convert the relatively unavailable native P in the soil or
residual fertilizer to more available chemical forms for the next crop. White lupin can take up
more P from soil than alfalfa, clover, peas, vetch and wheat grown in P- deficient soil [5, 16].
With respect to the capacity of mobilization and recycling of nutrients between legumes and
grasses used as green manure, pea and sunhemp showed better results compared to grass, due
to higher biomass and higher nutrient concentrations in biomass [25].

207
 Gray VB

Dwarf VB

Black VB

Lab lab

Millet

Xaraés

Sun Iarana

Sunn hemp

Jack bean

Marandú

Showy rattl

Calopogônio

Pigeon pea

Sun Uruguai

Ochroleuca

Ruziziensis

Breviflora

Smooth ratt

Dwarf PP

St CampoGr

St Mineirão

0 1 2 3
Euclidean distance

FIG. 3. Dendrogram obtained by hierarchical method of nearest neighbor based on distance

of 0.5 for the green manure species. Hierarchical Cluster Analysis (HCA) - Average Method -
for L-value discounting the P in the plant derived from the seed (L-s).

White lupin, pigeon pea and Stylosanthes are well adapted to acid soils deficient in P
[3, 5, 26]. The white lupin roots secrete large amounts of citric acid, which solubilizes the
fixed P [27–30], pigeon pea releases P from iron and aluminum phosphates through the
secretion of piscidic, malonic and oxalic acids by roots [12, 31, 32]. Stylosanthes roots exude
citrate and release P from Fe and Al phosphates. These organic acids solubilize the P, thus
increasing P uptake by the plant subsequently [3, 33–35], making these species resistant to P
deficiency in soil.

Legumes and grasses have been used as cover crops. However, they must have some
desirable features, for their beneficial effects to succeed in the main crops, and facilitate the
cultivation system [36]. The species to be used as green manures should provide enough
biomass to cover the surface area and improve the yield of the main crops, such as providing
nutrients as well as being compatible with the management of the main crop [37].

The L-values, considering or not the seed derived P, are illustrated in the Table 2.
There was increasing variation in the differences between the L and L-s values of the green

208
manure species with higher values of seed derived P, varying form 1% for Brachiaria
ruziziensis (lower seed P content, Table 1) up to 80% for grey velvet bean (Table 2),
suggesting that the higher the seed P, the greater the influence on the correction of the seed P
effect on the L-value (Tables 1 and 2). Reducing the L-value to more realistic levels by
correcting the seed derived P were also shown in an experiment with ryegrass grown in three
soils of very low, low and medium P status (soil I, II and III), containing 4, 14 and 37 mg
initial NaHCO3-soluble P kg-1 soil, and supplied with increasing amounts of added P. The L-
values after correcting the seed-derived P were reduced by 69% in soil I, 18% in soil II and
10% in soil III. This effect was still much higher, when L-values were reduced by increasing
levels of fertilizer P [22]. In a study on the dynamics of P uptake by
wheat using radioisotopes 32P and 33P, the specific activities of the plants in the early
development stage for both radioisotopes were much lower, presumably due to the high
amount of seed- derived P in the plant [38]. Chickpea, faba bean, white lupin, canola and
wheat were evaluated in the greenhouse for P uptake and growth in three soil types with
low, medium and high P content, by labeling with ³²P. It was concluded that the L-
value for faba bean grown in soils with low P was compromised by the large proportion of
the seed P in relation to P taken up in shoots, making the calculation of L-value very
sensitive to the estimated value of P derived from the seeds accumulated in the shoot [39].

Stylosanthes cv. Minerão and Stylosanthes cv. Campo Grande showed no difference
between L-value and L-s value, probably because they have lower seed P contents (Table 1).
Jack bean was classified in this study as the species that absorbed more P (Fig. 1), with higher
SDM, P content and P uptake (Table 1), and highly efficient in P absorbing capacity (Fig. 2).
However, when discounting the P from the seed (Table 1), it was the least efficient in P
uptake (Fig. 3), despite its high SDM and P uptake in SDM (Table 1). The L-s value was the
most trustworthy, because a fraction of the amount of P translocated from the seed is
subtracted from the total P uptake by the plant [40, 41].

The specific activity which is inverse to the L-value showed no correlation between
the L-s value, seed SA and L-value (–0.095ns) and L-s value and L-value (–0.0012 ns), due to
differences in the seed P contents of the green manures studied. Therefore, the results could
be erroneous when the L-values are calculated without considering the P in the plant derived
from the seed.

4. CONCLUSIONS

- Jack bean was more efficient in P utilization, i.e., the most productive under low soil
available P conditions.

- Stylosanthes, regardless of cultivar Mineirão or Campo Grande, was the most efficient
green manure species for P uptake.

- The seed-derived P in the green manures, when L-value is used, affects the
identification and classification of P uptake efficiency

ACKNOWLEDGEMENTS

We thank IAEA for providing financial support (Project RC 13779), CAPES for the
research scholarship granted to the first and third authors and CNPq for the research
fellowship granted to the second author.

209
 REFERENCES

[1] CALEGARI, A., et al., “Aspectos gerais da adubação verde”, Adubação Verde no Sul
do Brasil, (COSTA, M.B.B., Ed.), AS-APTA, Rio de Janeiro (1992) 346 p.
[2] RICHARDSON, A.E., et al., Plant mechanisms to optimize access to soil phosphorus,
Crop Pasture Sci. 60 (2009) 124–143.
[3] LI, X.F., et al., Secretion of citrate from roots in response to aluminum and low
phosphorus stresses in Stylosanthes, Plant Soil 325 (2009) 219–229.
[4] FULLER, W.H., DEAN, L.A., Utilization of phosphorus from green manures, Soil Sci.
68 (1949) 197–202.
[5] CAVIGELLI, M.A., THIEN, J.S., Phosphorus bioavailability following incorporation
of green manure crops, Soil Sci. Soc. Am. J. 67 (2003) 1186–1194.
[6] BAH, A.R., ZAHARAH, A.R., HUSSIN, A., Phosphorus uptake from green manures
and phosphate fertilizers applied in an acid tropical soil, Commun. Soil Sci. Plant Anal.
37 (2006) 2077–2093.
[7] NACHIMUTHU, G., et al., Isotopic tracing of phosphorus uptake in corn from ³³P
labeled legume residues and ³²P labeled fertilisers applied to a sandy loam soil, Plant
Soil 314 (2009) 303–310.
[8] NURUZZAMAN, M., et al., Phosphorus benefits of different legume crops to
subsequent wheat grown in different soils of Western Australia, Plant Soil 271 (2005)
175–187.
[9] LARA CABEZAS, W.L., et al., Influência da cultura antecessora e da adubação
nitrogenada na produtividade de milho em sistema plantio direto e solo preparado,
Ciência Rural 34 (2004) 1005–1013.
[10] SILVA, E.C., et al., Aproveitamento do nitrogênio (15N) da crotalária e do milheto
pelo milho sob plantio direto em Latossolo Vermelho de Cerrado, Ciência Rural 36
(2006) 739–746.
[11] FRANZINI, V.I., MURAOKA, T., MENDES, F.L., Ratio and rate effects of ³²P-triple
superphosphate and phosphate rock mixtures on corn growth, Scientia Agricola 66
(2009) 71–76.
[12] HOCKING, P.J., et al., “Comparation of the ability of different crop species to access
poorly-available soil phosphorus”, Plant Nutrition for Sustainable Food Production
and Environment, (ANDO, T., et al., Eds), Kluwer Academic Publishers (1997) 305–
308.
[13] DOS SANTOS, H.G., et al., Sistema Brasileiro de Classificação de Solos, 2nd Edn,
EMBRAPA-Solos, Rio de Janeiro (2006) 306 p.
[14] VAN RAIJ, B., et al., Análise Química para Avaliação da Fertilidade de Solos
Tropicais, Instituto Agronômico, Campinas (2001) 285 p.
[15] EMBRAPA-SOLOS, Manual de Métodos de Análises de Solos, 2nd Edn, Rio de
Janeiro, (1997) 212 p.
[16] BRAUM, S.M., HELMKE, P.A., White lupin utilizes soil phosphorus that is
unavailable to soybean, Plant Soil 176 (1995) 95–100.
[17] MURAOKA, T., et al., “Comparison of the ability of different plant species and corn
hybrids to access poorly-available soil phosphorus in an Oxisol of the Cerrado region,
Brazil”, Management Practices for Improving Sustainable Crop Production in Tropical
Acid Soils, Proceedings Series, IAEA, Vienna (2006) 137–146.
[18] VOSE, P.B., Introduction to Nuclear Techniques in Agronomy and Plant Biology,
Pergamon Press, London (1980) 391 p.
[19] SARRUGE, J.R., HAAG, H.P., Análises Químicas em Plantas, Departamento de
Química, ESALQ, Piracicaba, SP (1974) 57 p.

210
[20] LARSEN, S., The use of ³²P in studies on the uptake of phosphorus by plants, Plant
Soil 4 (1952) 1–10.
[21] DI, H.J., CONDRON, L.M., FROSSARD, E., Isotope techniques to study phosphorus
cycling in agricultural and forest soils: a review, Biol. Fertil. Soils 24 (1997) 1–12.
[22] BROOKES, P.C., Correction for seed-phosphorus effects in L-value determinations, J.
Sci. Food Agric. 33 (1982) 329–335.
[23] SAS INSTITUTE, SAS User’s Guide: Statistics, Veriono 8.2, SAS Institute, Cary
(2001).
[24] BEEBE, K.R., KOWALSKI, B.R., An introduction to multivariate calibration and
analysis, Anal. Chem. 59 (1987) 1007A–1017A.
[25] DE ALCÂNTARA, F.A., et al., Adubação verde na recuperação da fertilidade de um
Latossolo Vermelho-Escuro degradado, Pesq. Agropec. Bras. 35 (2000) 277–288.
[26] WANG, B.L., et al., Citrate exudation from white lupin induced by phosphorus
deficiency differs from that induced by aluminum, New Phytol. 176 (2007) 581–589.
[27] GARDNER, W.K., BARBER, D.A., PARBERY, D.G., The acquisition of phosphorus
by Lupinus albus L. III. The probable mechanism by which phosphorus movement in
the soil / root interface is enhanced, Plant Soil 70 (1983) 107–124.
[28] DINKELAKER, B., RÖMHELD, V., MARSCHNER, H., Citric acid excretion and
precipitation of calcium citrate in the rhizosphere of white lupin (Lupinus albus L.),
Plant Cell Environ. 12 (1989) 285–292.
[29] JOHNSON, J.F., VANCE, C.P., WEIBLEN, G., Root carbon dioxide fixation by
phosphorus-deficient Lupinus albus: Contribution to organic acid exudation by
proteoid roots, Plant Physiol. 112 (1996) 19–30.
[30] SHANE, M.W., et al., Impact of phosphorus mineral source (Al-P or Fe-P) and pH on
cluster-root formation and carboxylate exudation in Lupins albus L., Plant Soil 304
(2008) 169–178.
[31] OTANI, T., AE, N., TANAKA, M., Phosphorus uptake mechanisms of crop growth in
soils with low P status, Soil Sci. Plant Nutr. 42 (1996) 553–560.
[32] HOCKING, P.J., RANDALL, P.J., “Better growth and phosphorus nutrition of
sorghum and wheat following organic acid secreting crops”, Plant Nutrition, Food
Security and Sustainability of Agro-ecosystems Through Basic and Applied Research,
(HORST, W.J., SCHENK, M.K., BÜRKERT, A., Eds), Kluwer Academic. Dordrecht
(2001) 548–549.
[33] NZIGUHEBA, G., et al., Soil phosphorus fractions and adsorption as affected by
organic and inorganic sources, Plant Soil 198 (1998) 159–68.
[34] NZIGUHEBA, G., et al., Organic residues affect phosphorus availability and maize
yields in a Nitisol of western Kenya, Biol. Fertil. Soil 32 (2000) 328–339.
[35] HAYNES, R.J., MOKOLOBATE, M.S., Amelioration of Al toxicity and P deficiency
in acid soil by additions of organic residues: a critical review of the phenomenon and
the mechanisms involved, Nutr. Cycl. Agroecosyst. 59 (2001) 47–63.
[36] FAGERIA, N.K., BALIGAR, V.C., BAILEY, B.A., Role of cover crops in improving
soil and row crop productivity, Commun. Soil Sci. Plant Anal. 36 (2005) 2733–2757.
[37] BALIGAR, V.C., FAGERIA, N.K., Agronomy and physiology of tropical cover crops,
J. Plant Nutr. 30 (2007) 1287–1339.
[38] MCLAUGHLIN, M.J., ALSTON, A.M., MARTIN, J.K., Phosphorus cycling in wheat-
pasture rotations. I. The source of phosphorus taken up by wheat, Aust. J. Soil Res. 26
(1988) 323–321.
[39] HENS, M., HOCKING, P., “An evaluation of the phosphorus benefits from grain
legumes in rotational cropping using ³³P isotopic dilution”, Proc. Int. Crop Sci. Congr.
4, Brisbane (2004). www.cropscience.org.au

211
[40] LARSEN, S., GUNARY, D., The determination of labile soil phosphate as influenced
by the time of application of labeled phosphate, Plant Soil 20 (1964) 135–142.
[41] RUSSELL, R.S., RUSSELL, E.W., MARAIS, P.G., Factors affecting the ability of
plants to absorb phosphate from soils. I. The relationship between labile phosphate and
absorption, J. Soil Sci. 8 (1957) 248–267.

212
ROOT TRAITS FOR BETTER NITROGEN ACQUSITION AND
USE IN LOW-N SOILS
 ROLE OF TRANSLOCTED SIGNALS IN REGULATING ROOT
DEVELOPMENT AND NUTRIENT UPTAKE IN LEGUMES

C.A. ATKINS

School of Plant Biology,

the University of Western Australia,
Crawley, WA,
Australia

Abstract

Uptake of nutrients is achieved through the expression and activity of specific

carrier/transporter mechanisms localized in the root system and distributed as a consequence of the
development of the architecture of the system. Both root system development and the nutrient
transport mechanisms are responsive to environmental factors that include nutrient supply and
availability, water supply, salinity, soil acidity and compaction together with a wide range of biotic
stresses. The response to each may be regulated at the molecular level by both local and systemic
signals. These signals include the classical plant growth regulators but also low molecular weight
compounds such as sugars and amino acids as well as macromolecules, including peptides, proteins
and nucleic acids. Among the latter, recent research has shown that small RNA species and especially
small interfering RNAs (siRNA) and microRNAs (miRNA) are potent and effective regulators of gene
expression which, in the context of root development as well as nutrient uptake, have central and
critical roles. Systemic (translocated) signals that specifically regulate root development and function
are less well defined but analyses of phloem exudate in species of lupin (Lupinus albus and L.
angustifolius) and species of Brassica and cucurbits have demonstrated that a wide range of
macromolecules, including miRNAs, are present and potentially translocated from source organs
(principally leaves) to sinks (shoot apical meristems, developing fruits and seeds, roots and nodules).
While specific signaling roles for many of these macromolecules are yet to be discovered there are
some that have been documented and their regulatory activity in organ development and functioning,
as well as in nutrition, confirmed. The following article provides an up to date review and presents the
results of recent research using lupin with emphasis on the analysis of small RNAs and their likely
role(s) in regulation of root development and function.

1. ANALYSIS OF MACROMOLECULAR COMPOSITION OF PHLOEM EXUDATE

IN LUPIN

Exudates were collected from incisions to the major vascular bundles associated with
the sutures of developing fruits of L. albus and proteins separated using 2D PAGE.
Proteomic analysis identified 83 proteins with the largest group comprising those involved in
metabolism (24%), followed by redox regulation (8%), protein synthesis, turnover and sorting
(9%), cell wall synthesis (6%). More prominent proteins were cyclophilin, ubiquitin, a
glycine-rich RNA-binding protein, a group of proteins that comprise a glutathione/ascorbate-
based mechanism to scavenge O2 radicals, enzymes of glycolysis and other metabolism
including methionine and ethylene synthesis [1]. Included was the ‘florigenic’ signal,
flowering locus T-protein (FT). One third of the proteins sequenced remained unknown with
some of these probably unique to lupin phloem. Both subunits of Rubisco and a chlorophyll
a/b binding protein were also identified in exudate indicating some contamination with the
contents of chloroplast-containing cells damaged by incision.

215
 Analysis of a cDNA library constructed from lupin phloem exudate sequenced 1063
clones and identified 609 unique transcripts that were also classified into functional groups.
Transcripts coding for proteins with unknown functions formed the largest category (280
sequences, 39% of all ESTs) while among those coding for proteins of known function were
metabolism (15% of all ESTs; 11% general and 4% energy metabolism), protein
modification/turnover (11%), and redox regulation, signaling and stress response and defense-
related with around 5% of all ESTs in each of these groups. Potential signaling
macromolecules included transcripts encoding proteins mediating Ca ion level and that for
FT. For 31 of the transcripts that were identified their corresponding protein was also
detected in phloem exudate. Although questions as to whether the many transcripts are
translocated in phloem and, if indeed they act as signals as a consequence of their
translocation, are unresolved. Their long distance movement is believed to occur as a
ribonucleoprotein complex dependant on the presence of suitable RNA binding proteins
(RBPs; [2]). As noted above, proteomic analysis of exudate from lupin identified a glycine-
rich RNA-binding protein, and Ham et al. [2] have shown translocation of a 50kD RNA
binding protein (RBP50) together with six different mRNA species in heterografting
experiments with pumpkin stocks and cucumber scions.

A number of microRNAs (miRNA) were enriched in exudate and differed in exudate

collected from different sites on the lupin plant. These data are discussed in detail below.

One group of molecules that lie between the low molecular weight solutes (sugars,
amino acids, organic acids and alkaloids, etc) and macromolecules in phloem are the bioactive
peptides. A recent review has highlighted the significance of an ever increasing list of these
potentially powerful signals/regulators [3] and Hoffmann-Benning et al. [4] have shown that
lupin exudates, for example, contains a large number of small peptides that to date have not
been studied further. There is compelling evidence that a CLAVATA-like peptide is
translocated in phloem from leaves to the root system in legumes where it exerts a regulatory
role over the intensity of nodulation [5], and the possibility that there are other systemic
signals among the peptides in phloem should not be overlooked.

These analyses from lupin, as do those reported by other researchers using exudates
from Brassica species (B. napus) or cucurbits (e.g. melon, pumpkin and cucumber; reviewed
in [6]), rely on incising the vasculature and collection of exudate from the site of the wound.
While it is reasonable to assume that the contents of the sieve elements (SE) constitute the
bulk of solutes and macromolecules in exudate there will necessarily be contamination from
cells other than those of phloem that are damaged by the incision. The methods of collecting
exudates and the likely levels of contamination have been recently reviewed [7]. Only the use
of severed stylets from sap sucking insects (leaf hoppers and, mainly, aphids) offers the
possibility of analyzing exudate that is minimally contaminated by cell contents other than
SE. However, stylectomy in situ is technically difficult and is effective mainly with
monocotyledon species yielding total volumes of a few microlitres at best. Nevertheless,
confirmation of the macromolecular composition of exudate from severed vasculature using
the few species that ‘bleed’ freely following a wound awaits comparative analysis of stylet
exudate. These considerations emphasize the need for caution in interpreting data from
species like lupin particularly in assigning translocated signaling roles for macromolecules
that are present in very small amounts.

2. A CENTRAL REGULATING ROLE FOR MIRNAS

miRNAs function as post transcriptional regulators of gene expression both in plants

and animals. They are small (18-26 nt) non-protein coding RNAs that are processed from

216
hairpin precursors by the ribonuclease III like enzyme, Dicer, and bind to cognate sequences
in mRNAs. In plants, miRNAs bind to their targets through perfect or close to perfect
complementarity causing degradation of the message or repression of translation. There is an
ever increasing list of developmental processes that involve regulation by miRNAs including
development of leaves, stems and roots (as well as legume nodules) together with flowering,
fruit development and seed set. A recent review by Khan et al. [8] has assembled the growing
evidence for miRNA regulation of gene expression as having a central role in root
development and architecture. They have further identified plant response to a wide variety
of stresses including both biotic and abiotic stresses as mediated through miRNA, in many
cases in the root system.

Small RNAs ranging in size from 8 to 35 nt, with the majority 19 to 23 nt, were
isolated from lupin phloem exudate. Those in the 18 to 26 nt size class were purified and used
to construct a small RNA library, yielding sequences for 330. Seventeen sequences from the
phloem library showed strong similarity to known miRNAs from seven different families,
with 12 identified as probable miRNAs by homology with those from other species. While
not all miRNAs identified in this study targeted transcription factors, the majority did, and
their activity could potentially have a widespread impact on gene expression. Among the 11
miRNAs analyzed by northern blot and real time PCR in lupin tissues, some were prominent
in phloem exudate while others were either absent or much lower than in the other tissues
studied, suggesting that there is a specific spectrum of miRNAs in lupin phloem (Fig. 1).
Furthermore, the miRNA composition in phloem exudates collected from the fruits or close to
apical meristems differed markedly from the composition in exudate collected from the base
of the stem close to the root system [1]. The most reasonable explanation for the presence of
small RNAs in phloem is that they are transported from the companion cell (CC) to the SE.
Thus, the fact that the pattern for five miRNAs in exudate collected at three different sites was
not the same, suggests either their differential expression in the adjacent phloem CC at
different sites and/or specificity in phloem loading. Interestingly, in another study [9] parallel
analyses of miRNAs in stylet exudate and phloem tissue extracts of apple revealed that, while
seven miRNAs were common, four that were amplified from the tissue (including CC) were
not detected in exudate, lending further support to the idea that transfer to the SE is specific.
If miRNAs serve as translocated signals it is not too surprising that the downward-moving
(collected at the base of stem) and upward-moving phloem streams (collected from pods and
axillary branches) showed distinct differences in their miRNA composition [1]. Thus lupins
offer the possibility of sampling exudate from phloem translocating from ‘source’ organs of
the shoot, including leaflet midribs and petioles, to ‘sinks’ such as fruits and apices, as well as
to those of the root system separately [10]. It would be interesting to further exploit this
ability by extending the range of miRNAs assayed to include the many that have now been
identified in exudates (reviewed, [7]). MiRNA-guided post-transcriptional gene regulation is
particularly interesting because miRNAs can regulate several protein-coding genes implicated
in the same pathway, altering whole metabolic pathways and complex response networks in
significant ways.

217
 6x10^6

5x10^6

4x10^6

3x10^6

2x10^6

1x10^6
m m
i i
R R
1 m 1
6 i 6
4 R 8
3
9 m
10.0 Pod 12.0 Pod A Pod B Pod
0 0
Phloem Phloem Phloem Phloem

FIG. 1. Expression of miRNAs in phloem exudate (phloem) and extracts of tissue sampled from
adjacent to the suture vasculature of lupin fruits (pod). The exudate assayed was collected from the
incised vasculature of the same fruits. (Taken from Rodriguez-Medina et al. [1]). The vertical axis is
expressed as molecules/ul template used in the PCR-based assays.

3. NUTRIENT UPTAKE AND HOMEOSTASIS

Physiological studies of plant mineral nutrition over many years have identified
homeostasis as an important feature both in the uptake and accumulation of nutrients.
Schachtman and Shin [11] reviewed the evidence for signaling events in relation to internal
nutrient availability providing strong support for the idea that plant roots respond to
fluctuations in the environment to enhance and optimize the acquisition of water and limiting
nutrients. While self regulation of uptake by roots has been postulated, it was not until
relatively recently that the tools of molecular biology have permitted the identification of ion
transporters and their encoding genes with a range of affinities for the transported species.
Thus, both high and low affinity transporters for Pi, NO3–, SO42– and a range of cations have
been described. Their differential expression in response to nutrient status has long been
thought to involve both local and systemic signaling, the latter specifically through
translocation in phloem, but until very recently these ideas remained speculative.

218
 Perhaps the best documented relationship between a phloem-mobile miRNA and the
regulation of nutrient uptake is that relating to Pi [12, 13]. The regulator in this case was
identified as miR399. Also involved in the Pi response was miR398, which has been reported
to participate in more abiotic and biotic stresses than any other miRNA [14]. In addition to
regulating gene expression associated with Pi uptake, miR398 has a role in a number of stress
responses including both bacterial and fungal pathogens, drought and salinity stress, UV-B
radiation, responses involving abscisic acid (ABA) and responses associated with oxidative
stress. The important target genes for miR398 are two closely related Cu/Zn superoxide
dismutases (CSD1 and CSD2), a subunit of mitochondrial cytochome oxidase and the Cu
chaperone for SOD. As a consequence Cu homeostasis in the plant is also regulated by this
miRNA. In lupin phloem exudate miR399 was most prominent in samples of the downward
moving stream collected at the base of the stem just above the root system, and was
responsive to Pi nutrition, increasing significantly under limitation [1].

Huang et al. [15] have identified miR395 as a regulator of SO42– uptake and as a
mediator in regulating response of Brassica napus to heavy metal stress (Cd2+).
Transcriptional analysis showed 13 miRNAs, including miR395, differentially expressed at
elevated levels in these two responses. Five genes, BnSultr2;1 and BnAPS1-4, which encode
a low-affinity SO42– transporter and a family of ATP sulphurylases, respectively, were
identified as the targets of miR395. SO42– metabolism in roots through cysteine synthesis
leads to the formation of glutathione, a central component of ROS metabolism, and in the
formation of complexes with xenobiotic molecules entering the plant. Lupin phloem was
found to contain miR395 as well as enzymes essential for redox regulation through
ascorbate/glutathione and also glutathione-S-transferase that would mediate conjugation of
xenobiotics [7]. That phloem exudate contains a high concentration of glutathione was
discovered many years earlier in cucurbits [16].

It is also clear that systemic regulation of NO3– uptake by the N status of the plant is
genetically regulated [17]. Split root experiments using transgenic Arabidopsis (hni mutants)
have shown clearly that the overall plant N status systemically regulates NO3– uptake by the
root system [17] through repression of the high affinity NO3– transporter gene, NRT2.1. The
nature of the systemic signals have not been clearly identified but organic solutes of N (amino
acids or amides) as well as micro RNAs, specifically miR167, have been suggested. MiR167
targets the auxin response factor gene (ARF8) as part of a regulatory mechanism modulating
lateral root emergence, but further research is required to establish the relationship
specifically for NO3– homeostasis. This miRNA was a prominent species in lupin phloem
exudate [6] and occurred also in exudate from both Brassica and cucurbit species, as well as
in stylet exudate from apple [9]. Vidal et al. [18] have also examined the regulation of N
status in Arabidopsis and identified miR393 as a central controlling element. This micro
RNA targets a transcription factor and one of the root expressed auxin receptors (AFB3) that
provide a regulatory module integrating N status and auxin signaling to control root
architecture. The picture is made more complex because cytokinins (CK) are increased in
response to increased NO3– supply as a consequence of inducing one of the CK synthesis
genes isopentenyltransferase (IPT3), leading in turn to changes in N uptake and in lateral root
development (architecture). CK are translocated both in xylem and phloem [19] and as such
could provide a further systemic signal that communicates N status from roots to shoots and
vice versa. Ruffel et al. [20] have recently provided convincing evidence that NO3– sensing
triggers long distance systemic signaling to report plant N demand through a NO3–-CK relay
system. They speculate further that indeed auxin may be part of the long distance signal as a
means to inform the root system of the N status of the shoot. Clearly the signals that regulate
N uptake, distribution and utilization in the plant are complex, and will require considerable

219
further research before the features are sufficiently defined to develop genetic markers that
might assist in enhancing N economy in crop species.

A broad study of the differential expression of some 68 known miRNAs in common

bean (Phaseolus vulgaris) using a macro-array hybridization approach identified 33 miRNAs
of which five were only expressed under nutrient stress conditions [21]. The study used
leaves, roots and nodules from plants suffering Pi or Fe deficiency, acidic pH or Mn toxicity.
MiR157, miR156, miR167, miR319, and miR398 responded to all the nutrient stresses in
each of the plant organs, and of this group all have been detected in phloem exudates from
lupin, Brassica spp, cucurbits and importantly from stylet exudate of apple (reviewed in [7]).
It is perhaps not surprising that miRNAs involved in nutrient status are at least in part among
the translocated signals that traverse the distance between sources and sink tissues in a plant.

In the study by Valdez- Lopez [21] discussed above, most of the miRNAs that were
responsive to the stresses in nodules were up regulated with the expression of some specific to
this organ. A number of miRNAs have been found to be involved in the initiation and
development of legume nodules in a wider range of legumes (reviewed in [21]) and definitive
studies have used small RNA libraries from root tips and nodules of Medicago truncatula to
reveal 100 putative new and 36 conserved miRNAs that are expressed. There seems little
doubt that these powerful regulators of gene expression are involved in establishing and
sustaining the symbiotic relationship.

4. CONCLUDING REMARKS

The functional significance of each of the proteins, transcripts and small RNAs in
phloem provides new and exciting prospects to identify the many hypothetical ‘signals’
postulated by the wealth of physiological research that has described many of the processes
involved in plant growth and development. However, the presence of this bewildering array
of macromolecules in phloem also poses new questions about the biochemical features and
maintenance of the SE itself. Not only does there appear to be specific metabolic components
that might support the SE, there is also a wide range of proteins and transcripts that are, in
theory at least, involved in defending the tissue against pathogens and predators (either for the
plant generally or the phloem in particular). One of the prominent groups of mRNAs found in
lupin phloem exudate was 10 transcripts (3% of defined ESTs) for the polyprotein of the bean
yellow mosaic virus (BYMV) [1]. These were most likely due to the presence of feeding
aphids but their presence and abundance in exudate indicates that the phloem stream is also a
pathway for pathogen attack. How the ‘protective’ macromolecules/metabolism in phloem
interact with a pathogen at the molecular level remains as another interesting question.

A recent compilation identified 13 miRNAs involved in plant responses to

drought/salt stress [22]. Eight of these were present in lupin phloem exudate [1] and,
importantly, six were also recovered from PCR amplification of apple stylet exudate [9].
There is thus a possibility that the responses in lupin to both drought and salinity are mediated
through miRNAs translocated from sites where the stress is sensed to sites where a response
is initiated. Participation of the root system in both sensing and responding to these stresses is
clearly indicated. Both drought and salinity stresses are significant constraints to lupin
production in Western Australian cropping systems and perhaps translocated miRNA signals
can provide a basis for marker assisted crop improvement. Furthermore, sampling the phloem
stream (as in lupin) for sensitive analysis of translocated low molecular weight solutes,
transcripts and small RNAs, as well as specific proteins, is non-destructive and could
conceivably be adapted for use as an effective selection tool in a breeding program.

220
 REFERENCES

[1] RODRIGUEZ-MEDINA, C., et al., Macromolecular composition of phloem exudate

from white lupin (Lupinus albus L.), BMC Plant Biol. 11 (2011) 36–54.
[2] HAM, B-K., et al., A polypyrimidine tract binding protein, pumpkin RBP50, forms the
basis of a phloem-mobile ribonucleoprotein complex, Plant Cell 21 (2009) 197–215.
[3] MATSUBAYASHI, Y., Post-translational modifications in secreted peptide hormones
in plants, Plant Cell Physiol. 52 (2011) 5–13.
[4] HOFFMANN-BENNING, S., et al., Comparison of peptides in the phloem sap of
flowering and non-flowering Perilla and lupine plants using microbore HPLC
followed by matrix-assisted laser desorption/ionization time-of-flight mass
spectrometry, Planta 216 (2002) 140–147.
[5] SEARLE, I.R., et al., Long-distance signaling in nodulation directed by a CLAVAT1-
like receptor kinase, Science 299 (2003) 109–112.
[6] ATKINS, C.A., SMITH, P.M.C., Translocation in legumes: assimilates, nutrients, and
signaling molecules, Plant Physiol. 144 (2007) 550–561.
[7] ATKINS, C.A., et al., Macromolecules in phloem exudates - a review, Protoplasma
248 (2011) 165–172.
[8] KHAN, G.A., et al., MicroRNAs as regulators of root development and architecture,
Plant Mol. Biol. 77 (2011) 47–58.
[9] VARKONYI-GASIC, E., et al., Characterisation of microRNAs from apple (Malus
domestica “Royal Gala”) vascular tissue and phloem sap, BMC Plant Biol. 10 (2010)
159–173.
[10] PATE, J.S., et al., “Transport physiology and partitioning”, Lupins as Crop Plants:
Biology, Production and Utilization, (GLADSTONES, J.S., ATKINS, C.A.,
HAMBLIN, J., Eds), CAB International, UK (1998) 181–226.
[11] SCHACHTMAN, D.P., SHIN, R., Nutrient sensing and signaling: NPKS, Ann. Rev.
Plant Biol. 58 (2007) 47–69.
[12] LIN, S., et al., Regulatory network of microRNA399 and PHO2 by systemic signaling,
Plant Physiol. 147 (2008) 732–746.
[13] PANT, B.D., et al., MicroRNA399 is a long-distance signal for the regulation of plant
phosphate homeostasis, Plant J. 53 (2008) 731–738.
[14] ZHU, C., et al., MiR398 and plant stress responses, Physiol. Plant. 143 (2011) 1–9.
[15] HUANG, S.Q., et al., A set of miRNAs from Brassica napus in response to sulphate
deficiency and cadmium stress, Plant Biotech. J. 8 (2010) 887–899.
[16] ALOSI, M.C., et al., The regulation of gelation of phloem exudate from Cucurbita
fruit by dilution, glutathione, and glutathione reductase, Plant Physiol. 86 (1988)
1089–1094.
[17] GIRIN, T., et al., Identification of arabidopsis mutants impaired in the systemic
regulation of root nitrate uptake by the nitrogen status of the plant, Plant Physiol. 153
(2010) 1250–1260.
[18] VIDAL, E.A., et al., Nitrate-responsive miR393/AFB3 regulatory module controls
root system architecture in Arabidopsis thaliana, Proc. Nat. Acad. Sci. USA 107 (2010)
4477–4482.
[19] EMERY, R.J.N., et al., The forms and sources of cytokinins in developing Lupinus
seeds and fruits, Plant Physiol. 123 (2000) 1593–1604.
[20] RUFFEL, S., et al., Nitrogen economics of root foraging: Transitive closure of the
nitrate-cytokinin relay and distinct systemic signaling for N supply vs. demand, Proc.
Nat. Acad. Sci. USA 108: 18524–18529.

221
[21] VALDÉS-LÓPEZ, O., et al., MicroRNA expression profile in common bean
(Phaseolus vulgaris) under nutrient deficiency stresses and manganese toxicity, New
Phytol. 187 (2010) 805–818.
[22] COVARRUBIAS, A.A., REYES, J.L., Post-transcriptional gene regulation of salinity
and drought responses by plant microRNAs, Plant Cell Environ. 33 (2010) 481–489.

222
 ASSESSMENT OF ROOT MORPHOLOGICAL TRAITS OF 16 TROPICAL
AND FOUR TEMPERATE MAIZE CULTIVARS FOR NITROGEN
EFFICIENCY IN SHORT-TERM NUTRIENT SOLUTION EXPERIMENTS
WITH THE CIGAR ROLL AND GROWTH POUCH METHODS

S. SAIFU, G. SCHULTE auf’m ERLEY, W.J. HORST

Institute for Plant Nutrition,
Leibniz University of Hannover,
Federal Republic of Germany

Abstract

Genotypic differences in N efficiency of maize have been reported by many authors. One of
the reasons responsible for genotypic differences in N efficiency is differences in N uptake efficiency
after anthesis. Continuous root growth and N uptake activity are responsible for the high N uptake
efficiency of N-efficient genotypes. This study was conducted mainly to identify root parameters
which could be used as secondary selection traits for genotypic differences in N efficiency of maize
established in field experiments. The specific objective of the first experiment was to establish a
relationship between root parameters with genotypic differences in N efficiency in the field, and to
identify root traits to be used as secondary selection criteria for N efficiency. Four temperate and 16
tropical genotypes were grown in low-N nutrient solution with a cigar roll and a growth pouch culture
for 9 and 10 days, respectively. In the cigar roll experiment individual root fractions (adventitious,
seminal and primary root fractions) and in growth pouch experiment root distribution and root
branching angle were of primary interest. Genotypic differences were found in most of the root traits,
but the differences were not clear cut between N-efficient and inefficient genotypes with few
exceptions. The N-efficient genotypes had the highest percentage of root length in the deepest (>20
cm) interval in the growth pouch, which also positively correlated with N uptake after anthesis and
grain yield. The N-efficient genotypes also had a high percentage of roots in the root branching angle
interval of 60-90o. It was concluded that the high N uptake efficiency of N-efficient genotypes might
be related to a higher percentage of roots growing downwards (high branching angle) and a high
percentage of root length in deeper soil layers enabling them to exploit nitrate in the subsoil more
efficiently. These two root parameters were found promising to use as selection criteria for N
efficiency.

1. INTRODUCTION

Nitrogen (N) recovery efficiency in annual crops averages only about 42 and 29 % in
developed and developing countries, respectively [1]. This low N recovery efficiency is
associated with the loss of applied N by leaching, volatilization, denitrification and soil
erosion [2]. In addition, use of inadequate crop management practices, as well as biotic and
abiotic stresses are also responsible for low N use efficiency. Breeding of N-efficient cultivars
is one of the strategies to improve the N efficiency of crops.

Selection for grain yield at low N seems to be necessary, but encounters some
difficulties due to possible cultivar by N interactions [3]. With a decreasing level of N in the
soil, non-genetic variation of yield is increasing due to greater soil variability. Therefore,
selection based on additional plant traits contributing to N efficiency may be more precise and
enhance the selection progress [4]. A better knowledge of morphological and physiological
traits controlling N uptake and N utilization efficiencies in plants are essential for both
selection of cultivars and well-defined breeding strategies to improve the N efficiency of
crops [5]. Selection for root growth characteristics may provide a direct means for improving
N uptake [4].

223
 Nitrate-N is the most mobile nutrient in the soil. Plants develop different adaptations
to exploit this temporally and spatially/differentially available nutrient. Roots in nutrient-rich
zones can demonstrate both physiological and morphological alterations compared with those
roots outside the zone [6]. Increasing the root system is an adaptation to increase the total
volume of soil explored, and thereby increases the total amount of nitrate which the plant can
access [7, 8]. A highly significant correlation between the size of the root system and grain
yield at low N supply was observed [5]. Maize hybrids respond to N deficiency by increasing
total root length [9]. The N-efficient maize line ‘478’ showed higher values of total root
length at low N compared to an inefficient cultivar [10].

Maize is a cereal and has a fibrous root system; however it can be divided into three
distinct root fractions: primary root, seminal roots and adventitious roots. The primary root is
the first single root that emerges from the seed, also called the radicle. Seminal roots also
emerge from the seed, but later, and are numerous in number. Adventitious roots emerge from
the stalk nodes. However, it is not well known how N supply affects these different
components of the root system.

Nitrate supply has a pronounced influence on the total length of primary roots [10].
The primary root increased to the maximum at 0.4 mM level, and then declined sharply with
increasing nitrate supply. It was confirmed that longer primary roots were the main reason for
the enhanced root growth at low N [10]. When N supply is low uptake of N by maize plants
depends on a deeper root system resulting from longer primary roots.

Roots respond to localized nitrate supplies by proliferating lateral roots within the
nitrate-rich zone [11]. In barley, this ability is due to a combination of increased numbers of
lateral roots and increased rates of lateral root elongation [12]. Low-input cultivars of maize
develop more and longer seminal roots [13]. Since the seminal roots grow more in the
downward direction than adventitious roots [14], they may particularly contribute to the
exploration of the sub-soil during later stages of development. Maize hybrids respond to N
deficiency by increasing the elongation of individual seminal roots and enhancing lateral root
growth while reducing the number of seminal roots [9]. A negative correlation between yield
and root number, particularly at low N supply, has been reported [15].

Nitrogen deficiency resulted in significant decreases both in shoot and root fresh and
dry matter weight of young plants [16]. Low nitrogen generally suppresses shoot growth and
increases root: shoot ratios with and without increasing the root biomass in maize [9]. Higher
root dry matter at low N has been reported [13]. N deficiency induces higher dry matter
allocation to the roots during the vegetative stage [17]. Deficiency in N results in a shift in dry
matter allocation in favour of root growth [18]. In solution culture where N buffering ability is
low, however, it is common that low N decreases root growth [10, 19–21].

A plant’s ability to explore the soil and to compete effectively for soil resources is
critically dependent on architecture of the root system [22]. Root architecture is determined
by the pattern of root branching and by the rate of growth of individual roots [23]. These
properties of a root system are not only under genetic control but are also highly plastic, being
influenced by a wide range of physical, chemical and biological factors [11, 24, 25].
Simulation modelling revealed that deep-rooted architecture acquired more nitrate than a
shallow-rooted structure [7]. Root architecture is mainly determined by the distribution of
seminal roots, and seminal roots determine the architecture of the root system in the soil
volume [9].

224
 The fineness of a root is expressed as specific root length (length per unit weight of
root). Roots with high specific root lengths (i.e., fine roots) are often found in plants grown
under nutrient-deficient conditions [26]. Fine roots allow the root system to explore the soil
volume effectively while minimizing the investment needed to construct and maintain the root
system. On the other hand, thicker roots are more costly to produce, but have greater transport
capacity and are less vulnerable to desiccation and physical damage, and thus are generally
longer-lived [27]. The root diameter generally shows less plasticity than the stem diameter,
and in many species specific root length does not change significantly as a function of nutrient
availability [28]. In grasses, thicker roots and a high tissue density (dry matter per unit root
volume) have also been associated with increased longevity. In nutrient-poor environments
thicker roots with a longer lifespan may increase the residence time of nutrients in the plant,
and provide an important means of nutrient conservation [29].

Many authors have discussed the response of root parameters under low-N conditions,
but few reports have been published about the relationship of these root parameters to N
efficiency at low-N. This study was designed to test the suitability of root morphological traits
as secondary plant traits for N efficiency in short-term nutrient solution experiments.

2. MATERIALS AND METHODS

2.1. Experiment 1

Plants were grown in an environmentally controlled growth chamber of the Institute of

Plant Nutrition, Leibniz Universität Hannover, Germany. The growth chamber was adjusted
to 30 / 20oC day / night temperature and 80 % relative humidity.

2.2. Plant material

A total of 4 temperate and 16 tropical maize (Zea mays L.) cultivars were chosen to
study the relationship between root parameters and N efficiency. The experiment was
conducted with seedlings grown on filter paper in nutrient solution. The temperate cultivars
were obtained from KWS SAAT AG (Einbeck, Germany) and categorized according to N
efficiency by the breeder (Tables 1). The tropical cultivars were obtained mainly from
CIMMYT and categorized for their N efficiency by Worku [30] (Table 2).

TABLE 1. TEMPERATE MAIZE CULTIVARS USED IN THE STUDY

Number Cultivar Pedigree Efficiency class

1 TH TH x Tester N-inefficient
2 NUE NUE- N-inefficient
3 ASK ASKET N-efficient
4 SL SL x Tester N-efficient

225
TABLE 2. TROPICAL MAIZE CULTIVARS USED IN THE STUDY

Cultivar code Pedigree Efficiency class Source

C1 CML444/CML445//CML440 N-efficient CIMMYT-Zimbabwe
C2 CML395/CML444//CML440 N-efficient CIMMYT-Zimbabwe
C3 CML202/CML395//CML205 N-inefficient CIMMYT-Zimbabwe
C4 SC515 N-inefficient Seed-CO-Zimbabwe
C5 CML395/CML444//CML442 N-efficient CIMMYT-Zimbabwe
C6 CML444/CML197//CML443 N-efficient CIMMYT-Zimbabwe
C7 SC633 N-inefficient Seed-CO-Zimbabwe
C8 CML181/CZL01005//CZL01006 N-inefficient CIMMYT-Zimbabwe
C9 CML181/CML182//CML176 N-efficient CIMMYT-Zimbabwe
C10 CML144/(16304/6303Q)-B-6-1-3-3-B*6 N-inefficient CIMMYT-Mexico
C11 CML247/CML254 N-efficient CIMMYT-Mexico
C12 CML78/CML373 N-efficient CIMMYT-Mexico
C13 CML264/CML311//CML334 N-inefficient CIMMYT-Mexico
C14 CML442/CML444//[MSRXPL9]C1F2- N-efficient CIMMYT-Kenya
205-1(OSU23I)-1-1-X-1-X-B-B
C15 LPSC4F273-2-2-1-B-B- N-efficient CIMMYT-Kenya
B/CML202//CML384
C16 CML312/CML247//CML78 N-efficient CIMMYT-Kenya

2.3. The cigar roll experiment

Maize seeds were surface-sterilized with 10 % NaOCl solution for 2 minutes and then
soaked in deionized water for 4 hours. Germination was carried out on moistened filter paper
with the size of 23 x 35 cm. Seeds were placed at 1 cm below the top center of the filter paper
and rolled like a cigar. The cigar rolls were placed upright in an 18 l capacity bucket
containing 7 l of tap water and allowed to germinate. One seed placed per cigar roll represents
one replication and a total of 8 replications were used. Germinated seedlings with an
emerging radicle were transferred to a box filled with 5 l nutrient solution. The nutrient
solution was composed of (µM): K2SO4 (500); KH2PO4 (100); MgSO4 (325); NaCl (50);
H3BO3 (8); MnSO4 (1.0); ZnSO4 (0.4); CuSO4 (0.4); Na2MoO4 (0.1) and Fe-EDDHA (85).
Ca(NO3)2 with a concentration of 100 µM was used as a low N source. The cigar rolls were
placed randomly in the solution and the bucket was covered with aluminum foil to avoid
drying. The experimental layout was a randomized complete block design, each bucket
considered as a block. Plants were harvested 9 days after planting (DAP) for tropical and 10
DAP for temperate cultivars when they produced 3 visible leaves.

2.4. The growth pouch experiment

Growth pouches were made of filter paper (23 × 35 cm) placed in a randomly
punctured plastic bag. Seeds were sterilized and soaked using the same procedure as for the
cigar roll experiment.

Tropical cultivars were germinated in potato dextrose agar (PDA). The agar solution
was prepared in the following way: 5 g agar was added to a 500 ml flask containing 200 ml
tap water. The solution was placed in a microwave until it formed a clear solution and allowed
to cool. 100 ml of 1mM CaSO4 and 0.8 ml of 8 µM H3BO3 were added to the solution as a
source of Ca and B for the germinating seeds. The solution was filled to the final volume of
500 ml with tap water. The agar solution was poured into plastic plates. Maize seeds were
placed immediately into the agar solution before the agar solidified. The agar solution was
added to the plate until the entire seeds were covered. One slice of agar per seed was cut and

226
placed in petri dishes. The petri dishes were then placed in the dark in a growth chamber in an
upright position. Germinated seedlings together with the agar slices were fixed with pins on
moist filter paper at the top center position 3 DAP and covered with randomly punctured
plastic foil. A hole was made into the plastic pouch above the seeds for the shoots to emerge
and the pouches were frequently sprayed with water to avoid drying. The pouches were
hanged in a plastic box (42 × 30 × 27 cm size) containing 5 l nutrient solution using wooden
sticks. The experimental layout was a randomized complete block design, each box
considered as a block. The composition of the nutrient solution was the same as for the
cigarroll experiment described in section 2.3.

Except for germination, the same procedure was followed for both tropical and
temperate cultivars. Temperate cultivars were germinated in a sandwich experiment, because
fungal growth was observed for the tropical cultivars in the agar. Seeds were placed on
moistened filter paper at the top center position. The filter paper was covered with randomly
punctured plastic foil and placed between foam and PVC on both sides. The PVC plates were
fixed at the top and bottom side with plastic bands to avoid seed displacement. One seed per
pouch representing one replication was placed upright in a plastic box containing 7 l tap
water. After germination only the filter paper covered with plastic foil was transferred to a
nutrient solution of the same concentration as described above. Treatments were replicated 6
times for both tropical and temperate cultivars. Harvesting and measurements were done at 10
DAP when plants had 3 visible leaves.

2.5. Measurements

After harvesting, shoots and roots were separated. In the cigar roll experiment, each
root fraction (adventitious, seminal and primary roots) was counted and separated from the
root system for scanning. Primary root length was measured manually with a ruler. Scanning
was done by floating roots in a transparent plastic tray (30 × 20 × 2 cm size) filled with 1 cm
layer of water. Larger root samples that could not fit into a tray were cut and dispersed to
avoid overlapping. Images were analyzed for total root length, average root diameter, number
of root tips and root volume using WinRHIZO image analysis software (WIN MAC, Regent
Instruments, Inc.). Root tip diameter was measured on 1 cm root tips.

In addition to parameters taken in the cigar roll experiment, root branching angle from
the horizontal surface and root distribution were measured for the plants grown in growth
pouches. The branching angle of roots was measured in reference to the horizontal surface.
Two vertical lines were marked 2 cm left and right from the kernel, and the distance between
the roots crossing these lines and the horizontal layer at the height of the kernel was measured
with a ruler. The branching angles were calculated according to the formula:

Tan (root branching angle) = distance from the horizontal layer (cm) / 2 cm

Root distribution was analyzed by dividing the whole length of the growth pouch into
6 different layers each of 4 cm length with the exception of the bottom layer. The intact root
system was separated carefully from the filter paper by moistening the paper with water and
scanned as a digital image using the WinRHIZO scanner. The root lengths of the 6 different
layers were analyzed separately.

Other root parameters were derived from the measured root data; tips per length was
calculated by dividing the number of root tips by the total root length (TRL); specific root
length is the ratio of TRL and root dry weight; tissue density was calculated by dividing the
root dry weight to the root volume. Root and shoot dry weights were measured after oven

227
drying the sample for 48 hr at 60 oC. Data on biomass N uptake at anthesis, post-anthesis, at
harvest and grain yield were obtained from Worku [30]. The field experiments were carried
out in Harare, Zimbabwe (2003 and 2004). The 16 maize cultivars were tested using three N
treatments (severe N stress - N1, medium-N stress - N2, and optimum N supply - N3).

2.6. Statistical analysis

The results of the cigar roll and growth pouch experiments were analyzed as a
randomly complete block design (RCBD). The General Linear Model Procedure (GLM) of
the Statistical Analysis Systems (SAS) Package (SAS Institute Inc., 2002-2003) was used for
statistical analysis of all root parameters. When significant genotypic differences (P<0.05)
occurred as indicated by the ANOVA result, the Tukey test was used for comparing
individual means. r2 values were obtained from Sigma plot (Version 8.0). Finally simple
correlation coefficients were calculated to assess the relationships between root parameters
and N uptake and grain yield of the same tropical cultivars in field experiments conducted in
Zimbabwe at low-N stress in 2003 and 2004.

3. RESULTS

3.1. Performance of tropical maize cultivars

3.1.1. Root parameters measured both in the cigar roll and growth pouch experiments

3.1.1.1. Shoot and root dry weight

Genotypic variations (P<0.001) were detected in shoot and root dry weight per plant
in the cigar roll experiment (Fig. 1A). The N-efficient cultivars C5 and C1 accumulated the
highest shoot dry weight in comparison to N-efficient C14, C15 and N-inefficient C4 and
C10. Mean shoot dry weight of efficient cultivars was higher than for inefficient cultivars.
The root dry weight of the efficient cultivars C5 and C2 was higher than that of the efficient
cultivar C11 and the inefficient cultivar C10. Genotypic variation was also observed (P<
0.01) in shoot dry weight in the growth pouch experiment (Fig. 1A). N-efficient cultivars C6,
C1 and C12 produced a higher shoot dry weight compared to the N-efficient cultivar C15. No
variation was detected in root dry weight among the cultivars. However, N-efficient cultivars
tended to produced higher root dry weight than inefficient cultivars. In all cultivars, higher
shoot than root dry weight was observed in the cigar roll experiment, while in the growth
pouch experiment shoot dry weight was lower than root dry weight.

3.1.1.2. Total root length

No genotypic variation was detected in total root length per plant in the cigar roll
experiment (Fig. 1B). However, N-efficient cultivars tended to produce a higher mean total
root length (307.5 cm) than inefficient cultivars (294.1 cm). However, cultivars differed in
root length (P<0.001) in the growth pouch experiment (Fig.1B). Higher root length was
achieved by N-inefficient C8 and N-efficient C1 cultivars in comparison to N-efficient
cultivars C16, C14, C15 and N-inefficient cultivars C13 and C7. Comparing the two
experiments, higher mean total root length (302.7 cm) was observed in the cigar roll
experiment compared to growth pouch experiment (257.1 cm).

3.1.1.3. Primary root-axis length

228
 Genotypic variation (P<0.01) was observed in the primary root-axis length in the cigar
roll experiment (Fig. 1C). The N-efficient cultivar C1 produced the longest while another N-
efficient cultivar C16 and N-inefficient cultivar C3 produced the shortest primary root.
Genotypic variation was also detected in the growth pouch experiment (Fig. 1C). N-efficient
cultivar C5 produced higher primary root length compared to another N-efficient cultivar
C14. Comparing the two experiments, a higher primary root length (35.8 cm) was observed in
the cigar roll compared to the growth pouch (30.5 cm) experiment.

3.1.1.4 . Number of adventitious and seminal roots

Genotypic variation (P<0.001) was observed in the number of adventitious and

seminal roots per plant in the cigar roll experiment (Fig. 1D). The N-inefficient cultivars C7,
C4, C8 and N-efficient cultivars C2, C6 and C12 produced more adventitious roots than the
N-inefficient cultivar C10. The N-efficient cultivar C2 and N-inefficient cultivar C13
produced more seminal roots compared to the N-efficient cultivars C11, C9 and N-inefficient
cultivars C8 and C4. The mean number of seminal roots produced by efficient cultivars was
higher (5.4) than by inefficient cultivars (4.9). Genotypic variation was also detected (P<
0.01) in the growth pouch experiment (Fig. 1D). The highest number of adventitious roots
was recorded by N-efficient cultivars C1 and C2 compared to the N-inefficient cultivar C13.
The same efficient cultivars C1 and C2 also produced the highest number of seminal roots
compared to the N-efficient cultivars C9, C14, C16 and N-inefficient cultivar C10.

3.1.1.5. Average root diameter

Genotypic variation was observed in average root diameter per plant in the cigar roll
experiment (Fig. 2A). N-efficient cultivars C1, C9 and C2 produced the greatest root diameter
compared to another efficient cultivar C14. Genotypic variation was also detected (P<0.001)
in average root diameter per plant in the growth pouch experiment (Fig. 2A). The N-
inefficient cultivar C13 and N-efficient cultivar C15 produced a higher root diameter than N-
inefficient cultivars C7 and C8. Higher mean root diameter was recorded in growth pouches
(0.647 mm) compared to the cigar rolls (0.584 mm).

3.1.1.6. Tips per length

Genotypic variation (P<0.01) was observed in tips per length in the cigar roll
experiment (Fig. 2B). The N-efficient cultivar C1 produced the highest number of tips per
root length in comparison to the other efficient cultivars C11, C5, C12 and C14. Genotypic
variation (P<0.001) was also detected in the growth pouch experiment (Fig. 2B). Higher tips
per root length were exhibited by the N-efficient cultivars C1, C9 and the N-inefficient
cultivars C13, C10 and C3 compared to the N-efficient cultivar C16. Higher tips per length
was observed in the growth pouch (3.77 tips cm-1) compared to the cigar roll experiment (2.78
tips cm-1).

3.1.1.7. Specific root length

Genotypic differences in specific root length existed in the cigar roll experiment (Fig.
2C). The N-efficient cultivar C11 had a higher specific root length compared to the other
efficient cultivar C2. Genotypic variation (P<0.01) was also observed in the growth pouch
experiment (Fig. 2C). Higher specific root length was achieved by the N-efficient cultivar
C11 in comparison to another N-efficient cultivar C16 and N-inefficient cultivar C13. The
mean specific root length of the growth pouch experiment (60.4 m g -1) was higher than the
cigar roll experiment (29.1 m g-1).

229
 Cigar roll Growth pouch

0.20
A 0.20 F-test
F-test N-efficient Adv.
A
Shoot and root dry weight (g)

N-efficient Seminal Cult (shoot): ***

0.15 Adv.
N-inefficient
0.15 Cult (shoot):*** Cult (root): ns
N-inefficient Seminal
Cult (root): *** N-efficient (shoot)
N-efficient (root)
0.10 0.10
N-inefficient (shoot)
N-inefficient (root)

0.05 0.05

0.00 0.00
C16C12C2C15C9 C5 C1 C6C11C14C3 C4 C7 C8C13C10 C16C12 C2 C15 C9 C5 C1 C6 C11C14 C3 C4 C7 C8 C13C10

600 500 F-test

F-test
B
Total root length (cm) per plant

500
400 Cult: ***
Cult: ns
400
N-efficient
300
N-inefficient
300
200
200

100
100

0 0
C16C12 C2 C15 C9 C5 C1 C6 C11C14 C3 C4 C7 C8 C13C10 C16C12 C2 C15 C9 C5 C1 C6 C11C14 C3 C4 C7 C8 C13C10

60 50
F-test F-test
C
Primary root axis length (cm)

50
Cult: ** 40 Cult: *
40
30
N-efficient
N-inefficient
30
20
20

10
10

0 0
C16C12 C2 C15 C9 C5 C1 C6 C11C14 C3 C4 C7 C8 C13C10 C16C12 C2 C15 C9 C5 C1 C6 C11C14 C3 C4 C7 C8 C13C10

10 12
F-test F-test
D 10
8 Cult (adv): *** Cult (adv): **
Cult (seminal):*** 8 Cult (seminal): ***
Root number

6
N-efficient Adv.
6
N-efficient Seminal
4 N-inefficient Adv.
4 N-inefficient Seminal

2
2

0 0
C16C12 C2 C15 C9 C5 C1 C6 C11C14 C3 C4 C7 C8 C13C10
C16C12 C2 C15 C9 C5 C1 C6 C11C14 C3 C4 C7 C8 C13C10
Cultivar
Cultivar
FIG. 1. Shoot and root dry weight (A), total root length (B), primary root-axis length (C) and
adventitious and seminal root number (D) of 16 tropical maize cultivars grown in the cigar roll (n =
8) at 9 DAP and in the growth pouch experiment (n = 6) at 10 DAP at low N supply (100 µM N).
Cultivars are arranged according to descending N efficiency in the field experiments from the left to
the right side.

230
3.1.1.8. Tissue density

Genotypic variation in tissue density was not detected in the cigar roll experiment
(Fig. 2D) due to the observed large standard deviation. However, the N-inefficient cultivars
C4, C7 and C8 and efficient cultivars C5 and C9 tended to have a higher tissue density than
other cultivars. Similarly, no genotypic variation was detected in the growth pouch
experiment (Fig. 2D). Higher mean tissue density was recorded from plants grown in the
growth pouch (0.12 g cm-3) compared to those grown in the cigar roll experiment (0.07 g
cm-3).

3.1.2. Root parameters measured only in the cigar roll experiment

3.1.2.1. Relative root length of different root fractions

Cultivars were evaluated for the relative contribution of adventitious, seminal and
primary roots to the total root length, but no genotypic differences were detected (Fig. 3A).
However, N-inefficient cultivars tended to have higher mean percentages of adventitious and
primary root length compared to N-inefficient cultivars. On the other hand, N-efficient
cultivars tended to produce a higher percentage of seminal roots compared to N-inefficient
cultivars.

3.1.2.2. Number of tips of the different root fractions

Genotypic differences (P<0.01) were detected in the number of adventitious, seminal

and primary root tips per plant (Fig. 3B). The N-efficient cultivar C8 had more adventitious
root tips than the other N-efficient cultivars C13 and C10. The N-efficient cultivars C2, C6
and C1 produced a higher number of seminal roots tips compared to the N-efficient cultivars
C12, C14 and N-inefficient cultivar C8. A higher number of primary root tips was recorded
by N-efficient cultivar C1 when compared to the other N-efficient cultivars C12, C15, C14
and the N-inefficient cultivar C13. The mean number of primary root tips of efficient cultivars
were higher (266.7) than of the inefficient cultivars (236.1).

3.1.2.3. Average diameter of root fractions

Genotypic differences existed in mean diameter of adventitious and seminal roots, but
no variation was detected in mean diameter of primary roots (Fig. 3C). The N-inefficient
cultivar C3 had a higher diameter of adventitious roots than the N-efficient cultivar C15.
Higher diameters of seminal roots were recorded by the efficient cultivars C2, C9, C1, C16,
C6, C15, C5 and N-inefficient cultivars C3, C13, C10, C4, and a small diameter was
measured for the N-efficient cultivar C14. In general, the mean diameters of primary roots
were less than that of adventitious roots.

231
 Growth pouch
F-test F-test
0.8 1.0
A Cult: * Cult: ***
Average root diameter (mm)

0.8
0.6
N-efficient
0.6 N-inefficient
0.4

0.4

0.2
0.2

0.0 0.0
C16C12 C2 C15 C9 C5 C1 C6 C11C14 C3 C4 C7 C8 C13C10 C16C12 C2 C15 C9 C5 C1 C6 C11C14 C3 C4 C7 C8 C13C10

8
8
F-test F-test
B
Tips per length (tips cm-1)

6 6
Cult: ** Cult: ***
N-efficient
4 N-inefficient
4

2 2

0 0
C16C12 C2 C15 C9 C5 C1 C6 C11C14 C3 C4 C7 C8 C13C10 C16C12 C2 C15 C9 C5 C1 C6 C11C14 C3 C4 C7 C8 C13C10

140
60

120 F-test F-test

C 50
Specific root length (m g-1)

100 Cult: * Cult: **

40
80
N-efficient
30 N-inefficient
60

20
40

20 10

0 0
C16C12 C2 C15 C9 C5 C1 C6 C11C14 C3 C4 C7 C8 C13C10 C16C12 C2 C15 C9 C5 C1 C6 C11C14 C3 C4 C7 C8 C13C10

0.16 0.25

0.14
F-test
D 0.20 F-test
Tissue density (g cm -3)

0.12
Cult: ns Cult: ns
0.10 0.15
0.08 N-efficient
0.10 N-inefficient
0.06

0.04
0.05
0.02

0.00 0.00
C16C12 C2 C15 C9 C5 C1 C6 C11C14 C3 C4 C7 C8 C13C10 C16C12 C2 C15 C9 C5 C1 C6 C11C14 C3 C4 C7 C8 C13C10
Cultivar Cultivar
FIG. 2. Average root diameter (A), tips per length (B), specific root length (C) and tissue density (D)
of 16 tropical maize cultivars grown in the cigar roll (n = 8) at 9 DAP and in the growth pouch
experiment (n = 6) at 10 DAP at low N supply (100 µM N). Cultivars are arranged according to
descending N efficiency in the field experiments from the left to the right side.

232
 100
F-test
A
80 Cult (adv): ns
Cult (seminal): ns
Root length (%)

Cult (primary): ns
60

Adv. roots
Seminal roots
40 Primary roots

20

0
C16 C12 C2 C15 C9 C5 C1 C6 C11 C14 C3 C4 C7 C8 C13 C10

Cultivar
1000 F-test
B
Cult (adv): **
800
Cult (seminal): ***
Number of root tips

Cult (primary): ***

600

N-efficient Adv.
N-efficient Seminal
400
N-efficient Primary
N-inefficient Adv.
200 N-inefficient Seminal
N-inefficient Primary

0
C16 C12 C2 C15 C9 C5 C1 C6 C11 C14 C3 C4 C7 C8 C13 C10

Cultivar
1.0
F-test
C
Cult (adv): *
0.8
Cult (seminal): *
Average diameter (mm)

Cult (primary): ns
0.6

N-efficient Adv.
N-efficient Seminal
0.4 N-efficient Primary
N-inefficient Adv.
N-inefficient Seminal
N-inefficient Primary
0.2

0.0
C16 C12 C2 C15 C9 C5 C1 C6 C11 C14 C3 C4 C7 C8 C13 C10

Cultivar
FIG. 3. Root length (%) of the total root length per plant of the different root fractions (A), number of
root tips (B) and average diameter of adventitious, seminal and primary roots (C) of 16 tropical maize
cultivars grown in the cigar roll experiment (n = 8) at 9 DAP at low N supply (100 µM N). Cultivars
are arranged according to descending N efficiency in the field experiments from the left to the right
side.

233
3.1.2.4. Root tip diameter of root fractions

Root tip diameter was analyzed from root tips of 1 cm length cut of individual root
fractions. No genotypic variation was detected in root-tip diameter of adventitious roots while
genotypic variation (P<0.01) was observed for seminal root-tip diameter (Fig. 4). A higher
seminal root-tip diameter was recorded by N-efficient cultivar C12 in comparison to N-
inefficient cultivars C10 and N-efficient cultivars C1, C16, C15, C14 and C9.

0.8

F-test
Root tip diameter (mm)

0.6
Cult (adv): ns
Cult (seminal): **
0.4
N-efficient Adv.
N-efficient seminal
N-inefficient adv.
0.2 N-inefficient seminal

0.0
C16 C12 C2 C15 C9 C5 C1 C6 C11 C14 C3 C4 C7 C8 C13 C10

Cultivar

FIG. 4. Root tip diameter of adventitious and seminal roots of 16 tropical maize cultivars grown in the
cigar roll experiment (n = 8) at 9 DAP at low N supply (100 µM N). Cultivars are arranged according
to descending N efficiency in the field experiments from the left to the right side.

3.1.3. Root parameters measured only in the growth pouch experiment

3.1.3.1. Root length (%) distribution at different root-systems depth

Cultivars were evaluated in the percentage of root length produced in each layer of the
growth pouch, which was partitioned into six layers with 4 cm interval from the surface.
Genotypic variation was observed in all growth pouch layers except the fourth layer (12-16
cm from the surface) (Fig. 5A). In the first 0-4 cm interval, the N-inefficient cultivar C7
produced a higher (P<0.01) percentage of root length than the N-efficient cultivars C11, C5,
C2, C6, C12, C1 and N-inefficient cultivars C13, C3 and C10. In the deepest growth pouch
layer (>20 cm), the N-efficient cultivars C12, C6, C9 and C1 produced a higher (P<0.001)
percentage of root length than the N-inefficient cultivar C8. Generally, N-inefficient cultivars
tended to produce a higher percentage of root length in the upper layer and N-efficient
cultivars in the lowest layer of the growth pouch.

3.1.3.2. Root tips (%) distribution at different root-systems depth

Cultivars were also evaluated for the percentage of root tips in each layer of the
growth pouch. Genotypic variation was detected only in the first two layers (0-4 and 4-8 cm)
(Fig. 5B). In the 0-4 cm interval, the N-inefficient cultivar C7 produced a higher (P<0.001)
percentage of root tips than all other cultivars with the exception of cultivar C4. N-efficient
cultivar C15 produced a higher (P<0.01) percentage of root tips in the growth pouch layer of
4-8 cm than N-inefficient cultivar C3 and N-efficient cultivar C1. No genotypic variation was

234
detected in the remaining growth pouch layers, but the mean percentage of root tips of the
efficient cultivars tended to be higher than that of inefficient cultivars in the last two growth
pouch layers. There were no differences in the remaining intervals.

3.1.3.3. Root length (%) in different diameter classes

Genotypic differences were detected (P<0.001) in percentage of root length of roots

with a diameter class of 0-1 mm, 1-2 mm and >2mm (Fig. 5C). The dominant root types were
fine roots with a diameter class of 0-1 mm comprising about 83.3% followed by roots in a
diameter class of 1-2 mm (13.5%) and >2 mm (2.7%). In the diameter class of 0-1 mm, a
higher percentage of roots were produced by the N-inefficient cultivars C8, C4 and N-
efficient cultivars C16, C14, C15 and C11 in comparison to N-inefficient cultivars C13 and
C3. In the diameter class of 1-2 mm, the N-inefficient cultivars C3 and C13 produced a higher
percentage of roots than N-efficient cultivars C16, C5, C14 and N-inefficient cultivars C4 and
C8. In the diameter class >2 mm diameter, the N-efficient cultivars C1 had a higher
percentage of roots than N-efficient cultivars C16, C12 and N-inefficient cultivars C8 and C4.

3.1.3.4. Roots branching angle

Cultivars were evaluated for the percentage of roots they produced in a branching
angle interval of 0-30o, 30-60o and 60-90o (Fig. 6). No genotypic variation was detected in a
root branching angle interval of 0-30o, while genotypic variation was observed in the
branching angle intervals of 30-60o and 60-90o. In the root branching interval of 30-60o, a
high percentage of roots was exhibited by the inefficient cultivar C7 compared to the efficient
cultivar C16. The N-efficient cultivars C16 and C2 produced higher percentage of roots in a
branching angle interval of 60-90o compared to the inefficient cultivar C7.

3.3. Relationships between genotypic performances in short term nutrient solution in

the cigar roll and field experiments

Correlation were calculated to observe the relationship between the cigar roll
experiment parameters with N uptake and grain yield of this cultivars grown at low N in field
experiments in Zimbabwe (Table 3). Total root length was positively and significantly
correlated with N uptake at anthesis. Adventitious and seminal root numbers were not
significantly correlated with N uptake at anthesis. A significant negative correlation was
detected between average root diameter and N uptake at anthesis. None of the parameters
showed correlations with N uptake after anthesis and grain yield.

No significant correlations were observed between root length in different layers of

the growth pouch and N uptake at and after anthesis (Table 4). However, a trend for a positive
correlation was detected between relative root length >20 cm and N uptake after anthesis and
grain yield. No correlation was observed between roots within 0-30 o branching angle interval
and N uptake after anthesis and grain yield. Percentage of roots within a branching angle
interval of 30-60o were significantly (P<0.1) and negatively correlated with N uptake after
anthesis. A tendency of negative correlation was also observed between the percentage of
roots within this branching angle interval and grain yield. A significant (P<0.1) positive
correlation was observed between percentage of roots within the branching angle interval of
60-90o and N uptake after anthesis. In addition to this, a trend for a positive correlation was
observed between the percentage of roots within the branching angle interval of 60-90o and
grain yield.

235
 100
A F-test

80 Cult (0-4 cm): **

Cult (4-8 cm): *
Cult (8-12 cm): ***
Root length (%)

60 Cult (12-16 cm): ns

Cult (16-20 cm): **
Cult (> 20 cm): ***
40
0-4 cm
4-8 cm
8-12 cm
12-16 cm
20 16-20 cm
> 20 cm

0
C16C12 C2 C15 C9 C5 C1 C6 C11C14 C3 C4 C7 C8 C13C10

100 F-test
B
Cult (0-4 cm): ***
80
Cult (4-8 cm): **
Cult (8-12 cm): ns
Cult (12-16 cm): ns
Root tips (%)

60
Cult (16-20 cm): ns
Cult (> 20 cm): ns
40 0-4 cm
4-8 cm
8-12 cm
12-16 cm
16-20 cm
20 > 20 cm

0
C16C12 C2 C15 C9 C5 C1 C6 C11C14 C3 C4 C7 C8 C13C10
Cultivar

100
Root length (%) by different diameter classes

C F-test

80 Cult (0-1 mm): ***

Cult (1-2 mm): ***
Cult (> 2 mm): ***
60
0-1 mm
1-2 mm
> 2 mm
40

20

0
C16C12 C2 C15 C9 C5 C1 C6 C11C14 C3 C4 C7 C8 C13C10

Cultivar

FIG. 5. Root length (%) of the total root length per plant (A), root tips (%) (B) at different root-
systems depth from the surface and root length (%) in different diameter classes (C) of 16 tropical
maize cultivars grown in growth pouches (n = 6) at 10 DAP at low N supply (100 µM N). Cultivars
are arranged according to descending N efficiency in the field experiments from the left to the right
side.

236
Number of roots (%) within specified angle interval
100
F-test

Cult (0-30 o): ns

80 Cult (30-60 o): *
Cult (60-90 o): *
0-30o
60 30-60 o
60-90 o

40

20

0
C16 C12 C2 C15 C9 C5 C1 C6 C11 C14 C3 C4 C7 C8 C13 C10
Cultivar
FIG. 6. Number of roots (%) within a specified branching angle interval of 16 tropical maize cultivars
grown in growth pouch experiment (n=6) at 10 DAP at low N supply (100 µM N). Cultivars are
arranged according to descending N efficiency in the field experiments from the left to the right side.

TABLE 3. CORRELATION COEFFICIENTS BETWEEN THE CIGAR ROLL ROOT-

PARAMETERS AND N UPTAKE AND GRAIN YIELD OF THE SAME TROPICAL MAIZE
CULTIVARS AT LOW N IN THE ZIMBABWE FIELD EXPERIMENTS

Parameter N uptake Grain yield

Antheses Post-antheses Harvest
Total root length (cm) 0.50* –0.27 0.24 0.16
Adventitious root number –0.15 0.11 –0.04 0.39
Seminal root number –0.07 0.25 0.17 0.13
Average root diameter (mm) –0.52* 0.23 –0.30 0.10
* denotes significance at P<0.05

TABLE 4. CORRELATION COEFFICIENTS (R) BETWEEN THE GROWTH POUCH ROOT-

PARAMETERS AND N UPTAKE AND GRAIN YIELD OF THE SAME TROPICAL CULTIVARS
AT LOW N IN THE ZIMBABWE FIELD EXPERIMENTS

Parameter N uptake
Anthesis Post-anthesis Harvest Grain yield
Relative root length (0-4 cm) 0.07 –0.10 –0.03 –0.03
Relative root length (4-8 cm) 0.29 –0.28 0.01 –0.22
Relative root length (8-12 cm) –0.05 –0.22 –0.27 –0.30
Relative root length (12-16 cm) –0.20 –0.08 –0.28 –0.08
Relative root length (16-20 cm) –0.31 0.31 0.00 0.22
Relative root length (>20 cm) –0.04 0.40 0.36 0.35
Relative root branching angle (0-30o) 0.09 –0.22 –0.13 –0.22
Relative root branching angle (30-60o) 0.16 –0.46+ –0.30 –0.42
Relative root branching angle (60-90o) –0.16 0.45+ 0.29 0.41
+ denotes significance at P<0.1

237
3.4. Performance of temperate maize cultivars

3.4.1. Root parameters measured in both the cigar roll and growth pouch experiments

3.4.1.1. Shoot and root dry weight

Genotypic variation (P<0.01) was observed in shoot dry weight per plant in the cigar
roll experiment (Fig. 7A). The N-inefficient cultivar TH produced a higher shoot dry weight
than the N-efficient cultivars ASK and SL and the N-inefficient cultivar NUE. No genotypic
variation was observed in root dry weight, but the N-inefficient cultivars tended to produce a
higher mean root dry weight than the N-efficient cultivars. Also, in the growth pouch
experiment, genotypic variation was observed (P<0.001) in shoot dry weight (Fig. 7A). The
N-inefficient cultivar TH produced higher shoot dry weight than the N-efficient cultivars
ASK, SL and the N-inefficient cultivar NUE. No genotypic variation was detected in root dry
weight. Similar to tropical cultivars, shoot dry weight of cultivars were higher than root dry
weight in the cigar roll, but the opposite holds true for the growth pouches.

3.4.1.2. Total root length

No genotypic variation was observed in total root length per plant both in the cigar roll
and growth pouch experiments (Fig. 7B). Comparing the two experiments, higher total root
length was observed in the growth pouch experiment.

3.4.1.2. Primary root axis length

No genotypic variation was observed in primary root length in the cigar rolls while
significant (P<0.001) variation between cultivars was observed in the growth pouch
experiment (Fig. 7C). Longer primary roots were produced by the N-inefficient cultivar TH
compared to the other three cultivars in growth pouches. Comparing the two experiments, a
longer primary root length was observed in the growth pouch experiment (45.9 vs 31.7 cm).

3.4.1.3. Number of adventitious and seminal roots

Genotypic variation existed (P<0.01) in the number of adventitious roots per plant in
the cigar roll experiment (Fig. 7D). The N-inefficient cultivar NUE and N-efficient cultivar
SL had a higher number of adventitious roots than the N-efficient cultivar ASK. No variation
was observed between cultivars in the number of seminal roots per plant in this experiment.
Genotypic variation in the number of adventitious roots was detected in the growth pouch
experiment (Fig. 7D). The two N-inefficient cultivars NUE and TH produced a higher number
of adventitious roots than the N-efficient cultivars SL and ASK. No variation was observed
between cultivars in the number of seminal roots per plant. On average, the N-inefficient
cultivars produced a slightly higher number of seminal roots compared to the efficient
cultivars. Comparing the two experiments, a higher mean number of adventitious roots was
produced in the growth pouch experiment (3.5 roots) compared to the cigar rolls (2.8 roots),
but the number of seminal roots was lower in the growth pouch (4.8 roots) than in the cigar
roll experiment (5.7 roots).

3.4.1.4. Average root diameter

No genotypic variation existed in average root diameter in the cigar roll experiment
(Fig. 8A). Genotypic variation (P<0.01) in this trait was observed in the growth pouch

238
 experiment (Fig. 8A). The N-inefficient cultivar NUE produced thicker roots than the N-
efficient cultivar ASK.

Cigar roll Growth pouch

0.10 0.16
F-test F-test
A 0.14
0.08
Cult (shoot): ** 0.12 Cult (shoot): ***
Cult (root): ns Cult (root): ns
Dry weight (g)

0.06 0.10
N-efficient (shoot)
0.08 N-efficient (root)
N-inefficient (shoot)
0.04 0.06 N-inefficient (root)

0.04
0.02
0.02

0.00 0.00
TH NUE ASK SL TH NUE ASK SL

600 400
F-test F-test
500
B
Total root length (cm)

Cult: ns 300 Cult: ns

400

N-efficient
300 200
N-inefficient

200
100
100

0 0
TH NUE ASK SL TH NUE ASK SL

40 60
F-test F-test
C
Primary root axis length (cm)

50
30 Cult: ns Cult: ***
40

N-efficient
20 30
N-inefficient
20
10
10

0 0
TH NUE ASK SL TH NUE ASK SL

8 8
F-test F-test
D
6 6 Cult (Adv): ***
Cult (Adv): **
Root number

Cult (Seminal): ns Cult (Seminal): ns

4 4 N-efficient Adv.
N-efficient Seminal
N-inefficient Adv.
2 2 N-inefficient Seminal

0 0
TH NUE ASK SL TH NUE ASK SL
Cultivar Cultivar

FIG. 7. Shoot and root dry weight (A), total root length (B), primary root-axis length (C) and
adventitious and seminal root number (D) of 4 temperate maize cultivars grown in the cigar roll (left
side) (n = 8) and growth pouch experiment (n = 6) at 10 DAP at low N supply (100 µM N).

239
3.4.1.5. Tips per root length

No genotypic variation was detected in tips per length in the cigar roll experiment
(Fig. 8B). Genotypic variation was observed (P<0.001) in the growth pouch experiment (Fig.
8B). The N-inefficient cultivar NUE produced a higher number of tips per root length than the
N-efficient cultivars ASK, SL and the N-inefficient cultivar TH. Generally, higher tip
numbers per length were observed in the cigar roll (2.6 tips cm-1) compared to the growth
pouch experiment (1.6 tips cm-1).

3.4.1.6. Specific root length

Specific root length is the total length of roots per unit of root dry weight. No
genotypic variation was detected in specific root length in both the cigar roll and growth
pouch experiments (Fig. 8C). Comparing the two experiments, a higher specific root length
was exhibited in the cigar roll (40.9 m g-1) compared to the growth pouch experiment (33.2 m
g-1).

3.4.1.7. Tissue density

Cultivars did not differ in tissue density in both the cigar roll and growth pouch
experiments (Fig. 8D). Nevertheless, N-efficient cultivar SL tended to have higher tissue
density than other cultivars in both experiments.

3.4.1.8. Root tip diameter of adventitious, seminal and primary roots

No variations were observed among cultivars in root tip diameter of adventitious,

seminal and primary roots both in the cigar roll and growth pouch experiments (Figs. 9A and
9B).

3.4.2. Root parameters measured only in the cigar roll experiment

3.4.2.1. Relative root length of different root fractions

Cultivars were evaluated for the relative contribution of root fractions to the total root
length in the cigar roll experiment (Fig. 10A). Genotypic variation (P<0.001) was detected in
relative contribution of adventitious root length to the total root length. The N-inefficient
cultivar NUE produced a higher percentage of adventitious root length than the N-efficient
cultivars SL and ASK. No variation was detected in the relative contribution of seminal and
primary root length to the total root length.

3.4.2.2. Number of tips of different root fractions

Genotypic variations existed in the number of adventitious root tips (Fig. 10C). The
N-inefficient cultivar NUE produced a higher number of root tips than the N-efficient cultivar
ASK. No variation was detected in the number of seminal and primary root tips.

3.4.2.3. Average diameter of root fractions

Cultivars were also evaluated for their thickness of adventitious, seminal and primary
roots but no genotypic variation was detected (Fig. 10B).

240
 Cigar roll Growth pouch
0.8 0.8
F-test F-test
A
Average root diameter (mm)

0.6 Cult: ns 0.6 Cult: **

0.4
N-efficient
0.4
N-inefficient

0.2 0.2

0.0 0.0
TH NUE ASK SL TH NUE ASK SL

4 2.5

B F-test F-test
Tips per length (tips cm -1 )

2.0
3
Cult: ns Cult: ***
1.5
2 N-efficient
1.0 N-inefficient

1
0.5

0 0.0
TH NUE ASK SL TH NUE ASK SL

80 50
F-test F-test
C
Specific root length (cm g-1)

40
60 Cult: ns Cult: ns
30 N-efficient
40 N-inefficient
20

20
10

0 0
TH NUE ASK SL TH NUE ASK SL

0.12 0.20
F-test F-test
0.10
D
Tissue density (g cm-3)

Cult: ns 0.15 Cult: ns

0.08
N-efficient
0.10
N-inefficient
0.06

0.04
0.05
0.02

0.00 0.00
TH NUE ASK SL TH NUE ASK SL
Cultivar Cultivar

FIG. 8. Average root diameter (A), tips per length (B), specific root length (C) and tissue density (D)
of 4 temperate maize cultivars grown in the cigar roll (left side) (n = 8) and growth pouch experiment
(n = 6) at 10 DAP at low N supply (100 µM N).

241
 0.8 1.0
F-test F-test
C Cult (Adv): ns
B Cult (Adv): ns
0.8
Root tip diameter (mm)

0.6 Cult (Sem): ns Cult (Sem): ns

Cult (Prim): ns Cult (Prim): ns
0.6

0.4
N-efficient Adv.
0.4
N-efficient Seminal
N-efficient Primary
0.2 N-inefficient Adv.
0.2
N-inefficient Seminal
N-inefficient Primary
0.0 0.0
TH NUE ASK SL TH NUE ASK SL
Cultivar Cultivar

FIG. 9. Root tip diameter of 4 temperate maize cultivars grown in the cigar roll (A) (n = 8) and
growth pouch experiments (B) (n = 6) at 10 DAP at low N supply (100 µM N).

3.4.3. Root parameters measured only in the growth pouch experiment

3.4.3.1. Root length (%) distribution at different depth intervals from the surface

Cultivars were evaluated on the percentage of root length produced in each layer of
the growth pouch, which was partitioned into six layers with 4 cm intervals. Genotypic
variation was found in the first two layers (0–4 and 4–8 cm) from the surface (Fig. 11A). In
the 0–4 cm interval, the N-inefficient cultivar NUE and the N-efficient cultivar SL produced a
higher percentage of root length than the N-inefficient cultivar TH. In the 4–8 cm interval, the
N-efficient cultivars ASK and SL produced a higher percentage of root length (P<0.001) than
the N-inefficient cultivar TH. No genotypic variation existed in the remaining growth pouch
layers.

3.4.3.2. Root tips (%) distribution at different root-systems depth from the surface

Genotypic variation was observed (P<0.01) in the relative number of root tips in the
growth pouch layer of 0–4 cm (Fig. 11B). The N-inefficient cultivar NUE produced a higher
percentage of root tips than the N-efficient cultivar ASK and the N-inefficient cultivar TH.
The N-efficient cultivar ASK produced a higher percentage of root tips than the N-inefficient
cultivar TH in the growth pouch layer of 4–8 cm. No genotypic variation was detected in the
growth pouch layers of 8–12, 12–16 and 16–20 cm. In the last growth pouch layer (>20 cm)
N-inefficient cultivar TH produced a higher percentage of root tips than the other N-
inefficient cultivar NUE.

3.4.3.3. Root length (%) in different diameter classes

No genotypic variation existed in the percentage of roots in the diameter classes of 0–

1, 1–2 and >2 mm (Fig. 11C). More than 88 % of the total root length consisted of roots
within the 0–1 mm diameter class, followed by the diameter classes of 1–2 mm and >2 mm
with 10.5 and 0.5%, respectively.

3.4.3.4. Root branching angle

Cultivars were evaluated for the relative number of roots within the root branching
angle interval of 0–30, 30–60 and 60–90o from the surface (Fig. 11D). No genotypic variation
was observed in any root branching interval. However, N-inefficient cultivars tended to
produce a higher percentage of roots in a root branching interval of 0–30o and 30–60o, while

242
N-efficient cultivars produced a higher percentage of roots in the root branching interval of
60–90o.

100
D
Root length (%) of root fractions

80

F-
60
Adv. roots
Seminal roots
40
Primary roots

20

0
TH NUE ASK SL
Cultivar
Average diameter (mm) of root fractions

1.0
B
A 0.8

0.6

N-efficient Adv.
0.4 N-efficient Seminal
N-efficient Primary
N-inefficient Adv.
0.2
N-inefficient Seminal
N-inefficient Primary
0.0
TH NUE ASK SL
Cultivar
600 F-test
Number of tips of root fractions

Cult (Adv): ***

A 500 Cult(Sem): ns
Cult (Prim): ns
400
N-efficient Adv.
N-efficient Seminal
300
N-efficient Primary
N-inefficient Adv.
200
N-inefficient Seminal
N-inefficient Primary
100

0
TH NUE ASK SL
Cultivar
FIG. 10. Root length (%) of root fractions (A), average diameter of root fractions (B) and number of
root tips of the different root fractions (C) of 4 tropical maize cultivars grown in the cigar roll
experiment (n = 8) for 9 DAP at low N supply (100 µM N).

243
A 100
F-test
Cult (0-4 cm): *
80
Cult (4-8 cm): ***

Root length (%)

Cult (8-12 cm): ns
60 Cult (12-16 cm): ns
Cult (16-20 cm): ns
40
Cult (> 20 cm): ns
0-4 cm
4-8 cm
8-12 cm
20 12-16 cm
16-20 cm
> 20 cm
0
TH NUE ASK SL
B Cultivar
100
F-test
Root tip (%) at different height from

Cult (0-4 cm): **

80 Cult (4-8 cm): *
the horizontal surface

Cult (8-12 cm): ns

60 Cult (12-16 cm): ns
Cult (16-20 cm): ns
Cult (> 20 cm): *
40
0-4 cm
4-8 cm
20 8-12 cm
12-16 cm
16-20 cm
> 20 cm
0
TH NUE ASK SL
C Cultivar
Root length (%) by different diameter class

100
F-test
Cult (0-1 mm): ns
80
Cult (1-2 mm): ns
Cult (> 2 mm): ns
60
0-1 mm
1-2 mm
40 > 2 mm

20

0
TH NUE ASK SL
D Cultivar

100
F-test
Number of roots (%) with specified

Cult (0-30 o): ns

80 Cult (30-60 o): ns
Cult (60-90 o): ns
angle interval

60
0-30o
30-60 o
40 60-90 o

20

0
TH NUE ASK SL
Cultivar

FIG. 11. Root length (%) (A), root tips (%) (B) at different depth intervals from the surface, root
length (%) by different diameter classes and number of roots (%) within a specified root branching
angle from the horizontal surface of 4 temperate maize cultivars grown in the growth pouch
experiment (n = 6) at 10 DAP at low N supply (100 µM N).

244
4. DISCUSSION

The present study was designed to assess the relationship between root parameters in
short-term experiments and the N-efficiency of tropical and temperate maize cultivars in field
experiments.

Due to the existence of large genotypic difference in carbon assimilation under low N,
it is quite possible that root growth may be enhanced, unchanged, or even suppressed [9].
Nitrogen deficiencies resulted in significant decreases both in shoot and root dry matter
weight of young plants [16]. On the other hand, after investigating a range of tropical
cultivars, no correlation was found between root dry matter of a nutrient solution experiment
and N uptake in the field [31]. It was concluded that shoot dry matter at early growth stages in
nutrient solution is not related to shoot biomass at anthesis and grain yield, and therefore,
could not be used as a selection parameter for N efficiency [31]. Other authors also found no
correlation between root dry matter of seedlings and root parameters or important agronomic
traits in the field [4, 32], and concluded that root dry matter is not a useful criterion for
indirect selection. In our experiments we have found inconsistent results for the two sources
of maize cultivars. High shoot dry weights were observed by N-efficient tropical cultivars
compared to inefficient cultivars (Fig. 1A), while efficient temperate cultivars produced lower
shoot dry weight than inefficient cultivars (Fig. 7A). The root dry weight results were also
inconsistent. No significant differences were observed between cultivars in root dry weight of
tropical and temperate cultivars in the cigar roll experiment, but higher root dry weights of
inefficient tropical and temperate cultivars were observed in the growth pouch experiment
(Fig. 7A). This implies that the shoot and root dry weight of cultivars at the seedling stage
might not be a reliable parameter for selection for N efficiency.

Genotypic variation in root length of maize seedlings was reported [13], and a highly
significant correlation was found between the size of the root system and grain yield at low N
supply in field experiments [5]. Maize hybrids respond to N deficiency by increasing total
root length (TRL) [9]. However, in our experiment no genotypic differences were observed in
this trait. Other authors reported no correlation between root length in nutrient solution and
total shoot nitrogen uptake [31, 32]. However, in our experiment we found positive and
significant correlations between root lengths of 16 tropical cultivars in the cigar roll
experiment and N uptake up to anthesis (Table 3). Nevertheless, no correlation was found
with total root length after anthesis, which is very important for N efficiency. Therefore, it
seems that a high root length at the early growth stage is not a suitable selection parameter for
N efficiency.

Maize hybrids respond to N deficiency by elongation of individual seminal roots

while reducing the number of seminal roots [9]. Since seminal roots grow more in a
downward direction than adventitious roots, they may particularly contribute to the
exploration of the sub-soil during later stage of development [14]. In our study, significant
differences were observed among tropical cultivars in the number of seminal roots both in the
cigar roll and growth pouch experiments; moreover the mean seminal root number of N-
efficient cultivars was higher than inefficient cultivars. Significant genotypic differences were
also observed among temperate cultivars in the cigar rolls, but the N-inefficient cultivars had
a higher number of seminal roots than the efficient cultivar ASK (Figs. 1 and 7D). Low-input
cultivars of maize developed more and longer seminal roots than the high-input cultivars [13].
A higher seminal root number which was found for some N-efficient tropical cultivars is
beneficial, because an efficient cultivar seems to be able to exploit the available N in deeper
soil layers. On the other hand, Heuberger [4] concluded that the number of seminal roots

245
might not be a good predictor for early root growth in the field, but a high length of seminal
roots seemed to be more promising to identify cultivars with high penetration capacity into
the subsoil. We also observed no correlation between the number of seminal roots in the cigar
rolls and N uptake and grain yield of tropical cultivars (Table 3). Taken into consideration the
inconsistency of the result and literature findings it is possible to conclude that this parameter
is not a suitable selection parameter for N efficiency at low N.

Heuberger [4] reported significant genotypic variation in the number of adventitious

roots of maize in one pot experiment, but also found inconsistent correlations between this
fraction of the root system and root parameters in the field. Genotypic variation in this
parameter was also observed for maize [33] and other cereals [34]. In our study the number of
adventitious roots showed a similar trend as the number of seminal roots i.e. N-efficient
tropical cultivars produced more adventitious roots while efficient temperate cultivars had
fewer adventitious roots (Figs. 1 and 7D). We also observed no correlation between the
numbers of adventitious roots of tropical cultivars in the cigar rolls and N uptake and grain
yield (Table 3). Heuberger) [4] stated that the number of adventitious roots of seedlings may
only be a weak indicator for later development, since the bulk of adventitious roots are
formed at later growth stages, and cultivars may have different patterns of development. It
was also suggested that the number of adventitious roots is less important for a deep rooting
system [35]. Taken together, the inconsistence of the result and literature findings, the number
of adventitious roots at early growth stages cannot be used as a selection criterion for N
efficiency.

It has been mentioned that at later growth stages (after flowering), the newly
developed roots are characterized by a reduced diameter associated with increased root length
in the deeper soil layers maintaining N uptake [36]. In our study we found significant negative
correlation between the average root diameter and N uptake up to anthesis in the cigar rolls
(Table 3), suggesting that thinner roots are more important for high N uptake at low N.

The fineness of roots is expressed as specific root length (length per unit dry weight),
and roots with high specific root length (fine roots) are found in plants grown under nutrient
deficient conditions [26]. In our experiment, significant differences were observed in this trait
among tropical cultivars both in cigar roll and growth pouch experiments. However, this
difference could be found only among a limited number of cultivars (Fig. 2C). No genotypic
differences were detected in this trait among temperate cultivars (Fig. 8C).) It has been
proposed that fine roots (high specific root length) allow the root system to explore the soil
volume effectively while minimizing the investment needed to construct and maintain the root
system [27]. However, our results do not support this hypothesis, and therefore this trait is not
suitable to be used as a selection parameter for N efficiency.

Low tissue density enables a fast resource acquisition as a plant can rapidly expand its
root system with low investment in dry matter; however, the low dry-matter produced by such
a fast growing plants is likely to have a short life span [37]. In grasses, thicker roots and a
high tissue density have been associated with increased longevity [29]. In our experiment, no
genotypic differences were observed among tropical and temperate cultivars in tissue density
(Figs. 2 and 8D). However, some N-efficient cultivars tended to have higher tissue density
than inefficient cultivars. High tissue density is a result mostly of a high amount of cell wall
material and lignin [38, 39]. A lower percentage of air space may also contribute to a high
density [40]. Nevertheless, in the absence of genotypic difference it would be difficult to
generalize that high N uptake efficiency is due to the presence of high tissue density.

246
Therefore, this trait in young seedlings also appears to have no relevance for screening of
cultivars for N efficiency.

Lynch [22] proposed that studies of the root distribution and root architecture are
important because soil resources are unevenly distributed, or subject to localized
accumulation or depletion, so that the spatial deployment of the root system will measure the
ability of a plant to exploit those resources. He stated that, since they are difficult to observe
and to quantify we know little about these parameters. In our study, we used growth pouches
to study the root distribution of cultivars in the pouch and correlate them with the field result.
Some N-inefficient tropical cultivars possessed a relatively higher percentage of root length in
the 0-4 and 4-8 cm layers; on the other hand, N-efficient tropical cultivars possessed a
relatively higher percentage of root length in the deepest >20 cm layer (Fig. 5A). Though
non-significant, a tendency for a positive correlation was also observed between the
percentage of root length in the lower layer of the pouch and N uptake after anthesis and grain
yield (Table 4). This result is in agreement with Worku [30] who reported that an N-efficient
cultivar developed relatively more roots in the subsoil (60-90 cm) than an N-inefficient
cultivar. Also, the N-efficient cultivar developed more vertically oriented fine roots and
exploited more mineral-N in the soil layer below 60 cm. Similar trends were observed in the
percentage of root tips in the upper 2 and the lowest grow-pouch layers for the tropical
cultivars. However, the result obtained for the temperate cultivars did not fit this trend: the N-
inefficient cultivar possessed a higher percentage of root length both in the first and the
deepest layer (Fig. 10B). Worku [30] stated that the root-length density determined for
tropical maize cultivars was much lower than has been previously reported for temperate
cultivars, implying that differences in root-length densities and distribution in the subsoil
between cultivars may be of greater importance in tropical than in temperate environments.

The angle of root growth has been reported as an important factor in determining the
distribution of the root system in the soil [41]. In our study, N-inefficient tropical and
temperate cultivars possessed high percentages of roots at a branching angle interval of 0-30
and 30-60o, while N-efficient cultivars produced high percentage of roots at a branching angle
interval of 60-90o (Figs. 6 and 11D). This suggests that N-efficient cultivars produce more
vertically grown roots than inefficient cultivars. Moreover, the percentage of roots in a
branching angle interval of 60-90o was positively correlated (P<0.1) with N uptake after
anthesis and grain yield (Table 4). Dunbabin and colleagues [7, 25] concluded on the basis of
their simulation model that plants with a deep-rooted architectures acquire more nitrate than
shallow-rooted plants. The high N uptake efficiency of N-efficient cultivars might be related
to this deep rooted architecture.

It is concluded form our work presented here, that among the many shoot and
particularly root parameters assessed in the growth pouch and cigar roll experiments with
seedlings, the only parameters which might be suitable to be used as a selection parameter for
N efficiency of tropical maize cultivars might be the percentage of root-length distribution in
a horizontal layer and the percentage of roots with a specified branching-angle interval
measured in the growth pouch experiment.

ACKNOWLEDGMENTS

The authors would like to thank IAEA, Vienna, for financial support.

247
 REFERENCES

[1] RAUN, W.R., JOHNSON, G.V., Improving nitrogen use efficiency for cereal
production, Agron. J. 91 (1999) 357–363.
[2] FAGERIA, N.K., BALIGAR, V.C., Enhancing nitrogen use efficiency in crop plants,
Adv. Agron. 88 (2005) 97–185.
[3] KLING, J.G., et al., “Potential for developing N use-efficient maize for low input
agricultural systems in the moist savannas of Africa”, Developing Drought and Low-N
Tolerant Maize, (EDMEADES, G.O., et al., Eds), Proc. Symp. Elbatan, Mexico
(1996) 490–501.
[4] HEUBERGER, H.T., Nitrogen Efficiency in Tropical Maize - Indirect Selection
Criteria with Special Emphasis on Morphological Characteristics, Ph. D Thesis,
University of Hannover, Germany (1998).
[5] HORST, W.J., et al., “Genotypic differences in nitrogen use-efficiency in crop plants”,
Innovative Soil-Plant Systems for Sustainable Agricultural Practices, (LYNCH, J.M.,
SCHEPERS, J.S., ÜNVER, I., Eds), OECD (2002) 75–92.
[6] BRADSHAW, A.D., Evolutionary significance of phenotypic plasticity in plants, Adv.
Genet. 13 (1965) 115–155.
[7] DUNBABIN, V., RENGEL, Z., DIGGLE, A.J., Simulating form and function of root
systems: Efficiency of nitrate uptake is dependent on root system architecture and the
spatial and temporal variability of nitrate supply, Funct. Ecol. 18 (2004) 204–211.
[8] MARGARITA, R., VILLAGARCIA, V., COLLINS, W., Nitrogen uptake and
nitrogen use efficiency of two sweet potato genotypes during early stages of storage
formation, J. Am. Soc. Hort. Sci. 5 (1998) 814–820.
[9] CHUN, L., et al., Genetic analysis of maize root characteristics in response to low
nitrogen stress, Plant Soil 276 (2005) 369–382.
[10] WANG, Y., et al., Response of root morphology to nitrate supply and its contribution
to nitrogen accumulation in maize, J. Plant Nutr. 27 (2004) 2189–2202.
[11] ROBINSON, D. Resource capture by localized root proliferation. Why do plants
bother? Ann. Bot. 77 (1996) 179–185.
[12] DREW, M.C., SAKER, L.R., Nutrient supply and the growth of the seminal root
system in barley: II. Localized, compensatory increases in lateral root growth and rates
op nitrate uptake when nitrate supply is restricted to only part of the root system, J.
Exp. Bot. 26 (1975) 79–90.
[13] FEIL, B., et al., Root traits of maize seedlings-indicators of nitrogen efficiency? Plant
Soil 123 (1990) 155–159.
[14] GORNY, A.G., LARSON, S., New aspects in root breeding, Vortr. Pflanzenzüchtung
16 (1989) 339–356.
[15] GALLAIS, A., COQUE, M., Genotypic variation and selection for nitrogen use
efficiency in maize. A Synthesis, Maydica 50 (2005) 531–537.
[16] VARGA, P., SARDI, K., BERES, I., Effects of N imbalances on shoot and root-
growth of corn and velvet leaf, Proc. 7th Hungar. Congr. Plant Physiol. 46 (2002) 213–
214.
[17] MARSCHNER, H., Role of root-growth, arbuscular mycorrhiza and root exudates for
the efficiency in nutrient acquisition, Field Crops Res. 56 (1995) 203–207.
[18] ERICSON, T., Growth and shoot: root ratio of seedlings in relation to nutrient
availability, Plant Soil 169 (1995) 205–214.
[19] MAIZLICH, N.A., FRITTON, D.D., KENDALL, W.A., Root morphology and early
development of maize at varying levels of nitrogen, Agron. J. 72 (1980) 25–31.

248
[20] SATTELMACHER, B., HORST, W.J., BECKER, H.C., Factors that contribute to
genetic variation for nutrient efficiency of crop plants, Z. Pflanzenernähr. Bodenk. 157
(1994) 215–224.
[21] VAMERALI, T., et al., A comparison of root characteristics in relation to nutrient and
water stress in two maize hybrids, Plant Soil 255 (2003) 157–167.
[22] LYNCH, J.P., Root architecture and plant productivity, Plant Physiol. 109 (1995) 7–
13.
[23] ZHANG, H., et al., Dual pathways for regulation of root-branching by nitrate, Proc.
Nat. Acad. Sci. USA 96 (1999) 6529–6534.
[24] SCHIEFELBEIN, J.W., BENFEY, P.N., The development of plant roots: New
approaches to underground problems, Plant Cell 3 (1991) 1147–1154.
[25] DUNBABIN, V., DIGGLE, A., RENGEL, Z., Is there an optimal root architecture for
nitrate capture in leaching environments? Plant Cell Environ. 26 (2003) 835–844.
[26] FITTER, A.H., “Functional significance of root morphology and root system
architecture”, Ecological Interactions in Soil, (FITTER, A.H., et al., Eds), Blackwell
Scientific, Oxford (1985) 87–106.
[27] FITTER, A.H., An architectural approach to the comparative ecology of plant-root
systems, New Phytol. 106 (1987) 61–77.
[28] HUTCHINGS, M.J., de KROON, H., Foraging in plants: The role of morphological
plasticity in resource acquisition, Adv. Ecol. Res. 25 (1994) 159–238.
[29] EISSENSTAT, D.M., YANAI, R.D., “The ecology of root lifespan”, Adv. Ecol. Res.
27 (1997) 1–60.
[30] WORKU, M., Genetic and Crop-Physiological Basis of Nitrogen Efficiency in
Tropical Maize: Field studies, Ph. D Thesis, University of Hannover, Germany (2005).
[31] AMBEBE, T.E., Genotypic Differences in Nitrogen Efficiency in Maize (Zea mays L.)
in relation to leaf senescence, M. Sc Thesis, University of Hannover, Germany (2006).
[32] BEGUM, N., Genotypic Differences in Nitrogen Efficiency in Maize (Zea mays L.) in
Relation to Leaf Senescence, M. Sc Thesis, University of Hannover, Germany (2004).
[33] TEYKER, R.H., HOBBS, D.C., Growth and root morphology of corn as influenced by
nitrogen form, Agron. J. 84 (1992) 694–700.
[34] EL BASSAM, N., A concept of selection for 'low input' wheat varieties, Euphytica
100 (1998) 95–100.
[35] OYANAGI, A., NAKAMOTO, T., MORITA, S., The gravitropic response of roots
and the shaping of the root system in cereal plants, Environ. Exp. Bot. 33 (1993) 141–
158.
[36] OIKEH, S.O., et al., Growth and distribution of maize roots under N fertilization in
plinthite soil, Field Crops Res. 62 (1999) 1–13.
[37] RYSER, P., LAMBERS, H., Root and leaf attributes accounting for the performance
of fast and slow-growing grasses at different nutrient supply, Plant Soil 170 (1995)
251–265.
[38] DIJKSTRA, P., LAMBERS, H., Analysis of specific leaf-area and photosynthesis of
two inbred lines of Plantago major differing in relative growth rate, New Phytol. 113
(1989) 283–90.
[39] GARNIER, E., LAURENT, G. Leaf anatomy, specific mass and water content in
congeneric annual and perennial grass species, New Phytol. 128 (1994) 725–736.
[40] KOIKE, T., Leaf structure and photosynthetic performance as related to the forest
succession of deciduous broad-leaved trees, Plant Species Biol. 3 (1988) 77–87.
[41] FORDE, B.L., LORENZO, H., The nutritional control of root development, Plant Soil
232 (2001) 51–68.

249
 EVALUATION AND SELECTION OF MAIZE (ZEA MAYS L.) GENOTYPES
TOLERANT TO LOW N SOIL

C. THE
West Africa Centre for Crop Improvement (WACCI),
Legon, Accra,
Ghana

M.L. NGONKEU, C. ZONKENG, H.M. APALA

Institute of Agricultural Research for Development (IRAD),
Yaoundé,
Cameroon

Abstract

The identification and/or the development of germplasm with traits which enhance N uptake
and N use efficiency in low N soil could significantly sustain maize production on stress
environments. The use of secondary traits highly correlated with grain yield and high heritability,
could speed up the development of genotypes adapted to low N environments. Arbuscular mycorrhiza
fungi are known to enhance P uptake, but its role on plant N nutrition has not been extensively studied.
The study aimed to (i) identify tolerant and/or low N responsive genotypes (ii) measure the correlated
response of grain yield with some agronomic plant characteristic under low N and under mycorrhiza
inoculation (iii) measure the combining ability and the gene effects of the lines under low and high N
and (iv) to identify stable and high yielding hybrids adapted to low and high N condition. Initial
screening of 99 genotypes for two years identified 30 inbred lines that were evaluated in split plots for:
grain yield, root volume, chlorophyll content, leaf area index, and mycorrhizal colonization.
Significant genotype x soil N level interactions were obtained among the tested inbreds for all
measured traits, except for chlorophyll content which exhibited similar ranking from one soil N level
to another. In addition to selection for grain yield, 5 lines were retained for their good root volume, 4
for their chlorophyll content and stay green traits, 3 for their leaf area index and the last 3 for their
mycorrhizal colonization. Diallel crosses among the 15 selected lines yielded 105 F1 hybrids
evaluated in split plots, with 3 soil treatment levels (20 kg−N ha-1, 20 kg−N ha-1 + mycorrhiza and 100
kg−N ha-1). Significant differences were detected among the 3 soil treatments as well as for genotypes
x soil interaction for all measured traits. On 20 N plots, 10 hybrids yielded at least as good as the
check hybrid: Expl24 x 87036 (3.0 t ha-1). Among the 20 parents involved in these crosses, 8 had been
retained for their chlorophyll content, six for their root volume, four for their leaf area index and two
for their mycorrhizal colonization. On plots with 20 N, in addition to mycorrhizal inoculation, 8
hybrids yielded more than the check (3.2 t ha-1). Among the 16 inbreds involved, 5 were retained for
their root volume, 4 for their mycorrhizal colonization and leaf area index, respectively, and 3 for their
chlorophyll content. On 100 N plots, 13 hybrids yielded more than the check (5.4 t ha-1). Among the
26 inbreds involved, 13 (50%) were selected for their root volume, six (25%) for their mycorrhizal
colonization, four (15.4%) for their leaf area index, and three (11.5%) for high chlorophyll content. It
was concluded that on low N soil, selection for root volume was correlated with grain yield. However,
maximum benefit was obtained in hybrids when their inbred parents exhibited complementary plant
characteristics such as the efficient colonization by mycorrhiza, good leaf area index and high
chlorophyll content.

1. INTRODUCTION

Low nitrogen (N) stress in the soil is one of the most important abiotic stresses
causing yield reduction in maize production in tropical soils [1]. In these soils, 90% of maize
is grown by resource poor farmers who cannot afford expensive N fertilizers. Since 2007, the
price of N fertilizer has increased in developing countries mostly, making it inaccessible to

251
poor farmers. To alleviate this constraint, the identification and/ or the development of
germplasm with traits which enhance N uptake and utilization in poor soils could significantly
contribute to maize production on these stress environments. Several studies have indicated
that useful genetic potential for the improvement of N use efficiency exists in maize [2].
Therefore, the application of efficient farming techniques and/or the use of plant varieties that
have a better N use efficiency could reduce the use of N fertilizer [3]. Breeding strategies to
develop stress tolerant maize inbred lines include screening and selection of inbred lines
under managed stress conditions, multi-location testing of hybrids in a representation sample
of the target environment, and selection under high plant populations [4]. However, the per se
performance of maize inbred lines doesn’t predict the performance of hybrids for grain yield
or any other traits [5]. Given the low heritability of yield under stress conditions, it has been
suggested that the use of the secondary traits that possess high heritability and are correlated
with tolerance to low-N could help to speed up the development of high yielding, stable and
adapted varieties [6].

The ability of plants to access soil N varies among species and genotypes within
species and is affected by interactions of plants and microbes [7]. In addition to the inherent
genetic potential of a plant to extract N from soil, the ability to interact with microbes can also
be a major determinant of a plant’s ability to grow in soils with low levels of bioavailable N.
Arbuscular mycorrhizal (AM) fungi are plant symbionts that are known to enhance P uptake
in numerous plant species. The role of AM fungi on plant N nutrition has not been studied as
extensively as that of P, but it is now receiving more attention. It has been shown that AM
fungi directly affect N absorption and N assimilation [8] particularly in neutral soils [9]. Cruz
et al. [10] reported that AM fungi may improve plant N nutrition when grown in nutrient-poor
soil, and Mäder et al. [11] suggested that AM fungal hyphae may contribute substantially to
plant N nutrition.

The knowledge of the correlated response of plant characteristics (especially root

systems and leaf chlorophyll concentration) and grain yield on low-N soils, as well as the
combining ability and gene action of selected genotypes could help to speed up the
development of high yielding, stable and adapted cultivars. The potential use of nuclear
technique and the studies on rhizosphere characteristics, (root exudates, genotype x
mycorrhiza interaction, genotype x cultural system) will enhance the attainment of the
expected results. The advantages of N use efficiency would be the reduction of the quantity of
N fertilizer below the recommended rate of 100 kg−N ha-1 for the humid forest zone of
Cameroon. This would significantly contribute to sustainable maize production both
economically and environmentally. The overall objective of this study was to increase food
security in developing countries by sustaining maize production using low-N tolerant
germplasm and/or mycorrhizal (AM) fungi. The specific objectives were to (i) identify,
through intensive evaluation, maize germplasm tolerant and/or responsive to low N soils of
the tropics (ii) measure the correlated responses of maize grain yield with some plant
agronomic characteristics under low N and under mycorrhizal inoculation (iii) measure the
combining ability and the gene effects of selected lines exhibiting various levels of production
under low N.

2. MATERIALS AND METHODS

2.1. Site description

The experiment was conducted at Nkolbisson located at 11°36´ E and 3°44´ N in the
humid forest zone, 5 km from the capital city, Yaounde. The altitude is 650 masl. The annual
rainfall is 1560 mm, with a bimodal distribution. The average temperature is 23.5°C, the soil

252
is a sandy clay with pH (water) of 4.5, CEC (cmol (+) kg-1 of 4.79) and Al (cmol (+) kg-1 of
0.30). The main farming system is maize/groundnut/cassava as monocultures or intercropped.

Site preparation started in April 2007 at Nkolbisson with soil N depletion (Fig. 1).
This consisted of planting 90 000 plants of maize ha-1 in 2006 and early 2007. Soil liming was
carried out in 2008 by applying “dolomite” (35% CaO, 20% MgO) at the rate of 2000 kg ha -1
to remove acidity effect. Soil samples were collected before liming, and were analysed at the
Soil, Plant and Fertilizer laboratory of IRAD at Nkolbisson (Table 1).

TABLE 1. SOME OF THE SOIL CHARACTERISTICS AT NKOLBISSON

Soil property Value

Organic matter (%) 4.06
Organic carbon (%) 2.36
Total nitrogen (%) 0.10
C: N 23.6
CEC (cmol kg-1) 7.27
pH 4.8

FIG. 1. Soil depletion at Nkolbisson.

2.2. Mycorrhizal substrate and inoculum preparation

Sand from the Wouri sea (Littoral zone of Cameroon) was collected, sieved (<2 mm)
and washed with distilled water, resulting in a pH of 6.8, and an N content of 0.6 g kg-1, with
0.2 mg NO3 - kg-1. This substrate was autoclaved at 121°C for 20 min on three consecutive
occasions, and then inoculated with soil extract (20 ml pot -1 of sand: water mixture, 1: 1 v/v)
filtered through a Whatman No. 1 paper, to reintroduce the native microbial community, with
the exception of propagules of AM fungi. Arbuscular mycorrhizal fungi was provided by the
Regional Laboratory of Biological Control and Applied Microbiology of IRAD, Cameroon.
The mixed culture was composed of three AM fungi, Acaulospora tuberculata, Gigaspora
margarita and Glomus intraradices. All of these AM have demonstrated enhancement of
plant growth, phosphorus and P concentration in plants and high % colonization of maize and
sorghum roots in acid soils [12]. Therefore, the inoculum used consisted of soil, mycorrhizal
root fragments and spores and mycelia of the mixed species. Inoculation was carried out by
adding 20 g of inoculum in one plastic pot of 2 liters [12].

253
2.3. Plant material

Plant materials used in the 2009 experiment included 30 inbred lines selected from the
original 99 lines evaluated in 2007 and 2008. Their origin and characteristics are shown in
Table 2. Eleven (11) of these line originated from IRAD, Cameroon, twelve (12) were
obtained from IITA, five were obtained from CIMMYT, Mexico, and 2 were from CIMMYT,
Colombia. The 30 inbred lines were evaluated in 2009 at Nkolbisson on 20 N plots as well as
on 100 N plots. Three replications were used, one of which was for destructive sampling for
data collection on root characteristics.

TABLE 2. THE ORIGIN, AND PLANT CHARACTERISTICS OF INBRED LINES USED IN THE
2008 AND 2009 SCREENING EXPERIMENTS

Code Inbred lines Origin† Color Characteristics and heterotic pattern

L1 TZSTR 133 IITA White Striga Tolerant
L2 Ku 1409 IITA Yellow low N soil Tolerant
L3 M 131 IRAD White low N soil tolerant
L4 Exp1 24 IRAD White High yielding
L5 Cla 17 CIMMYT Yellow Tolerant to Al toxic soil
L6 88069 IRAD Yellow Good root volume
L7 TZSTR 131 IITA Yellow Striga tolerant
L8 TZSTR 137 IRAD White Good root lenght
L9 88094 IRAD White Mycorrhiza use
L10 1368 IRAD Yellow Good leaf area
L11 87036 IRAD Yellow Root volume
L12 CML 358 CIMMYT Yellow High chlorophyll
L13 CML 254 CIMMYT Yellow Good mycorrhize colonization
L14 TZSTR 140 IITA Yellow Good root volume
L15 ATP S425W IRAD White Tolerant to acid soil
L16 0114077 IITA White mycoorhize efficient
L17 0114079 IITA White Mycoorhize efficient
L18 0114075 IITA White Mycoorhize efficient
L19 ATP S7 33Y-2 IRAD Yellow Root volume
L20 ATP S5 31Y-1 IRAD Yellow Tolerant to acid soil
L21 ATP S6 22Y-2 IRAD Yellow Tolerant to acid soil
L22 ATP S8 26Y-3 IRAD Yellow Tolerant to acid soil
L23 ATP S9 30Y-3 IRAD Yellow Tolerant to acid soil
L24 ATP S5 26Y-1 IRAD Yellow Tolerant to acid soil
L25 CML 365 CIMMYT White Tolerant to acid soil
L26 9450 IITA Yellow Temperate adapted
L27 9071 IITA White Temperate adapted
L28 5012 IITA White Temperate adapted
L29 Entrada 3 CIMMYT White Tolerant to acid soil
L30 Entrada 29 CIMMYT White Tolerant to acid soil
†CIMMYT, Colombia

In 2010, 15 of these lines were selected based on their plants characteristics when
evaluated under low N (Table 3). Five (5) of these line were selected for their root
characteristics, four (4) were retained for their stay green character at 20 N as measured by
their chlorophyll content, three (3) lines each were retained for their mycorrhizal colonization
and for their leaf area index. These lines were crossed in a diallel to obtain 105 F1 hybrids.

254
TABLE 3. INBRED LINES, ORIGIN, PEDIGREE, HETEROTIC PATTERN, ADAPTATION AND
CHARACTERISTICS

Inbred line Origin† Pedigree Heterotic Adaptation Characteristics‡

pattern
87036 IRAD TZMSR x pop32 Eto Mid-altitude High root volume
M 131 IRAD TZMSR x POP32 Eto Mid-altitude High myc. colon.
ATP-S9-30-Y-1 IRAD Acid tolerant pop - Lowland High leaf area index
9450 IITA B73 Temperate Lowland High root volume
CML 358 CIMMYT Pop SA3 Eto Lowland High chlorophyll
88069 IRAD - - Mid-altitude High root volume
88094 IRAD TZMSR x pop43 Tuxpeño Mid-altitude Myc. colon.
1368 IITA Pop21 Tuxpeño Lowland Leaf area
ATP-S6-20-Y-1 IRAD Acid tolerant pop - Lowland Chlorophyll content
9848 IITA - Temperate Lowland Leaf area
CAM Inb gr1 17 IRAD SuwanI Tuxpeño Lowland Root volume
ATP-S5-31-Y-1 IRAD Acid tolerant pop - Lowland Chlorophyll content
CML 254 CIMMYT - - Lowland Myc. colon
9071 IITA N 28 Temperate Lowland Root volume
5012 IITA - Temperate Lowland Chlorophyll content
†CIMMYT: Centro International de Mejoramiento de Maíz y Trigo; IRAD: Institut de Recherche Agricole pour
le Développement; IITA: International Institute of Tropical Agriculture
‡Myc. colon., mycorrhizal colonization

2.4. Data collected

For all experiments, the following data were collected:

− Number of days from planting to 50% silking (DTS)

− Number of days from planting to 50% anthesis (DTA)
− Anthesis Silking Interval (ASI)
− Leaf width(LWIDTH)
− Leaf length(LLENGTH)
− Stem diameter(STEMDIA)
− Leaf area in cm2 (LAREA)
− Chlorophyll concentration (with a SPAD 502chlorophyll meter) (CHLOROCONIC)
− Root weight (ROOTWT)
− Plant height (PHT)
− Ear height (EHT)
− Plant aspect (PA)
− Plant stand at harvest (PAH)
− Number of ears at harvest (EAH)
− Mycorrhizal colonization
− Grain yield (YIELD)

2.5. Root characteristics

Root weight was measured on a destructive replication. Data were recorded on six
consecutive plants per plot at anthesis.

2.6. Experimental design and data analysis

The preliminary field evaluation of the 30 inbreds retained was carried out in 2009.
These lines were evaluated at Nkolbisson using a RCBD with 2 replications. For the

255
mycorrhiza experiment in pots, the screening consisted of a factorial arrangement; the factors
were: 2 mycorrhizal treatments: M0 = plot without mycorrhiza, M1= plot with mycorrhiza,
two N levels (N1 = 20 N and N2 = 100 N), and the factors consisted of the 30 maize inbred
lines evaluated in three replications. Fourteen days after sowing, a single application of N in
distilled water was made. Plants were watered every three days for 2 months. Rorison’s
nutrient solution without N was applied to supply deficient nutrients when observed.
Chlorophyll rate, plant height (cm) and leaf number were collected one month after planting.
Additional parameters collected included: root colonization [13], root length (cm), root dry
weight (g), plant fresh weight (g). Mycorrhizal colonization data were arc sine transformed to
normalize their distribution. Data were analyzed using the statistical SAS package v. 9.1, and
significant differences between means were made using the Student-Newman-Keuls test.

In 2010, the 15 inbred lines retained as described above were crossed in diallel to
obtain 105 F1. During the second rainy season, these hybrids were evaluated in a field
experiment. The experimental design consisted of a split plot with the main plot consisting of
3 soil treatments: 20 N, 20 N + mycorrhiza, 100 N. The sub-plots consisted of the tested 105
genotypes arranged in an alpha lattice. For all experiments, three seeds were planted per hill
and later on was thinned to two plants per hill. Row and hill spacing were 0.75 and 0.50 m,
respectively. The total plant density was 53 333 plants ha-1 and the experimental unit
consisted of a single row per plot, 5 m long.

ANOVA was performed for each soil correction using alpha-lattice. The percentage
grain yield reduction under different N levels was computed for all genotypes. Genotypes
with small difference were considered resistant to low N soil. Efficient genotypes on low N
soil were those which exhibited high grain yield on both low N and high N. The top 20% for
grain yield at each soil N level were identified, and high yielding common genotypes for all
soil treatments were determined. In addition, parents with high appearance frequency were
determined. This gave an indication of parents most adapted to low N as well as high N soil.
The GCA and the SCA of the parental line were estimated using the Griffing method. Finally,
the correlations between grain yield and measured plant characteristics at each soil correction
level were calculated. This suggested plant characteristics to be used when breeding and
selecting for low N tolerant and efficient genotypes

3. RESULTS AND DISCUSSION

3.1. Preliminary evaluation and selection of inbred lines

The combined analysis for the 30 lines indicated significant differences between the
two N levels for all measured characteristics (Table 4). All genotype performed differently for
all the measured traits at the two N levels. Significant genotypes x soil N levels interactions
were obtained for all the measured traits except chlorophyll concentration, indicating that the
tested genotypes had different relative performance from one soil nitrogen level to another.

3.2. Grain yield

Grain yield of the 30 inbreds used in the preliminary evaluation are shown on Table 5.
At 20 kg−N ha-1, 7 lines produced more than 2.0 t ha-1. These were 88069 (3.41 t ha-1), 114075
(3.02 t ha-1), 87036 (2.84 t ha-1), ATP S8 30 Y−3 (2.54 t ha-1), M 131 (2.48 t ha-1), CML 365
(2.43 t ha-1), and ATP S9 30 Y−1 (2.11 t ha-1). Five lines yielded between 1.5 and 2.0 t ha-1:
ATP S5 26 Y−1, Ku 1409, 9071, 114079, Cla 17. At 100 kg−N ha-1, 16 lines yielded at least
2.2 t ha-1. The best ones included: 88069 (4.98 t ha-1), 87036 (4.92 t ha-1), 114075 (4.04 t ha-

256
1
), 114077 (3.74 t ha-1), CML 365 (3.54 t ha-1), M 131 (3.26 t ha-1), ATP S8 30 Y−3 (2.97 t ha-
1
) and Cla 17 (2.8 t ha-1).

TABLE 4. MEAN SQUARES FROM THE COMBINED ANALYSIS OF THE 30 INBRED

PARENTAL CROSSES AT THE TWO N LEVEL FOR ALL MEASURED TRAITS AT 20 KG−N
HA-1 DURING THE 2009 AND 2010 GROWING SEASONS

Traits Mean squares†

N level (N) Varieties (V) NxV
DTA 42.1** 5.9** 3.8**
DTS 115.2** 7.05** 6.2**
ASI 11.25** 2.14** 1.46*
LAREA 480288.5** 40898** 4663.1*
STEMDIA 4.89** 0.27** 0.06*
CHLOROCONC 1445.9** 52.1** 0.13ns
ROOTWT 135766** 21363.3** 9750.2**
PHT 14.4** 2793.4** 1120.7**
EHT 523.6** 823.3** 486.8**
YIELD 9.7** 7.6** 2.1**
†Degrees of freedom for: N level = 1; Varieties = 29; N x V = 29;
* and ** denote significance at P<0.05 and P<0.01, respectively

The percent grain yield reduction due to fertilisation varied from 3.70 (TZSTR 137) to
87.5% (TZSTR 133). Thirteen lines showed grain yield reduction due to N fertilization of less
than 25%. Among these lines, 88069, ATP S9 30 Y−1 (6.2%), ATP S8 30 Y−3 (14.5%), Ku
1409 (15.8%) ATP S5 26 Y−1 (20.6%), 9071 (21.6%) and M 131 (23.9%) exhibited more
than 1.5 t ha-1 at 20 kg−N ha-1and were among the 10 best yielding at 100 kg−N ha-1.

3.3. Leaf area

Leaf area index at 20 N ranged from 160.3 (TZ−STR 137) to 508.4 (CML 365) (Table
5). The average of the trial was 319.9. Eight genotypes exhibited leaf area index of more than
352.4. They were: CML 365, 1368, ATP S9 30 y−1, 88069, 114075, 87036, M131, ATP
S6−22−Y−2. At 100 N, leaf area index varied from 311.0 (TZ−STR 133) to 682.5 (88069), with
a mean value of 423.2. Thirteen genotypes exhibited leaf area index of more than 425. These
were: 88069, ATP S9 30 y−1, CML 365, 87036, 114075, 00114077, M131, ATP−S6−20−y−1,
ATP−S6−22−y−2, Entrada 29, CML 358, 1368. Genotypes retained for best leaf area index at
both 20 N and 100 N included: 88069, 87036, 9848, 1368, 114075, ATP−S6−22−y−2, M131,
CML 365, ATP−S9−30−y−1.

3.4. Chlorophyll content

The chlorophyll contents of the 30 genotypes at 20 N and 100 N are shown in Table
5. In general, mean chlorophyll concentration was higher at 100 N than 20 N. At 20 N,
chlorophyll content varied from 26.9 for 88094 to 41.0 for CML 358, with a mean of 34.8.
Nine inbreds exhibited chlorophyll content of at least 36.0. These were: 114 075, ATP S5 26
y-1, KU1409, ATP-S6-20 y-1, ATP-S6-22-y-2, ATP-S5-31-y-2, Entrada 29, CML 358, 5012.
At 100N, chlorophyll content varied from 35.9 (Entrada 3) to 49.2 (CML 358), with a mean
of 40.5. Twelve inbreds had chlorophyll content of at least 42.0. Among these, 6 had high
chlorophyll content at 20 N and those selected were: CML 358, ATP-S5-26-y-1, 5012, Ku
1409, ATP-S6-20-y-1, ATP-S6-22 y-2.

257
TABLE 5. GRAIN YIELD AND PLANT CHARACTERISTICS OF THE ORIGINAL 30 MAIZE
INBRED LINES AT 20 AND 100 KG−N HA-1

Varieties Yield % yield Leaf area index Chlorophyll content Root weight
20 N 100 N loss 20 N 100 N 20 N 100 N 20 N 100 N
88069 3.4 5.0 19.7 471.1 682.5 35.2 39.7 161.2 206.3
114075 3.0 4.0 25.3 456.5 502.3 36.3 39.2 79.3 246.5
87036 2.8 4.9 42.3 397.3 556.8 35.7 42.2 74.0 291.1
ATP S8 30 Y-3 2.5 3.0 14.5 292.0 411.5 32.7 40.6 14.7 16.7
M 131 2.5 3.3 23.9 391.0 455.5 35.3 42.9 39.5 92.0
CML 365 2.4 3.5 31.4 508.4 567.0 32.5 42.0 59.3 177.6
ATP S9 30 Y-1 2.1 2.3 6.2 504.1 621.1 35.2 41.6 28.3 34.4
ATP S5 26 Y-1 1.9 2.4 20.6 323.5 410.0 37.7 45.2 38.3 31.7
Ku 1409 1.9 2.3 15.8 235.6 325.5 37.1 45.5 23.6 87.6
9071 1.7 2.2 21.6 345.9 387.1 35.9 38.1 55.2 92.0
114079 1.7 2.4 28.4 327.4 385.7 33.6 40.5 40.4 246.5
Cla 17 1.6 2.8 44.9 256.9 426.8 32.5 40.7 28.8 61.2
TZSTR 140 1.5 1.8 17.0 310.0 352.7 34.9 39.7 13.7 15.0
CML 254 1.5 1.6 7.1 311.4 415.3 31.9 42.6 38.3 43.3
114077 1.4 3.7 62.3 354.9 482.7 30.3 42.9 138.0 140.0
ATP S6 20 Y-1 1.3 1.4 5.7 313.6 425.5 38.3 44.0 39.5 43.0
Entrada 3 1.3 1.8 28.6 294.4 367.9 31.2 35.9 21.0 32.3
ATP S6 22 Y-2 1.2 2.1 42.7 352.4 445.2 37.2 47.3 43.0 138.4
9450 1.1 1.4 23.2 268.8 341.3 34.4 34.7 38.3 41.0
1368 1.0 2.0 46.7 278.7 406.4 34.8 37.8 18.0 33.0
ATP S5 31 Y-2 0.8 1.8 53.9 251.4 323.7 38.0 40.2 13.7 47.7
Entrada 29 0.7 2.0 65.2 220.6 488.4 36.7 37.4 8.7 39.0
4001STR 0.7 2.5 71.4 248.0 397.2 35.0 43.0 18.0 20.0
CML 358 0.6 0.8 23.2 307.2 427.1 41.0 49.2 20.0 92.0
Exp1 24 0.6 2.2 72.0 275.7 397.0 32.1 38.9 12.0 77.0
5012 0.5 1.7 69.0 226.4 323.0 37.7 42.1 39.5 48.0
TZSTR 131 0.5 2.1 73.9 345.9 387.1 35.9 37.0 54.5 39.4
TZSTR 137 0.5 0.6 3.7 160.3 337.3 33.5 34.7 6.5 20.0
88094 0.5 1.1 53.3 262.8 336.2 26.9 36.5 30.0 30.0
TZSTR 133 0.3 2.2 87.5 305.6 311.8 31.9 32.1 12.3 16.7
MEAN 1.7 2.2 319.9 423.2 34.8 40.5 40.0 94.9
CV (%) 55.2 32.4 15.9 12.4 3.2 7.8 0.5 0.0
Lsd 1.6 1.2 83.2 85.9 5.2 5.2 0.3 0.0
S.E 1.2 1.7 93.0 99.4 4.0 4.8 34.8 94.5

3.5. Root weight

Root weight of the 30 tested inbred lines taken at 20 N and 100 N are presented in
Table 5. At 20 N, genotype root weight varied from 6.5 g (TZ-STR-137) to 161.2 g (88069),
with a mean of 40.0 g. Nine inbreds exhibited root weights of at least 40.4 g. These were:
88069, 114077, 114075, 87036, CML 365, 9071, TZ-STR 131,114079, and Cam inb gp1.17.
At 100 N, genotype root weights ranged from 15.0 to 291.1 g, with an average of 94.9 g.
Eight genotypes exhibited root weights of at least 130.0 g. These were; 88069, 114075,
87036, CML 365, 114079, 114077, Cam inb gp1.17 and ATP-S6-22-y-2. It was noted that 7
lines ranked among the best at 20 N as well as for 100 N. These were: 88069, 114075, 87036,
CML 365, Cam inb gp1.17, 114079 and 114077.

258
3.6. Shoot weight and root weight response to mycorrhizal inoculation

In neutral sand with low nitrogen (20 kg−N ha-1) and without the addition of AM
fungi, maize inbred lines differed for shoot and root length (Fig. 2). The best lines were:
0114077, 88069, AT S7 33Y-2, ATP S4 25W, 87036 and TZSTR 131. The average shoot and
root length ratio was around 4.7 g.

A B
FIG. 2. Growth variation and root density of some maize inbred lines on sand (A) 20 N fertilizer (B)
20 N fertilizer with addition of mycorrhizae.

With the addition of AM fungi, SFW and root length increased significantly with 20
kg−N ha-1. Varieties CML 365, ATP S5 26 Y-1, CML 254, M 131, 88094, TZSTR133
exhibited high general performance (Fig. 2B). The addition of AM fungi to 100 kg−N ha-1,
improved SFW and RL of maize inbred lines except for 88069, ATP S6 22 Y-2 and CML
365. The selected inbred lines based on their plant characteristics are presented in Table 6.

TABLE 6. SIX BEST INBRED LINES FOR SELECTED PLANT CHARACTERISTICS

Rank Grain yield Root volume Leaf area index Chlorophyll content Mycorrhizal
1 88069 88069 1368 CML.358 CML 254
2 ATP-S9-30Y-1 87036 ATP-S9-30-Y-1 ATP-S6-20-Y-1 88094
3 87036 Cam Inb gp1 17 9848 5012 M131
4 M131 9450 88069 ATP-S9-30-Y-1 87036
5 9071 9071 87036 M 131 ATP-S9-30-Y-1
6 ATP-S5-26Y-1 CML 365 CML 365 ATP-S5-31Y-1 CML 365

Mean grain yield ranged from 0.2 t ha-1 for 5012 x 9450 to 4.3 t ha-1 for CML 254 x
1368 on 20 N. Twenty (20) hybrids yielded more than 2.5 t ha-1, with 8 yielding more than
the check variety 87036 x Exp1.24 (3.0 t ha-1). The top 20% for grain yield on 20 N plots are
shown in Table 7. Grain yield varied between 2.4 t ha-1 for 9450 x Cam inb to 4.3 t ha-1 for
1368 x CML 254. The frequency of appearance of individual inbred parents showed that ATP
SR S6 20 Y-1 retained for its high chlorophyll content appeared 8 times, CML 254 selected
for its mycorrhizal efficiency appeared 5 times, 9071, Cam inb gp1.17 and 87036 selected for
their root volume appeared 4 times each, as well as ATP-SR S5 31 Y retained for its
chlorophyll content. Inbreds 5012, ATP SR S9 30 Y-1 and 1368 appeared 3 times. They were
retained for their chlorophyll content and for their leaf area index, respectively. M131 and
CML 358 did not appear as parents among the 20% top hybrids.

259
 Ten (10) hybrids yielded as much as or better than the check hybrids (Expl.24 x
87036: 3.0 t ha-1). Four had 23% grain superiority. Among the twenty parents involved in the
ten top hybrids, 8 were retained for their chlorophyll content, 6 for their root volume, 4 for
their leaf area index and 2 for their mycorrhizal colonisation.

TABLE 7. GRAIN YIELD OF THE 20 TOPS % ON 20 N, AS COMPARED TO PERFORMANCE

ON 100 N AND 20 N + MYCORRHIZA

Rank Hybrids Criteria† Yield 20 N % check

1 1368 x CML 254 Leaf x myc 4.3 143.3
2 1368 x 9848 Leaf x leaf 3.7 123.3
3 ATP S6-20-Y-1 x 5012 Chlo x chlo 3.7 127.3
4 ATP S6-20-Y-1 x ATPS5-31-Y-1 Chlo x chlo 3.7 123.3
5 ATP S6-20-Y-1 x 9071 Chlo x root 3.6 120.0
6 ATP S6-20-Y-1 x CML 254 Chlo x myc 3.6 120.0
7 CAM INB GR1 17 x 9071 Root x root 3.2 106.7
8 88094 x ATP S6-20-Y-1 Myc x chlo 3.1 103.3
9 87036 x ATP S9-30-Y-1 Root x leaf 3.0 100.0
10 9450 x ATP S6-20-Y-1 Root x chlo 3.0 100.0
11 87036 x ATP S6-20-Y-1 Root x chlo 2.9 96.6
12 CML 254 x 9071 Myc x root 2.8 93.3
13 87036 x ATP S5-31-Y-1 Root x chlo 2.7 90.0
14 ATP S5-31-Y-1 x 9071 Chlo x root 2.7 90.0
15 88094 x 5012 Myc x chlo 2.7 90.0
16 1368 x CAM INB GR1 17 Leaf x root 2.6 86.7
17 5012 X ATP S5-31-Y-1 Chlo x chlo 2.5 83.3
18 87036 X CAM INB GR1 17 Root x root 2.5 83.3
19 ATP S9-30-Y-1 X 88069 Leaf x root 2.5 83.3
20 ATP S9-30-Y-1 X CML 254 Leaf x myc 2.5 83.3
21 ATP S6-20-Y-1 X CML 254 Chlo x myc 2.4 80.0
22 9450 x CAM INB GR1 17 Root x root 2.4 80.0
† Chlo, chlorophyll; Root, root volume; myc, Mycorrhiza; Leaf, leaf area index

Results obtained on plots which received 20N and were inoculated with AM showed
that grain yield varied from 1.8 t ha-1 for 9848 x Cam inb gp1.17 to 4.5 t ha-1 for 1368 x CML
254. The check variety yielded 3.3 t ha-1, which represented a 7% increase over its value on
20 N plots. Eight hybrids yielded at least 3% better than the check, and 2 of them exhibited at
least 38 % grain yield superiority.

The top 20% yielding hybrids, presented in Table 8, exhibited a mean of 3.04 t ha-1.
This represented a 35% grain yield increase over their performance on 20 N. Thirteen hybrids
had at least 25% grain yield superiority over their performance on 20 N. CML 254, retained
for its mycorrhizal colonization, appeared 7 times, followed by 9848, 87036 and ATP-S5-31-
Y-1 which appeared four time each. These lines were selected for their leaf area index, root
volume, and for their chlorophyll content, respectively. Eight (8) hybrids yielded more than
the check (3.2 t ha-1). Among the 16 parents involved in these crosses, 5 were selected for
their root volume, four (4) parents for their mycorrhizal colonization and leaf area index,
respectively, and three (3) for their chlorophyll content. In fact, 29.5% of the crosses involved
at least one inbred selected for mycorrhizal colonization. Twenty seven percent involved at
least one line retained for its root volume, 25% had at least one line with good chlorophyll
content, and 18% involved lines retained for their leaf area index.

260
TABLE 8. GRAIN YIELD OF THE TOP 20% ON 20 N + MYCORRHIZA AS COMPARED TO
THEIR PERFORMANCE ON 20 N AND 100 N

Genotypes Criteria† Yield % Yield 20 N

myc. check % of 20 N
1368 X CML 254 Leaf x myc 4.5 136 4.3 105
ATP S9-30-Y-1 X 9450 Leaf x root 4.4 133 1.6 275
88069 X CML 254 Root x myc 3.7 112 0.9 411
CML 358 X CML 254 Chlo x myc 3.6 109 1.0 260
5012 X CAM Inb gr1 17 Chlo x root 3.5 106 1.8 194
88094 X 1368 Myc x leaf 3.4 103 1.5 227
9848 X 9071 Leaf x root 3.3 100 0.8 413
87036 X ATP S5 31 Y-1 Root x chlo 3.3 100 2.7 246
M131 X 5012 Myc x chlo 3.2 97 1.3 246
CML 358 X 88069 Chlo x root 3.2 97 1.9 356
87036 X ATP S9 30 Y-1 Root x leaf 3.1 94 3.0 103
88094 x ATP S5-31-Y-1 Myc x chlo 2.7 82 1.5 180
87036 X CML 254 Root x myc 2.7 79 2.0 135
ATP S5-31-Y-1 X CML 254 Chlo x myc 2.7 79 3.7 73
M131 X 9848 Myc x leaf 2.7 79 1.9 137
87036 X CML 358 Root x chlo 2.5 76 2.0 125
9450 X CML 254 Root x myc 2.5 76 1.2 208
1368 X 9848 Leaf x leaf 2.4 73 3.7 65
M131 x ATP-S5-31-Y-1 Myc x chlo 2.4 73 2.3 104
88094 X 5012 Myc x chlo 2.4 73 2.7 89
CML 254 X 9071 Myc x root 2.4 73 2.8 86
9450 X ATP S6-20-Y-1 Root x chlo 2.3 70 3.0 77
9848 X CAM Inb gr1 17 Leaf x root 2.3 70 1.5 153
†Chlo, chlorophyll; Root, root volume; myc, Mycorrhiza; Leaf, leaf area index

The top 20% of hybrids on 100 N are shown on Table 9. Grain yield ranged from 5.0 t
ha-1 for M131 x 9071 to 9.5 t ha-1 for ATP-SR-30-Y-1 x CML 254. Average grain yield was
5.9 t ha-1 as compared to 2.3 t ha-1 on 20 N. This represented a 257% yield superiority,
ranging from 123% for 1368 x CML 254 to 560% for 88069 x Cam inb gp1.17. The inbred
CML 254, appeared 7 times, followed by 9071 which appeared 6 times. Inbreds 88069 and
ATP-SR-30-Y-1 appeared 5 times. Inbred 88094 was involved in 4 crosses. Inbreds 87036,
M131 and ATP-SR-20-Y-1 appeared 3 times each. ATP-SR-31-Y-1, 9450 and Cam inb
gp1.17 appeared 2 times each. Finally, 1368 and 5012 were represented only once each. CML
358 and 9848 did not appear among the top 20%. The frequency of appearance of plant
characteristics indicated that on 100 N, 18 F1 crosses (41%), had at least one parent selected
for its root volume. Fourteen F1 (31.8%) involved at least one inbred selected for its
mycorrhizal colonization. Six F1 (13.6%) used at least one parent selected for its leaf area
index and chlorophyll content, respectively.

3.7. Combining ability of the lines

Significant mean square values for the orthogonal partition of the hybrid sum of
squares indicated general combining ability (GCA). Differences were obtained at the two soil
correction levels (20 N and 100 N) for chlorophyll content and grain yield. GCA for leaf area
index was significant only at 100 N. This indicated that these parameters were controlled by
additive gene action. Significant specific combining ability (SCA) effects were detected only
for grain yield at 100 N, indicating the presence of non additive gene action for this trait.

261
TABLE 9. GRAIN YIELD OF THE 20 TOP HYBRIDS ON 100 N COMPARED TO THEIR
PERFORMANCE AT 20 N

Rank Hybrids Criteria† Yield % check Yield % grain yield

100 N 100N 100 N increase/20 N
1 ATP S9-30-Y-1 x CML 254 Leaf x myc 9.5 176 2.5 380
2 ATP S9-30-y-1 x 88094 Leaf x myc 7.2 133 1.6 450
3 CML 254 x 9071 Myc x root 8.0 148 2.8 286
4 ATP S5-31-Y-1 x CML 254 Chlo x myc 6.7 124 3.6 186
5 ATP S9-30-Y-1 x 88069 Leaf x root 6.4 119 2.5 256
6 ATP S6-20-Y-1 x 9071 Chlo x root 6.1 113 3.6 169
7 9450 x CAM INB GR1 17 Root x root 6.0 111 2.4 250
8 88069 x 9071 Root x root 6.0 111 1.4 429
9 88069 x 88094 Root x myc 5.8 107 1.9 305
10 87036 x ATP S9-30-Y-1 Root x leaf 5.8 107 3.0 193
11 88069 x CAM INB GR1 17 Root x root 5.6 103 1.0 560
12 87036 x CML 254 Root x myc 5.6 103 2.0 280
13 ATP S5-31-Y-1 x 9071 Chlo x root 5.5 102 2.7 204
14 1368 x CML 254 Leaf x myc 5.3 98 4.3 123
15 ATP S9-30-Y-1 x ATP S6-20-Y-1 Leaf x chlo 5.3 98 1.9 279
16 M131 X 9450 Myc x root 5.2 96 1.2 433
17 87036 x ATP S6-20-Y-1 Root x chlo 5.1 94 2.9 176
18 88069 X CML254 Root x myc 5.1 94 0.9 286
19 88094 x 5012 Myc x chlo 5.1 94 2.7 189
20 88094 x 9071 Myc x root 5.1 94 1.8 283
21 M131 X CML 254 Myc x myc 5.0 93 1.8 278
22 M131 X 9071 Myc x root 5.0 93 1.1 455
† Chlo, chlorophyll; Root, root volume; myc, Mycorrhiza; Leaf, leaf area index

The combining ability of the 15 lines used in the F1 evaluation is shown in Table 10.
On 20 N soil, the best combiners were: ATP-SR-20-Y-1 (0.6), 87036 (0.4), ATP-SR-31-Y-1
(0.4), CML 254 (0.4), 1368 (0.2), 9071 (0.2) and ATP S9-30-Y-1 (0.1). On 20 N added to
mycorrhize, the best positive combiners were: ATP S9-30-Y-1 (0.7), M131 (0.4), 87036
(0.02) and CML 254 (0.0). On 100 N soil, the best combiners were: CML 254 (1.3), ATP-SR
S9-30-Y-1 (1.1), 88069 (0.4), 9071 (0.70), M131 (0.3), 88094 (0.2), ATP-S6-20-Y-1 (0.2).
Good combiners on 20 N and 100 N included: ATP SR S9 30 y-1, ATP-SR-20-Y-1, CML
254. Good combiners on 20 N plus mycorrhiza and 100 N included: M131, ATP-SR S9-30Y-
1 and CML 254 (0.0). Good combiners on 20 N and 20 N plus mycorrhiza included: 87036 ,
ATP-SR-30Y-1and CML 254. Finally the best combiners on the 3 soil types were only ATP-
SR-S9-30Y-1 and CML 254.

The five best specific combiners on 20 N included: 1368 x CML 254, 1368 x 9848,
ATP SR S9 30 Y-1 x ATP-SR S5 31 Y-1, 1368 x ATP-SR S6 20 Y-1 and CAM INB gp1.17 x
CML 254. The five best specific combiners on 100 N included: 9450 x CAM INB gp1.17,
9450 x 9071, CAM INB gp1.17 x CML 254, ATP SR S9 30 Y-1 x CML 254 and ATP SR S9
30 Y-1 x ATP SR S5 31 Y-1.

3.8. Correlation between grain yield and plant characteristics

The correlation values between grain yield and some measured plant characteristics
are shown in Table 11. At 20 N, all measured characteristic were positively correlated with

262
grain yield. The highest values were obtained for days to silk, chlorophyll content and root
volume. At 100 N, only number of green leaves, total number of leaves, and leaf area index
showed significant correlations with grain yield. It could be suggested that indirect selection
criteria for grain yield at 20 N, which could have a correlated response at 100 N included:
green leaves, total leaves and leaf area index.

TABLE 10. GRAIN COMBINING ABILITY ACROSS THE THREE ENVIRONMENTS

Inbred lines Yield (t ha-1)

20 N 20 N + mycorrhiza 100 N
87036 0.42 0.02 –0.01
M131 –0.40 0.40 0.30
ATP S9-30-Y-1 0.10 0.70 1.10
9450 –0.10 –0.50 –0.50
CML 358 –0.70 –0.01 –0.80
88069 –0.40 –0.60 0.40
88094 –0.10 –0.10 0.20
1368 0.20 –0.04 –0.40
ATP S6-20-Y-1 0.60 –0.01 0.20
9848 –0.40 –0.20 –0.80
5012 0.00 –0.20 –0.90
cam inb gp1 17 –0.10 –0.30 –0.90
ATP-S5-31-Y-1 0.40 –0.30 0.00
CML 254 0.40 0.00 1.30
9071 0.20 –0.10 0.70

TABLE 11. CORRELATION COEFFICIENTS BETWEEN GRAIN YIELD AND SELECTED

PLANT CHARACTERISTICS

Plant characteristics Grain yield

20 N 100 N
Days to anthesis 0.3** 0.15 ns
Days to silk 0.4*** 0.13 ns
A.S.I 0.3*** 0.02 ns
Green leaves 2 0.2* 0.35***
Total leaves 0.3*** 0.29***
Leaf area 0.3*** 0.26***
Root Volume 0.4*** 0.12 ns
Chlorophyl 2 0.4*** 0.03 ns
Chlorophyl 3 0.4*** 0.14 ns
Leave Senescence 1 0.2* 0.03 ns
Leave Senescence 2 0.2* 0.08 ns
*, **, *** denote P<0.05, P<0.01, P<0.001, respectively; ns, not significant

4. CONCLUSIONS

This study revealed that four F1 crosses were high yielding on the 3 types of soil N
correction (20 N, 20 N + mycorrhiza, 100 N). These were: 1368 x CML 254; 87036 x ATP-
SR-S9-30-Y-1; CML 254 x 9071; and 88094 x 5012. Three out of the 4 crosses involved at
least one parent retained for its root volume. Three hybrids performed well on 20 N as well as
20 N + mycorrhizal inoculation. These were: 1368 x 9848; 9450 x ATP-S6-20-Y-1 and 87037
x ATP-S5-31-Y-1. These hybrids included parents selected for root volume, chlorophyll
content and for leaf area index. Based on performance between 20 N and 100 N, 6 hybrids
were identified: ATP-SR-S6-20-Y-1 x 9071; 87036 x ATP-SR-S6-20-Y-1; ATP-SR-S5-31-Y-

263
1 x 9071; ATP-S9-30-Y-1 x 88069; ATP-SR-S9-30-Y-1 x CML 254 and 9450 x Cam inb
gp1.17. These genotypes involved 6 parents selected for their root volume, 3 parents retained
for their leaf area index, 2 parents for their chlorophyll content and one parent with
mycorrhizal use efficiency. Finally, when comparing performance on 20 N with mycorrhiza
and 100 N, three hybrids were as good in the one treatment as in the other: 88069 x CML 254;
87036 x CML 254 and ATP-SR-S5-31-Y-1 x CML 254. These involved 2 inbreds with good
root volume crossed to inbreds with good mycorrhizal use efficiency. These findings
suggested that inbred selection for root volume was important. However, for maximum
utilisation in a breeding programme, those lines must be crossed to another line having
complementary characteristics such as mycorrhizal use efficiency, good leaf area index, or
good stay green character as evidenced by high chlorophyll content.

REFERENCES

[1] MOSISA, W., Genetic and Crop-Physiological Basis of Nitrogen Efficiency in

Tropical Maize: Field Studies, Ph. D Thesis, University of Hannover, Germany (2005)
109 p.
[2] HEUBERGER, H.T., KLING, J.G., HORST, W.J., “Effect of root growth
characteristics on nitrogen use efficiency of tropical maize (Zea mays L.) varieties”,
Maize Research for Stress Environments, (JEWELL, D.C., et al., Eds), Proceedings of
the Fourth Eastern and Southern Africa Regional Maize Conference, CIMMYT,
Mexico, D.F. (1995) 44–48.
[3] GALLAIS, A., HIREL, B., An approach to the genetics of nitrogen use efficiency in
maize, J. Exp. Bot. 55 (2004) 295–306.
[4] BECK, D.L., et al., “Progress in developing drought and low soil N tolerance in
maize”, Proc. 51st Ann. Corn and Sorghum Res. Conf., (WILKINSON, D., Ed.),
ASTA, Washington, DC (1996) 85–111.
[5] HALLAUER, A.R. MIRANDA, J.B., Quantitative Genetics in Maize Breeding, Iowa
State University Press, Ames, Iowa (1988).
[6] BANZIGER, M., LAFITTE, I.R., Efficiency of secondary traits for improving maize
for low nitrogen target environments, Crop Sci. 37 (1997) 1110–1117.
[7] KAEPPLER, S.M., et al., Variation among maize inbred lines and detection of
quantitative trait loci for growth at low phosphorus and responsiveness to arbuscular
mycorrhizal fungi, Crop Sci. 40 (2000) 358–364.
[8] BAREA, J.M., AZCON-AGUILAR, C., AZCON, R., Vesicular-arbuscular
mycorrhiza improve both symbiotic N2 fixation and N uptake from soil as assessed
with a N-15 technique under field conditions, New Phytol. 106 (1987) 717–726.
[9] AZCON R., RUIZ-LOZANO J., RODRIGUEZ R., Differential contribution of
arbuscular mycorrhizal fungi to plant nitrate uptake (N-15) under increasing N supply
to the soil, Can. J. Bot. 79 (2001) 1175–1180.
[10] CRUZ, C., et al., Functional aspects of root architecture and mycorrhizal inoculation
with respect to nutrient uptake capacity, Mycorrhiza 14 (2004) 177–184.
[11] MÄDER, P., et al., Transport of 15N from a soil compartment separated by a
polytetrafluoroethylene membrane to plant roots via the hyphae of arbuscular
mycorrhizal fungi, New Phytol. 146 (2000) 155–161.
[12] NGONKEU, M.E.L., Biodiversité et Potentiel des Mycorhizes a Arbuscules de
Certaines Zones Agro Ecologiques du Cameroun, These de Doctorat 3 ieme cycle,
Universite de Yaounde 2, Cameroun (2003) 259 p.
[13] BRUNDREH, M.C., et al., Working with Mycorhizas in Forestry and Agriculture,
ACIAR Monograph 32, Canberra (1996) 374 p.

264
 PRODUCTIVITY OF UPLAND RICE GENOTYPES UNDER DIFFERENT
NITROGEN DOSES

K. TRAORE, O. TRAORE
INERA /Station de Farakoba,
Bobo-Dioulasso,
Burkina Faso

V.B. BADO
Africa Rice Center (AfricaRice),
Saint Louis,
Senegal

Abstract

Nitrogen (N) deficiency is one of the most yield-limiting nutrients in upland rice growing area
in Burkina Faso. A field experiment was carried out from 2008 to 2010 in Farakoba research center
with the objective to evaluate 200 upland rice (Oryza sativa L.) genotypes from WAB, NERICA,
CNA, CNAX, IRAT and IR lines for N use efficiency. The treatments consisted of three levels of N:
low, medium and high at 20, 60 and 100 kg−N ha-1, respectively. Both grain and straw yield increased
with N application. The yields were highest for NERICA and WAB lines compared to the other lines,
and this was consistent over the N doses. A large variability was found among the genotypes. Three
groups of genotypes were identified according to N use efficiency. The high N use efficiency
genotypes were found in WAB and NERICA lines. The N concentration in the shoot at flowering
significantly increased with N doses and this was similar for N taken up by genotypes.

1. INTRODUCTION

The demand for rice in sub-Saharan Africa is growing faster (by 5.6% per year) than
for any other major food staple. Urbanization and changes in employment patterns are driving
changing consumer preferences, such that rice is no longer a luxury food. West Africa
currently imports 40 to 60 % of the total rice consumed. Urgent action is needed to ensure that
African countries will have enough rice to feed their rural and urban populations. Rice is a
strategic crop in West Africa; it constitutes an important source of food and farm income for
many rural households in the region. In the specific case of Burkina Faso, the average annual
paddy production is 100 000 t which is far below the need of the country, estimated to be
304 986 t. Eighty-two % of the demand in rice is imported, corresponding to 40 billion F
CFA. However, in the most appropriate ecological zones, farmers are increasingly involved in
producing upland rice. For instance, the area cropped with upland rice was increased by 50%
in the south Sudanian zone of Burkina Faso during the past 10 years [1]. The current level of
upland rice production can be significantly increased by using improved and adapted rice
cultivars in combination with low-cost agronomic practices.

The possibility to increase total rice production in Burkina is possible through the
improvement of the upland system. However, yields are very low under this system (around
800 kg ha-1) due mainly to inappropriate farming methods and low inherent soil fertility.
Phosphorus (P) and nitrogen (N) are the most limiting nutrients for crop production in the
Sahel. Nitrogen is one of the most important nutrients for rice production because it
contributes to chlorophyll, amino acid and protein synthesis [2]. Nitrogen is necessary during
the early growth stages of rice as it gives vigour and increases tillering. The development of
rice varieties tolerant to low soil nutrients and efficient in using N is important for the many
thousands of poor farmers in the Sahel. The current study was carried out to evaluate the N

265
use efficiency by selected upland rice cultivars developed by the Burkina National Institute
for Agricultural Research (INERA) in partnership with the West African Rice Network
(WARDA). These varieties are high yielding and are efficient in N and P recovery under poor
soil conditions. We assume that genotypes which are efficient in using N are also tolerant to
low N conditions.

2. MATERIAL AND METHODS

2.1. Sites characterization

The experiment was carried out from 2008 to 2010 in Farako-Bâ research center in
Burkina Faso, West Africa. The geographic coordinates of the centre are 11°06’ latitude
north, 4°20’ longitude west and 405m altitude. The climate is south Soudanian [3]
characterized by a long dry season (October to April) and a short wet season (May to
September). Average annual rainfall (1995-2005) in the area is 950 mm with a cropping
season of 5 months (May to September). Rainfall is characterized by large annual viability.
Average minimum and maximum temperatures are also variable and depend on the season.
During the dry season minimum and maximum temperatures vary between 17 and 37°C, and
between 10 and 32oC during the rainy season. Evaporation is also a function of the season and
varies from 8.7 mm d-1 in January-February to 3.7 mm d-1 in August. The main soil types in
the Farako-Bâ research center are Ferrasols (FAO classification) [4]. These soils are acidic
and very poor in organic matter, available P and total N (Table 1).

TABLE 1. SOME CHEMICAL PROPERTIES OF THE SOIL AT FARAKOBA IN 2007

Soil property Average values†

Total C (%) 0.29 (0.05)
Total N (%) 0.03 (0.004)
C: N ratio 8.6 (0.74)
Organic matter (%) 0.50 (0.08)
Total P (mg kg-1) 84 (10.9)
Available P (mg kg-1) 3.6 (0.79)
pHH2O 5.6 (0.15)
pHKCl 4.3 (0.17)
†Standard deviations are given in parentheses

2.2. Genotypes

Six (6) lines and 200 accessions of upland rice were used for the experiment (Table 2).

TABLE 2. LINES AND NUMBER OF ACCESSIONS

Line† Number of accessions

WAB 175
NERICA 18
CT 2
CNA 2
CNAX 1
IRAT 1
IR 1
Total 200
†Provided by AfricaRice, Senegal (the Sahelian center)

266
2.3. Experimental design and data collection

A pre-cropping with maize to deplete soil and homogenize soil N content was carried
out. The maize was planted in early May and was harvested 3 weeks later. The experimental
design was randomized blocks with 3 treatments, consisting of 3 doses of N as urea at low,
intermediate and high rates (20, 60 and 120 kg−N ha-1, respectively), with a split application at
15 and 45 days after sowing (DAS):

The size of blocks was 60 × 3 m = 180 m2. The blocks were separated by a 1 m
interval. There were four blocks. The distance between genotypes in the same block was 40
cm. The 200 genotypes were sown on continuous lines of 3 m. The field was amended with
400 kg ha-1 Burkina rock phosphate and 100 kg ha-1 of KCl to correct P and K deficiencies,
respectively. The Burkina rock phosphate and KCl were incorporated in the soil during land
preparation.

The following data were collected:

− Germination rate
− Number of tillers
− Day to flowering
− Day to maturity
− Panicle weight
− Grain weight
− Straw weight
Soil samples were taken before sowing at 0-20cm. Plant samples were collected at
flowering in each treatment (genotype × N).The samples (soil and plants) were pre-dried in
direct sun light and subsequently oven-dried at 60oC for 48 h. The samples were ground and
sieved to 2 mm.

2.4. Laboratory and statistical analysis

The soil and plant samples were analyzed for their N status in the soil-water-plant
laboratory at Farakoba research center. The N concentrations of samples were determined by
Kjeldahl digestion. Excel was used for primary data processing. The analysis of variance
(ANOVA) was performed using GENSTAT 7th Edition. The means were separated using the
least significant difference (LSD) at P<0.05.

3. RESULTS AND DISCUSSION

3.1. Environmental conditions

Rainfall conditions varied during the 3 years of the experiment. In 2008, the total
rainfall received was 1100 mm in 3 months with drought spells at the end of June and early
July. In 2009, 950 mm of rain was received at the site with an early ending (end of
September). Rainfall was adequate in 2010 (1172 mm) with a drought spell at the beginning
of the season (in June). This erratic rainfall affected the planting date which probably affected
the yield.

3.2. Yields

267
 The effect of N dose on yield (grain and straw) was different for genotypes and lines
(Table 3). N application increased both grain and straw yields for all the genotypes (P<0001),
and the higher the N dose, the higher the grain and straw yields (Table 3). These results are
consistent with those reported in the literature [5, 6] where significant yield differences
among upland rice genotypes under low, medium and high fertility levels were found in
Brazilian Oxisols under field conditions. The results are also supported by other findings [4,
7–9] where the most growth limiting nutrient was N. Over all the genotypes, the intermediate
dose generated an 11% grain yield increase compared with the low dose, while the high dose
gave a 29% increase compared with the low dose. However, there was a very large variability
among genotypes from the same line and between lines (P<0001). Some genotypes showed
very low harvest indices. In fact, lines CNA and CNAX produced twice as much straw as
grain. The genotypes from IR and IRAT produced >6 times more straw than grain (Table 3).

TABLE 3. UPLAND RICE GRAIN AND STRAW YIELDS

Line Grain yield (kg ha-1) Straw yield (kg ha-1)

N20 N60 N100 N20 N60 N100
CNA 681 915 972 1086 1417 1812
CNAX 692 746 974 1726 1948 2505
CT 692 746 974 1434 1607 1933
IR 156 282 417 1449 1881 2830
IRAT 44 77 283 1156 1208 1630
NERICA 889 1036 1231 1366 1577 1744
WAB 787 868 995 1337 1500 1816
Statistics Line: P<0.001, LSD = 78 Line: P = 0.17
N treatment : P<0.001, LSD = 452 N treatment : P<0.001, LSD = 87
Line x N treatment: P = 0.999, LSD = 783 Line x N treatment: P = 0.957

Under the low N conditions NERICA and WAB lines performed better compared to
the other lines (Table 3). The lowest yield was obtained with the IRAT genotypes and this
was consistent over the N treatments. The yields varied from 149 to 1743 kg ha-1 for NERICA
genotypes, but N application always increased grain yield (P<0.001) for all genotypes. The
response to N application was very high for NERICA 10, 11 and 13. Yields were lower for
NERICA 6, 7 and 15 independent of N treatments.

The yield increment from the intermediate to the highest N dose was only about 20%
(Fig. 1). The overall productivity of NERICA genotypes was very low compared to their
potential yield of around 2 t ha-1. Furthermore grain yields were comparables for NERICA 17
and NERICA 5 at N60 and N120. Two genotypes generated very low yields (Nerica 7 and
15). The low and intermediate doses were comparable for some genotypes. Except for Nerica
7 and 15, the grain yield was always above 500 kg ha-1 without N application. The genotypic
response to N doses allowed ranking in three N efficiency index groups (Fig. 1, Table 4): N
efficient genotypes (Efficiency index >0.26), medium N efficient genotypes (Efficiency index
between –0.224 and 0.014) and N inefficient genotypes (<–0.224). The efficient N genotypes
are those generating high yield under both low and high N dose.

Upland grain yields of CNA genotypes were lower than found for NERICA. The
yields for CNA genotypes varied from 400 to 1243 kg ha-1. N treatments significantly
increased upland grain yield. Again, the yield benefit was low except for the CNAX genotype
(Fig. 2). For the latter genotype, comparable yields were obtained for N20 and N60. The

268
response of WAB genotypes to N treatments varied the most compared to NERICA and CNA
genotypes, from 11 to 3023 kg ha-1. The following genotypes generated more than 2000 kg
ha-1: WAB 36-54, WAB 56-77, WAB 881-10-37-18-14-P1-HB, WAB 907-12-3-1-1-1-HB,
WAB 963-3A1.1, WAB711-B-2A1.1. The highest yield was obtained with the WAB 881-10-
37-18-14-P1-HB (3023 kg ha-1). These good yields were obtained with the high N dose. The
genotypes WAB788-58-1-2-HB and WAB963-B-12A1.2 and WAB775-21-5-2-HB showed
very low grain yield. The high dose of N generated the highest grain yield for all the
genotypes.

efficient genotypes Efficient genotypes at Low

genotypes at high N and High N
Grain yield standard deviation

Inefficient N Low N efficient

genotypes genotypes
N efficiency index

FIG. 1. N efficiency index vs. grain yield standard deviation

Variables coordinates (axes F1 and F2 : 67,74 %)

LHI sd

PR HI sd

LGY sd
F2 (21,08 %)

PR GY sd
LSY sd

PR SY sd

F1 (46,66 %)

FIG. 2. Potential contribution for each factor

269
 The yields for the efficient genotypes were around 1000 kg ha-1, and were almost
double with the high N dose, mainly for WAB 767 and WAB 881 (Fig. 3). These two
genotypes are potential genotypes for high grain production.

FIG. 3. Upland grain yield for N efficient genotypes at low and at high N doses.

3.3. N concentration

Genotypic N concentrations varied widely and were highest for the WAB and
NERICA genotypes. The concentration of N in the shoots varied from 0.7 to 1.6% for the
WAB genotypes (Fig. 4), and from 0.8 to 1.4 for the NERICA genotypes (Fig. 5). The
concentration of N in shoots was highest for WAB 340 BB 2H2 and lowest for WAB
711 132A1.1.5 (Fig. 4). For the NERICA genotypes the concentration of N in shoots was
stable, and NERICA 5 and NERICA 8 had a higher concentration of N in shoots with the high
N dose compared to the other genotypes from the same group (Fig. 5).

270
 TABLE 4. CLASSIFICATION OF GENOTYPES NITROGEN USE EFFICIENCY INDEX

Highly Efficient Moderately Efficient Inefficient

NERICA 13, CNA 6675, CNA 6675, CNA 6680, CNAX 17625J-48-B-1, IR 47686-13-2-2 IRAT 136, NERICA
NERICA 9, NERICA 1, NERICA 10, NERICA 11, NERICA 12, NERICA 14, NERICA 17, NERICA 18, NERICA 2, 15, WAB 1234-
NERICA5, WAB NERICA 3, NERICA 4, NERICA 6, NERICA 7, NERICA 8, WAB 1022-3-2-1-1-1-HB, WAB 1087-B-37A1.2, 1A9.2, WAB 1645-
1645-11A1.1 WAB 1092-B-40AB.1, WAB 1095-B-1A1.1, WAB 1275-6AB.1, WAB 1618-321-8-2A1.1, WAB 1645-3A1.1, 5A4.1, WAB 564-8-1-
WAB 1645-4AB.1, WAB 1645-3A1.2, WAB 1645-4A4.1, WAB 1645-5A1.2, WAB 1645-5A6.1, WAB 1645-6AB.1, WAB 1645- 2, WAB 712-56-4-1-
WAB 340-B-B-2- 7A13.2, WAB 1645-7A3.1, WAB 1645-7A5.2, WAB 1645-8A3.1, WAB 1645-8AB.1, WAB 272-B-B-2-H3, HB, WAB 725-27-3-
H2, WAB 36-54, WAB 450-12-2-BL1-DR3, WAB 502-12-2-1, WAB 502-18-4-1, WAB 56-77 1-1-2-HB, WAB 748-
WAB 515-B- WAB 616-30-3-2, WAB 704-17-4-HB, WAB 707-32-3-1-HB, WAB 709-26-4-1-HB, WAB 709-73-3-2-HB, 11-2-HB, WAB 767-
16A1.2 WAB 711-136-2, WAB 718-26-1-1-HB, WAB 748-13-2-HB, WAB 748-14-2-HB, WAB 767-4-2-1-HB, WAB 2-4-1-HB, WAB 880-
WAB 709-18-1-1- 775-84-3-2-HB, WAB 781-47-4-2-2-1-1-HB, WAB 788-18-2-1-HB, WAB 788-25-1-1-1-1-1-HB, WAB 788- 1-38-12-2-P1-HB,
HB, WAB 711- 58-2-1-HB, WAB 801-27-1-1-HB, WAB 854-B-50A1.1 WAB 880-1-38-19-
137-3, WAB 718- WAB 878 SG1, WAB 878 SG35, WAB 878 SG36, WAB 878 SG43, WAB 878-6-37-8-3-P1-HB, WAB 880 23-P1-HB, WAB 880-
27-2-1-HB, WAB SG34, WAB 880 SG37, WAB 880 SG62, WAB 880-1-131-1-16-P1-HB, WAB 880-1-32-1-2-P1-HB, WAB 1-38-20-26-P2-HB,
880 SG6 880-1-38-13-1-P1-HB, WAB 880-1-38-19-26-P2-HB, WAB 880-1-38-20-14-P1-HB, WAB 881-10-37-18-12- WAB1079-B-39A1.1,
WAB 880-1-38-18- P3-HB, WAB 881-10-37-18-14-P1-HB, WAB 881-10-37-18-15-P1-HB, WAB 881-10-37-18-5-P1-HB, WAB WAB569-35-1-2-1-
20-P3-HB, WAB 881-10-37-18-7-P3-HB HB, WAB569-36-1-1-
907-12-3-1-1-1- WAB 881-10-37-18-7-P4-HB, WAB 891 SG12, WAB 891 SG31, WAB 894-B-3A1.2, WAB 897-B-B-B-B-24, 1-HB, WAB616-54-
HB, WAB 963- WAB 950-B-93A1.2, WAB 995-6A1.1 11-1-1-1-HB,
3A1.1, WAB537- WAB 99-7, WAB1094-B-58AB.1, WAB375-B-12-H3-1, WAB450-11-1-1-P28-4-HB, WAB450-24-2-2-P33- WAB759-55-1-2-HB,
14-4-1-1-1-HB, HB, WAB450-I-B-P-422-HB, WAB570-32-2-1-1-1-HB, WAB570-35-2-1-HB, WAB616-53-4-3-1-1-HB, WAB775-21-5-2-HB,
WAB711-B-2A1.1 WAB721-13-1-1-1-3-HB, WAB757-17-1-HB, WAB767-2-5-1-HB, WAB775-104-2-1-HB WAB775-49-2-3-HB
WAB759-34-5-1- WAB775-49-1-1-HB, WAB781-75-3-1-HB, WAB788-19-1-1-2-HB, WAB804-23-1-1-2-HB, WAB880-1-38- WAB788-58-1-2-HB,
HB, WAB788-51- 20-15-P2-HB, WAB880-1-38-20-16-P2-HB, WAB880SG37, WAB880SG42, WAB880SG50, WAB881-10-37- WAB880SG73
3-2-HB, WAB881- 18-3-P1-HB, WAB881-10-37-18-8-P2-HB, WAB881SG1, WAB901-1A1.1, WAB901-1A2.1, WAB901-7A2.1, WAB923-B-6A1.1,
10-37-18-14-P2- WAB902-B-14A1.1, WAB902-B-16A1.1, WAB903-5-1-1-HB, WAB905-B-12AB.1, WAB910-B-3A1.1, WAB960-B-11A1.1
HB, WAB905-B- WAB910-B-4AB.1, WAB910-B-5AB.1, WAB915-B-3A1.1, WAB919-72-4-1-HB, WAB925-B-3A1.1,
2A1.1 WAB951-B-181AB.1, WAB952-B-47AB.1, WAB954-B-51AB.1, WAB963-B-12A1.2, WAB964-B-3A1.2,
WAB969-40-1-HB

271
 FIG. 4. Shoot N concentration (%) of WAB genotypes as a function of N treatments.

At lower N doses the N concentration was higher for NERICA genotypes than for the
other genotypes. The higher N taken up with increased N dose is related to deep root
development allowing the genotypes to explore more soil depth [10, 11]. N application
improves the use of carbohydrate by the plants and stimulates the development of the rooting
system, which increases the uptake of other nutrients necessary for plant growth [12].
Nitrogen is also essential for enzyme synthesis and photosynthesis [13]. The effect of N
nutrition is therefore an indirect effect through other processes.

FIG. 5. Shoot N concentration (%) of NERICA genotypes as a function of N treatments.

272
3.4. N uptake

Flowering is an important growth stage for rice because it’s when the translocation of
nutrients occurs from shoot to grain. The quantity of N taken up by upland rice shoots at
flowering showed highly significant differences between N treatments (P<0.001) and
genotypes (P = 0.030) (Table 5). No significant correlation was found between N treatments
and genotypes. N application increased N taken in the shoot at flowering this was consistent
for all genotypes and lines (Table 5). The highest quantity of N was taken up by genotype
WAB 340-B-B-2-H2 for all N treatments. The lowest quantity of N was taken up by NERICA
17. The variability between genotypes was lower for the NERICA compared to WAB
genotypes.

TABLE 5. NITROGEN TAKEN UP IN SHOOT BY GENOTYPES AT FLOWERING

Genotype N taken up in the shoot (kg−N ha-1)

N20 N60 N120
NERICA 15 10.8 16.0 31.8
NERICA 17 6.3 11.7 18.6
NERICA 18 12.1 19.7 27.5
NERICA 8 10.2 13.6 23.4
NERICA5 11.2 15.0 19.2
WAB 340-B-B-2-H2 16.2 24.7 48.8
WAB 748-11-2-HB 7.9 14.2 25.1
WAB 881-10-37-18-7-P4-HB 8.6 21.3 28.1
WAB450-24-2-2-P33-HB 13.3 18.8 25.2
WAB711-B-2A1.1 8.8 14.2 19.2
Statistics N treatment: P<0.001, LSD = 4.9
Genotype: P = 0.030, LSD = 9.0
N treatment x genotype: ns

4. CONCLUSIONS

N application significantly affected upland rice grain and straw yields; the higher the
N dose the higher the grain and straw yield. However, the level of yield increment was
different for genotypes and lines. The increment was lower for the NERICA than for the
WAB genotypes. At low N supply some genotypes from WAB and NERICA lines generated
sustainable upland grain yields. The harvest index of genotypes was only affected for IRAT,
and IR lines. With N application these genotypes accumulated more shoot than grain. A very
large variability was observed within the WAB and NERICA genotypes. The yields were
higher for the efficient WAB compared to the NERICA genotypes. N doses also affected the
N concentration and N uptake of genotypes. N concentration was higher at low N dose for
NERICA than for the WAB genotypes. The opposite situation was found with the WAB
genotypes. The high dose generated higher yields with WAB genotypes than with the other
lines. This study allowed the selection of nutrient use efficient genotypes, which will be
integrated into the farming system to boost productivity.

273
 REFERENCES

[1] INERA, Riz et Riziculture, Document préparatoire du plan stratégique du CNRST,

Programme riz et riziculture, Station de Farako–Bâ (2003) 46 p.
[2] ADRAO, Formation en Production Rizicole: Manuel du Formateur, A.D.R.A.O.
Bouaké, Cote d’Ivoire (1995) 305 p.
[3] GUNIKO, S., Végétation de la Haute Volta, Thèse de Doctorat ès-Sciences Naturelles,
Université Bordeaux III, Tome 1 (1984) 318 p.
[4] BADO, B.V., Rôle des Légumineuses sur la Fertilité des Sols Ferrugineux Tropicaux
des Zones Guinéennes et Soudaniennes du Burkina Faso, Ph. D thesis, Université
Laval (2002) 90 p.
[5] FAGERIA, N.K., et al., Differential response of upland rice genotypes to soil fertility,
Rev. Bras. Ciênc. Solo 19 (1995) 261–267.
[6] FAGERIA, N.K., SANTANA, E.P., MORAIS, O.P., Response of favourable upland
rice genotypes to soil fertility, Pesq. Agropec. Bras. 30 (1995) 1155–1161.
[7] BATIONO, A., LOMPO, F., KOALA, S., Research on nutrient flows and balances in
West Africa: state of the art, Agric. Ecosyst. Environ. 71 (1998) 19–35.
[8] TAPSOBA, D., Caractérisation Événementielle des Régimes Pluviométriques Quest-
Africains et de Leur Récent Changement, Thése de l'Université Paris XI (1997) 145 p.
[9] ARRAUDEAU, A., Le Riz Irrigué, édition Maisonneuve et Larose : le technician
d’agriculture tropicale (1998) 321p.
[10] KONDO, M., et al., Genotypic and environmental variations in root morphology in
rice genotypes under upland field conditions, Plant Soil 255 (2003) 189–200.
[11] DUAN, Y.H., et al., Responses of rice cultivars with different nitrogen use efficiency
to partial nitrate nutrition, Ann. Bot. 99 (2007) 1153–1160.
[12] STEVENSON, F.J., Cycles of Soil: Carbon, Nitrogen, Phosphorus, Sulfur,
Micronutrients, John Wiley & Sons, New York (1986).
[13] LAMAZE, T., et al., “Effet d’une limitation en N sur la photosynthèse chez le maïs”,
Physiologie et Production du Maïs, INRA, Paris (1990) 113–121.

274
 USING UPLAND RICE ROOT TRAITS TO IDENTIFY N USE EFFICIENT
GENOTYPES FOR LIMITED SOIL NUTRIENT CONDITIONS

K. TRAORE, O. TRAORE
INERA / Station de Farakoba,
Bobo-Dioulasso,
Burkina Faso

V.B. BADO
Africa Rice Center (AfricaRice),
Saint Louis,
Senegal

Abstract

Crop production in the Sahelian countries of Africa is limited by many factors. The most
important are low potential yields of local varieties, low inherent soil fertility and low applications of
external inputs (organic and mineral fertilizers). A field experiment was conducted from 2007 to 2008
with the objective to develop and validate screening protocols for plant traits that enhance N
acquisition and utilization in upland rice grown in low N soils of two hundred (200) upland rice
(Oryza sativa L.) genotypes from WAB, NERICA, CNA, CNAX, IRAT and IR lines. An experiment
in small pots was carried out in a greenhouse of Farakoba research center. The pots were filled with a
sandy soil and upland rice genotypes were grown during three weeks, harvested and studied for their
root characteristics (seminal root length, adventitious root number, lateral root length and number and
roots hair density). The small pot method was reliable for root trait characterisation at the seedling
stage. A large variability among genotypes was exhibited for the root characteristics. The variability
was larger within the NERICA and WAB lines compared to the other lines. The length of the seminal
roots varied from 10 to 40 cm, the lateral root number ranged between 3 and 15 and the number of
adventitious roots varied between 2 and 7. The selected root traits can be used to identify high
nutrients and water use efficient genotypes.

1. INTRODUCTION

Crop production in the Sahelian countries of Africa is limited by many factors [1]. The
most important are low potential yields of local varieties, low inherent soil fertility and low
applications of external inputs (organic and mineral fertilizers) [2–5]. The upland soils in the
area are deficient in N and P [6], although considerable efforts were devoted to the
improvement of soil conditions for crop production through the design and diffusion of best
soil-nutrient management practices. The adoption rate of these technologies remained low due
to mainly limited incomes of small poor farmers and socio-economic constraints. The design
and the use of crop genotypes adapted to these hard environments and the use of organic
amendments combined with small quantities of mineral fertilizers are of importance for
thousands of smallholders in the Sahel. The current level of upland rice production can be
significantly increased by using improved and adapted rice cultivars in combination with low-
cost agronomic practices. Recently, many varieties of upland rice were developed by the
Burkina National Institute for Agriculture Researches (INERA) in partnership with the West
African Rice Network (WARDA). These varieties are high yielding and are efficient in N and
P recovery under poor soil conditions. In most cases, the agronomic performances of these
new upland rice genotypes were proved by previous studies. But, specific traits that could
justify the high performances of these new genotypes are less studied. Furthermore,
information on appropriate fertilizer doses or optimal combination of agronomic practices for
optimum performances of these new genotypes is not available.

275
 Plant roots play an important role in water and nutrient acquisition. Both constitutive
and adaptive root growth have been implicated in the improved performance of rice under
rainfed conditions [7]. For the genetic improvement of the root system, the information on
genotypic diversity is required on the traits related to the size and distribution of the root
system [8]. This study aimed to clarify the magnitude of the effect of genotype on the
variations in root traits related to the size and distribution of the upland rice root systems. The
main objective was to identify the best upland rice varieties able to improve the smallholders’
cropping system productivity through a better use of nutrients, particularly N from soil and
fertilizers. The study aimed specifically to develop and validate screening protocols for plant
traits that enhance N acquisition and utilization in upland rice grown in low fertility soils.

2. MATERIAL AND METHODS

The experiment was carried out in Farako-Bâ research center in the south-Sudanian
zone of Burkina Faso from 2007 to 2008.

2.1. Plant material

Six (6 lines) and 200 genotypes of upland rice were used in the experiments (Table 1).
The upland lines were received from AfricaRice in 2007.

TABLE 1. LINES AND NUMBER OF ACCESSIONS

Line Number of accessions

WAB 175
NERICA 18
CT 2
CNA 2
CNAX 1
IRAT 1
IR 1
Total 200

2.2. Pot experiments

Two screening experiments were carried out. The first one started on May 21, 2007
and ended on June 30, 2008. The second screening was from January 8, 2008 to February 14,
2008. The soil used to grow the genotypes was a sandy soil collected in the center (bulk
density, 1.5 g cm-3, organic matter, 7%; total P, 323 mg kg-1, available P, 16 mg kg-1, pHKCl,
6.0 and pHwater, 6.7). The soil was thoroughly mixed to make it uniform. Cylindrical PVC pots
of l liter capacity were filled with 700 g of sandy soil, the pots were watered to field capacity
and allowed to equilibrate before rice was panted. The pots were perforated at their base to
allow adequate water and air circulation. Five (5) grains of each of the 200 upland rice
genotypes were planted in each pot. The pots were thinned after emergence to 3 plants per
pot. The pots were randomly arranged in a screen house at ambient temperature in three
blocks corresponding to repetitions. Pots were watered once a day (morning).

2.3. Root parameter measurements

Fourteen days after sowing, the soil was carefully washed with tap water and roots
were sampled. The following parameters were measured:

− plant height

276
 − number of leaves
− seminal root length (SRL) was measured as the length of the longest root
− adventitious root numbers (ARN) were counted for each plant
− the lateral root length (LRL) and lateral root number (LRN)
− root hair density
− the above and below ground biomass of each genotype was separated, sun dried for 2
days and oven dried at 70°C during 24 h

3. RESULTS AND DISCUSSION

3.1. Screening at the seedling stage

Genotypes were significantly different for all variables except for the straw dry weight
(Table 2). Positive correlations were found between plant height and the number of lateral
roots, the number of leaves and genotype straw dry weight (Table 3), and between seminal
root length and lateral root length. The straw dry weight and root dry weight were also
positively correlated (Table 3).

TABLE 2. ANOVA ON GENOTYPE ROOT CHARACTERISTICS

Variables Probability LSD

Plant height (cm) <0.001 7.60
Number of leaves <0.001 0.82
Seminal root length (cm) <0.001 6.18
Number of lateral roots <0.001 3.87
Number of adventitious roots <0.001 1.41
Lateral root length (cm) <0.001 6.12
Number of root hairs <0.001 0.68
Straw dry weight (mg) 0.075 75.2
Root dry weight (mg) <0.001 48.7

The genotypes were ranked in 6 groups according to the root characteristics (Fig. 1).

3.2. Rooting system and physiological characteristics of upland rice genotypes

3.2.1. Plant height

Genotype heights ranged between 25 and 55 cm (Fig. 2). Most of the genotypes
showed plant height around 40 cm. Only one genotype was higher than 50 cm and two
genotypes showed plant height below 30 cm. The genotype WAB 537 144 11 HB was the
shortest one while 12 genotypes from the WAB line showed a plant height above 50 cm.
Genotypes WAB 5677 and WAB 788 18 2 1 HB were the highest.

277
278
TABLE 3. CORRELATION COEFFICIENTS AMONG ROOT CHARACTERISTICS

Parameter Plant Seminal Lateral root Number of Number of Number of Straw dry Root dry Roots hair
height root length length leaves adventitious roots lateral roots weight weight density
Plant height 1
Seminal root length 0.196 1
Lateral root length –0.251 0.436** 1
Number of leaves 0.310* 0.289 0.168 1
Number of adventitious roots 0.218 0.087 –0.053 0.172 1
Number of lateral roots 0.461** 0.01 –0.341 0.168 0.169 1
Straw dry weight 0.308* 0.069 –0.092 0.139 0.077 0.175 1
Root dry weight –0.007 0.110 0.084 0.035 0.035 0.040 0.410** 1
Roots hair density 0.116 0.163 0.194 0.092 0.107 0.110 0.109 0.137 1
*, P<0.05; **, P<0.01
 FIG. 1 . Classification of genotypes according to root characteristics.

FIG. 2. Heights (cm) of genotypes 3 weeks after sowing.

3.2.2. Root length

The seminal root length showed different pattern for genotypes (Fig. 3). The data are
scattered and ranged between 10 and 40 cm. This root trait could be a potential index for
difference between genotypes. Deep rooting has been emphasized as an important adaptation
to stress in rice [9], and it was also reported that upland rice genotypes have seminal root
lengths above 10 cm in Burkina Faso [10]. Long seminal roots mean future deep rooting,
which is necessary for water uptake [11–13].

279
 FIG. 3. Seminal root length (cm) of genotypes 3 weeks after sowing.

3.2.3. Number of lateral roots

The number of lateral roots varied from 3 to 15 (Fig. 4). This character seems to be an
important trait marking difference between genotypes that will probably be important for
nutrient uptake and accumulation by genotypes. The lateral roots show the possibility of the
genotype to explore the upper layer of the soil. The number of lateral roots was lower than the
number indicated in the literature. This is due to the fact that the plants were very young and
this was also observed previously [14, 15]. Genotypes with a high number of lateral roots will
develop a shallow rooting system which is not very good for N and P uptake. A higher lateral
root number can be also a characteristic for adaptation to drought [16].

3.2.4. Lateral root length

Lateral root length showed a less scattered pattern compared to the number of lateral
roots. The number varied between 9 and 25 cm (Fig. 5). Most of the genotypes showed a
length between 15 and 20 cm. Only one genotype showed a lateral root length around 30 cm.

3.2.5. Number of adventitious roots

The adventitious roots are precursors of lateral and seminal roots, and are therefore
important for the crop during the vegetative and the reproductive phases. Adventitious root
numbers differed among genotypes and varied between 2 and 7. Most of the genotypes had 2
to 5 adventitious roots. One genotype had more than 6 adventitious roots (Fig. 6).

3.2.6. Number of leaves

The number of leaves per genotype varied between 3 and 7. Most of the genotypes had
4 to 6 leaves. This physiological character was not different between genotypes (Fig. 7).
These results are in line with those from [17] who found 5 leaves 20 days after sowing. The
number is therefore a genotype dependent trait.

280
 FIG. 4. Number of lateral roots of genotypes.

FIG. 5. Lateral root lengths (cm) of genotypes.

3.2.7. Roots hair density

Roots hair density differed highly among genotypes and varied between 1 and 3 (Fig.
8). Based on this root trait the genotypes can be ranked in different groups. This character is
also important for water and nutrient uptake from the soil. Root hairs form from root
epidermal cells. It is presumed that root hairs contribute to the adhesion of the growing root to
the rhizosphere and assist in the uptake of nutrients and water from the soil by increasing the
absorptive surface area [18]. This character can be of importance for plant N uptake.

281
 FIG. 6. Number of adventitious roots of genotypes.

FIG. 7. Number of leaves of genotypes.

FIG. 8. Root hair density of genotypes.

282
3.3. Rooting system for upland rice lines

The lines differed for all variables except for plant height, number of leaves and root
hair density (Table 4). The lowest plant height was obtained for the CNA line and plant
heights were comparable for the other lines. For lateral roots and seminal root lengths, the
lowest value was found for the CNAX line. The CNA line had a very long seminal root. The
lengths of seminal roots were comparable for the other lines.

All of the genotypes presented on average 4 well developed leaves, and this character
was not significantly different for lines. The CNAX line showed a very high number of
adventitious roots while the CNA and IR lines showed the lowest number. The number of
lateral roots was comparable for CNA, CNAX, NERICA and WAB, but lower compared to
the other lines. Straw and roots dry weight showed the same pattern. The values were lowest
for CNAX and CT and highest for NERICA.

The data showed also large variability among lines, especially for WAB and NERICA,
with respect to minimum and maximum values (Table 5). The variability was consistent over
all variables and was very high for lateral and seminal root lengths, the number of lateral roots
and straw and roots dry weight.

4. CONCLUSIONS

From the current study we concluded that the small pots experiment is a reliable
method for root trait characteristics for seedlings. The results showed that:

− Upland rice genotypes differed for root traits, and adventitious roots, lateral root
number, and root hair density are potential traits to characterize nutrients and water
uptake from the soil which is very important for low inherent fertility soils in the
Sahel.

− Very large variability among root characteristics within the upland rice lines. The
variability was much higher for WAB and NERICA genotypes compared to the other
lines, and this was consistent over all the measured variables.

− The correlations between yield components and roots traits showed no clear pattern
even though positive correlations were found between yields and seminal root length
and number of leaves.

− Further investigation should be carried out on the selected root traits and also on plant
physiological characteristics (nutrient uptake, nutrient use efficiency and nutrient
content in the shoots and the roots) for possible significant correlation with root traits.

283
284
TABLE 4. ROOT CHARACTERISTICS OF UPLAND RICE GENOTYPES AT THE SEEDLING STAGE 3 WEEKS AFTER EMERGENCE

Line Plant Lateral root Seminal root Number of Number of Number of Shoot dry Roots Roots hair density
Height length length leaves adventitious lateral weight weight
(cm) (cm) (cm) roots roots (mg) (mg)
CNA 41.10 22.01 31.01 4.16 2.92 6.67 193.3 99.3 2.25
CNAX 37.02 11.80 18.25 5.00 6.67 5.67 116.8 55.2 2.00
CT 39.08 14.67 26.77 5.00 3.67 12.33 167.7 100.7 2.00
IR 41.03 20.27 25.03 4.50 2.83 11.33 207.0 138.5 2.50
IRAT 40.27 16.10 24.60 4.25 3.75 11.50 214.2 139.0 2.50
NERICA 38.69 17.10 23.74 4.31 3.47 8.23 227.2 111.1 2.01
WAB 38.27 17.31 25.56 4.77 3.38 7.34 195.0 111.8 2.13
F pr. 0.24 0.037 0.002 0.071 <.001 0.003 0.028 0.120 0.57
LSD 7.94 6.38 7.01 0.89 1.52 4.12 98.63 55.83 1.18

TABLE 5. MINIMUM (MIN) AND MAXIMUM (MAX) VALUES FOR LINES SCREENED AT THE SEEDLING STAGE

Line Plant Root length (cm) Number of Dry weight plant-1 (mg)
height (cm) Lateral Seminal Leaves Adventitious roots Lateral roots Root hairs Straw Roots
Min Max Min Max Min Max Min Max Min Max Min Max Min Max Min Max Min Max
CNA 25.0 48.0 12.0 33.6 23.0 43.5 3 5 1 4 0 12 1 3 100 300 53 128
CNAX 24.0 39.5 3.0 21.3 7.0 31.0 4 6 3 12 4 7 1 3 54 200 4 100
CT 30.5 46.5 12.0 19.0 19.0 32.3 4 6 2 5 7 17 2 2 43 233 29 138
IR 39.7 42.0 14.0 25.0 19.5 32.0 4 6 2 4 10 13 1 3 100 300 100 171
IRAT 28.0 50.3 11.5 25.9 19.9 32.0 4 5 3 5 9 14 1 9 100 280 95 171
NERICA 20.0 54.0 1.5 37.0 11.5 48.0 2 6 0 6 0 17 1 3 40 442 12 200
WAB 14.5 69.5 1.5 42.6 4.4 52.5 2 9 0 13.5 0 22 1 3 21 560 7 740
 REFERENCES

[1] BATIONO, A., LOMPO, F., KOALA, S., Research on nutrient flows and balances in
West Africa: state of the art, Agric. Ecosyst. Environ. 71 (1998) 19–35.
[2] BADO, B.V., Rôle des Légumineuses sur la Fertilité des Sols Ferrugineux Tropicaux
des Zones Guinéennes et Soudaniennes du Burkina Faso, Ph. D thesis, Université
Laval (2002) 90 p.
[3] ZOUGMORE, R.B., Integrated Water and Nutrient Management for Sorghum
Production in Semi-arid Burkina Faso (2003) 205 p.
[4] OUEDRAOGO, E., Soil Quality Improvement for Crop Production in Semi-arid West
Africa. Tropical resource management papers no 51, Dissertation, Wageningen
University (2004) 193 p.
[5] TRAORE, O., et al., Crop rotation and soil amendments: impacts on cotton and maize
production in a cotton based system in western Burkina Faso, Int. J. Biol. Chem. Sci.,
1 (2007) 143–150.
[6] LOMPO, F., Etude de cas au Burkina Faso de l'initiative phosphates naturels, Rapport
provisoire, INERA, Burkina Faso (1995) 36 p.
[7] ZHENG, B.S., et al., Mapping QTLs and candidate genes for rice root traits under
different water-supply conditions and comparative analysis across three populations,
Theor. Appl. Genet. 107 (2003) 1505–1515.
[8] KONDO, M., et al., Genotypic and environmental variations in root morphology in
rice genotypes under upland field conditions, Plant Soil 255 (2003) 189–200.
[9] NGUYEN, H.T., BABU, R.C., BLUM, A., Breeding for drought resistance in rice:
physiology and molecular genetics consideration, Crop Sci. 37 (1997) 1426–1434.
[10] ARRAUDEAU, M.A., VERGARA, B.S., Manuel Illustré de Riziculture Pluvial (1997)
284 p.
[11] MAMBANI, B., LAL, R., Responses of upland rice varieties to drought stress. II.
Screening rice varieties by means of variable moisture regimes along a toposequence,
Plant Soil 73 (1983) 73–94.
[12] LILLEY, J.M., FUKAI, S., Effect of timing and severity of water deficit on four
diverse rice cultivars. 1. Rooting pattern and soil water extraction, Field Crops Res. 37
(1994) 205–213.
[13] KATO, Y., et al., Growth of rice (Oryza sativa L.) cultivars under upland conditions
with different levels of water supply. 3. Root system development, water extraction
and plant water status, Plant Prod. Sci. 10 (2007) 3–13.
[14] ANGLADETTE, A., LE RIZ, G-P., Maisonneuve et Larousse, Paris (1966) 930 p.
[15] LACHARME, M., Le plant de riz: données morphologiques et cycle de la plante
(2001) 1–36.
[16] ARRAUDEAU, A., Le Riz, Irrigué edition, Maisonneuve et Larousse: le Technicien
d’agriculture Tropicale (1998) 321 p.
[17] ADRAO, Formation en Production Rizicole: Manuel du Formateur, A.D.R.A.O.
Bouaké, Cote d’Ivoire (1995) 305 p.
[18] MA, J.F., et al., Role of root hairs and lateral roots in silicon uptake by rice, Plant
Physiol. 127 (2001) 1773–1780.

285
 EVALUATION OF UPLAND RICE GENOTYPES FOR EFFICIENT UPTAKE
OF NITROGEN AND PHOSPHORUS

A.R. ZAHARAH, M.M. HANAFI

Universiti Putra Malaysia,
Serdang, Selangor,
Malaysia

Abstract

Upland rice grown by subsistence farmers in the tropics and subtropics is known to produce
very low yields due to it being planted on low fertility soils and under drought-prone conditions. Little
information is available on upland rice cultivar differences in response to N and P fertilization in Asia,
thus screening for P (PUE) and N use efficiency (NUE) of upland rice genotypes is a necessary first
step. The objectives of the study were: (i) to identify upland rice genotypes with root characteristics
favorable for efficient N and P uptake and utilization, (ii) to evaluate the selected genotypes for their
grain yield, and (iii) to assess the variability of N and P use efficiency in upland rice genotypes grown
under field conditions. Several laboratory, glasshouse and field experiments were carried out from
2007 to 2011 at Universiti Putra Malaysia to achieve the above objectives. Fifteen local and 15 upland
rice genotypes from WARDA were identified to have long roots, and it was observed that some of the
WARDA lines showed longer root length than the local landraces. This is a good trait since it is
known that longer root length will enhance the absorption of easily mobile nutrients such as nitrate
and potassium. Glasshouse and field evaluation of N use efficiency by these upland rice genotypes
showed that high N is utilized (40–80% of applied N), with good grain yield, and P use efficiency is
similar to other crops (4–8%).

1. INTRODUCTION

Upland rice is grown as a staple food by subsistence farmers on approximately 17

million hectares worldwide [1]. One of the key constraints to food production under these
conditions is N and P deficiency, often associated with low organic matter, high P fixation
and severe soil acidity. Although indigenous upland rice cultivars appear well suited to these
conditions, being both low-P and high-acidity tolerant relative to other cereals, repeated
cropping without fertilizer addition can only be sustained by shifting cultivation and long
fallow periods. A key objective for upland rice research is therefore to improve cultivar yield
potential under the prevailing conditions [2].

Average yields of rain fed upland rice are 1.1 t ha_1 but this varies according to soil
type, fertilizer use, rainfall and agronomic practices [1]. Increased root length density in
upland rice is important for maintaining plant water status. Deeper and thicker roots are
hypothesized to improve performance of rain-fed rice under drought by more efficient water
extraction from deep soil layers. Drought-adapted varieties tend to have longer roots [3]. The
low yield of upland rice is largely a consequence of its production being limited to infertile or
drought-prone uplands, and to low harvest index of traditional cultivars [4]. Traditional
cultivars are generally tall, have few tillers, and produce low but stable yields under
unfavorable environments. They tend to lodge under favorable conditions and are thus not
suited to high-input management.

Genetic selection and plant breeding techniques helped to develop rice varieties that
are resistant to pests, diseases, and adverse environmental conditions such as drought, nutrient
deficiencies, toxicities, and salinity. However, genetic selection to improve the rice crop's
nutrient-use efficiency has received little attention. The need for developing and identifying
superior nutrient efficient genotypes, especially for N and P is evident because of the low

287
availability of these nutrients in tropical soils. For P, it was found that total root length of
winter wheat was positively correlated to grain yield [5]. However, few breeders have paid
attention to the possibility of manipulating root length for making crop genotypes P efficient.
This may be partly due to the technical difficulties and expense involved in studying root
systems. Model simulations of plant root systems for predicting optimum root architecture for
P acquisition in low P conditions [6] are promising. However, validation of such models
through fioun experiments still remains a problem as development of root systems is
influenced by a large number of soil and climatic factors [7],

Another limitation to plant growth in many of the highly weathered tropical and sub-
tropical humid soils (Ultisols and Oxisols) is the availability of phosphorus (P) due to
inherently low P content which is exacerbated by high P fixing capacities of added P
fertilizers. Malaysia is no exception, and therefore P deficiency is the major constraint to
upland rice production resulting in low yields. Conventional amelioration of P deficiency by
application of fertilizer is becoming increasingly uneconomical and ecologically unsound as
the efficiency of added fertilizer is low ≈ 10% [8]. Consequently, improvement of P uptake
and P use efficiency (PUE) by crops is critical as an economic, environmental-friendly and
sustainable strategy. The key to increase P recovery from less accessible forms is using crop
cultivars that are more efficient in P use. High PUE is achieved through efficient internal P
utilization and/or increased acquisition of more P from soil. The mechanisms by which these
are achieved include: (i) Exudation of chemical compounds such as phosphatase into the
rhizosphere to increase the solubility of P in the soil [8]. (ii) Changes of root morphology
under P deficiency to permit the plant to explore a greater volume of soil [9]. (iii) Increase in
the P uptake kinetics via activating special high affinity root transporters in P deficient
conditions. Little information is available on upland rice cultivar differences in response to N
and P fertilization in Asia, thus screening for high P and N use efficiency of upland rice
genotypes is necessary to develop varieties with high PUE and NUE. The objectives of the
study were:

(i) to identify upland rice genotypes with root characteristics favorable for efficient N
and P uptake and utilization
(ii) to evaluate the selected genotypes for their N and P response on grain yield
(iii) to assess the variability of N and P use efficiency in upland rice genotypes grown
under field conditions.

2. MATERIALS AND METHODS

2.1. Gigar roll method

Upland rice seeds from each land race (30 local and 200 from WARDA) were soaked
in water and allowed to germinate using the cigar-roll method. Roots that developed were
measured after 14 d using the WINRhizo Image Scanner EPSON Perfection V700 PHOTO.
Root characteristics such as length, diameter, surface area, volume and number of root tips
were measured.

2.2. Nitrogen uptake and N use efficiency

2.2.1. Glasshouse experiments

288
 Two glasshouse experiments determined the N use efficiency of the upland rice
genotypes. The first batch was planted on 3rd September 2007. The following genotypes
were used in four replications:

1. WAB788-56-2-1-2-HB (WARDA No. 17)

2. WAB767-2-5-1-HB (WARDA No. 51)
3. WAB915-B-3A1.1 (WARDA No. 69)
4. NERICA 7 (WARDA No. 143)
5. KENINGAU AN749
6. BATU GARAM AN1249
7. NABAWAN AN773
8. BECOR
9. PULUT PETAI
10. SELAYANG

The first fertilizer application was made on 12th Sept 2007 using 8 g TSP; 2 g MOP; 2
g urea with 2.03 atom % 15N excess. The second fertilizer application was on 29th Oct 2007
with 1g MOP and 1 g 15N labeled urea, and the third fertilizer application was on 12th
November 2007 using 1g MOP and 1 g 15N-labeled urea. The total N added was (4 × 0.46 g)
= 1.84 g pot-1. Harvesting was done on 8th January, 2008.

A second glasshouse experiment was planted using another 15 genotypes on 5th

November 2007 in four replications. They were:

1. WAB 878 SG 43 (WARDA No. 192)

2. WAB 748-11-2-HB (WARDA No. 57)
3. WAB 880-1-38-20-15-P2-HB (WARDA No. 113)
4. WAB 905-B-2A 1.1 (WARDA No. 26)
5. WAB 801-23-2-2-HB (WARDA No. 61)
6. WAB 788-58-2-1-HB (WARDA No. 56)
7. WAB 709-26-4-1-HB (WARDA No. 80)
8. Bukit Garam AN1334
9. Bukit Garam AN 753
10. Bukit Garam AN 582
11. Kinabatanagn AN1084
12. Bertih
13. Satang
14. Wangi
15. Lebar Dukong

The soil used in the glasshouse experiment was a Munchong series soil (Oxisol)
having a pH of 4.38, Bray-1P of 4.16 mg kg-1 and total N of 0.16%. The fertilizers applied
were the same as the first experiment for P and K, while N was doubled due to the first
experiment which did not perform well. The nitrogen applied was urea with 5 atom % 15N
excess. The plants were harvested at the end of March 2008 except for variety Wangi and
Lebar Dukong which were still at the booting stage. The parameters recorded for the
harvested plants were plant height, panicle length, panicle number and grain weight. All data
obtained were analyzed statistically using SAS package version 9. Tukey Studentized Range
test (HSD) was used for comparison of means.

289
2.2.2. Field experiment

This experiment was conducted on a Bungor series soil (Typic Kadiudult) at Ladang
Puchong, University Putra Malaysia. Sixteen soil samples were taken randomly in each block
using an auger, and were thoroughly mixed and sub-sampled. The subsamples was air-dried,
ground and sieved to pass through a 2.0 mm sieve size and kept for analysis. Plots 2.1 × 2.1m
were marked and arranged in a complete randomized block design, with 8 plots in each block.
Eight upland rice genotypes were planted on 13th March, 2008. For each plot, a subplot
measuring 1.5 × 1.5 m in the centre of the main plot was marked for the application of 15N-
labelled fertilizer. The remaining area was fertilizer with unlabelled ammonium sulphate .
Upland rice seeds were planted at a spacing of 30 x 30cm. First application of fertilizers was
made at 20 days after planting (60 kg−N ha-1,100 kg−P ha-1, 60 kg−K ha-1) and second
application was at 40 days after planting (60 kg−N ha-1, 60 kg−K ha-1).

The genotypes used for this field experiment were:

Nabawan, Tenom, WAB804-23-1-1-20HB (WARDA No. 20), WAB 880 SG37

(WARDA No. 99), Sintok, Pulut Petai, Merah dan Kuku Belang

Plant samples were taken at harvest. Parameters collected were total grain yield, plant
height, panicle number (no/clump),1000 grain weight (g), filled grains (%), growth duration
(days to maturity) and tiller number. The central six plants from each sub-plot were used for
N-15 analysis.

The chemical properties of the original soil used were: pH(water), 4.75; Bray-1
extractable P, 6.70 mg kg-1; total N, 0.18% ; organic C, 1.025%; exchangeable cations (cmol
(+) kg-1 soil), K, 0.16, Mg, 2.25, Ca, 0.65 and CEC, 6.7 cmol kg -1 soil.

2.3. Phosphorus uptake and P use efficiency

Ten genotypes, Tenom, Nabawan, Keningau, Kinabatangan, Kuku Belang, Merah,

Wangi, BG582, Pulut petai and BG1334A, were grown in 25 kg polybags in the glasshouse.
The soil was an Ultisol labelled with 100 µCi 32P containing 1 mg P as KH2PO4 prior to
planting of the seeds. P fertilizer at 50 kg P ha-1 was applied to the treated polybags. The
other half was not applied with P fertilizer. N and K was added as urea and muriate of potash
at 150 kg−N and 200 kg K ha-1, respectively, to all polybags at 30, 60 and 90 d after planting
in three split-applications. The response of these plants under glasshouse condition were not
satisfactory, thus another six genotypes were replanted in polybags with the same treatment,
but were placed under field conditions.

3. RESULTS

3.1. Root characteristics (cigar roll method)

Distinct differences were observed from these land races of upland rice. Principal
Component Analysis carried out on the data collected for all the 220 genotypes of upland rice
collected showed that the two components (root region, and root length) showing the highest
eigen values of >1.0 and accounted for 81.6% of the standardized variance. FACTOR retains
the two components based on the eigen values >1.0 rule. The first component (root region)
has large positive loadings for all variables except average root diameter. The second factor
(root length) only correlated well with root volume.

290
 Greater root length has been found to relate to the ability of roots to penetrate deep
into the soil to absorb nutrients such as nitrate which is easily leached into the soil, especially
in tropical areas which have heavy rainfall of >2000 mm rain per year. This “deep root”
characteristic would also help in the plants being able to withstand low rainfall (drought)
periods. The top 15 local landraces that showed good root length characteristics are shown in
Table 1.

TABLE 1. ROOT LENGTH AMONG LOCAL RICE LANDRACES GROWN UNDER UPLAND
CONDITIONS

No. Landrace Origin Root length (mm)

1 Pulut Petai Pahang 155 a
2 Bukit Garam 1249 Sabah 143 ab
3 Lebar Dukong Pahang 132 bc
4 Wangi Pahang 122 c
5 Kuku Belang Kelantan 121 c
6 Padi Burung Pahang 118 c
7 Satang Pahang 115 c
8 Tenom AN1214 Sabah 114 c
9 Merah Kelantan 113 c
10 Pulut Galah Pahang 110 c
11 Keningau AN763 Sabah 108 c
12 Bertih Pahang 103 c
13 Bukit Garam AN753 Sabah 97 c
14 Nabawan AN773 Sabah 97 c
15 Pungop Sarawak 97 c
Data within a column followed by the same lower case letter are not significantly different (P<0.05)

The top 15 lines obtained from WARDA that showed good root length characteristics
are shown in Table 2. It can be seen that the roots produced by WARDA lines were
significantly longer than those of the local landraces except for Pulut Petai and Bukit Garam
1249 (Table 1).

TABLE 2. ROOT LENGTH AMONG WARDA RICE LINES GROWN UNDER UPLAND
CONDITIONS

No. Line Number in WARDA List Root length (mm)

1 NERICA 14 150 206 a
2 WAB1094-B-58AB-1 100 197 ab
3 WAB326-B-B-19-H1-H1-HB 68 162 bc
4 WAB707-32-3-1-HB 74 156 cd
5 WAB716-27-2-1-HB 33 156 cd
6 CNA6675 180 149 d
7 WAB781-75-31-HB 18 144. d
8 WAB995-6A1.1 184 143 d
9 NERICA 5 134 140 d
10 WAB880-1-38-19-33-P1-HB 188 139 d
11 WAB 165 137 139 d
12 WAB340-B-B-2-H2 123 138 d
13 WAB880-1-38-18-20-P3-HB 198 138 d
14 WAB564-8-1-2 79 137 d
15 WAB503-18-4-1 179 136 d
Data within a column followed by the same lower case letter are not significantly different (P<0.05)

291
3.2. Nitrogen characteristics of upland rice

3.2.1. Glasshouse experiment 1

The average dry matter, N concentration and total N uptake together with the mean
15
N enrichment in the genotypes are shown in Table 3.

TABLE 3. TOTAL DRY WEIGHT, N CONCENTRATION, TOTAL N UPTAKE AND MEAN 15N
ENRICHMENT OF GENOTYPES
15
Genotypes Total DW %N Total N N (atom
(g pot-1) (mg pot-1) % excess)
WAB788-56-2-1-2-HB (17) 18.36 ab 2.83 531 ab 0.464
WAB767-2-5-1-HB (51) 10.56 bcd 2.89 309 abc 0.485
WAB915-B-3A1.1 (69) 3.69 d 2.78 103 c 0.404
NERICA 7 (143) 8.03 cd 3.14 257 bc 0.400
KENINGAU 749 4.88 d 2.85 131 c 0.465
BATU GARAM 1249 13.14 abcd 3.55 469 ab 0.421
NABAWAN 773 15.00 abc 3.51 504 ab 0.461
BECOR 20.85 a 2.54 521 ab 0.454
PULOT PETAI 20.41 a 2.76 565 a 0.471
SELAYANG 17.62 ab 2.86 506 ab 0.455
ns ns
Data within a column followed by the same lower case letter are not significantly different (P<0.05); ns, not
significant (P<0.05)

The highest dry matter produced was from Becor and Pulot Petai (2 local landraces).
No significant difference in the N concentration among the 10 lines was observed. Also the
15
N enrichment did not show any significant difference. The % and amount of N derived from
the fertilizer are given in Table 4.

TABLE 4. PERCENTAGE N DERIVED FROM FERTILIZER AND N TAKEN UP FROM

FERTILIZER

Genotypes NdfF (%) NdfF (mg pot-1)

WAB788-56-2-1-2-HB (17) 21.7 116 a
WAB767-2-5-1-HB (51) 22.7 94 ab
WAB915-B-3A1.1 (69) 18.9 34 b
NERICA 7 (143) 18.7 69 ab
KENINGAU 749 21.8 33 b
BATU GARAM 1249 19.7 93 ab
NABAWAN 773 21.6 121 a
BECOR 21.3 113 a
PULOT PETAI 22.0 122 a
SELAYANG 21.3 108a
ns
Data within a column followed by the same lower case letter are not significantly different (P<0.05); ns, not
significant (P<0.05)

292
 On average, 20% of the N in rice came from the applied fertilizer, with Pulot Petai
showing the highest percentage. This is expected since this landrace did show good root
length and good root branching habit.

3.2.2. Glasshouse experiment 2

The dry matter yield and N concentration of straw and grain and given in Table 5.

TABLE 5. DRY MATTER AND N CONCENTRATION OF STRAW AND GRAIN

Genotypes Straw DW Straw Grain DW Grain

(g pot-1) (% N) (g pot-1) (% N)
WAB 878 SG 43 42.93 cde 2.96 d 12.84 ab 3.78 ab
WAB 748-11-2-HB 52.93 cd 3.04 bcd 23.08 a 3.53 ab
WAB 880-1-38-20-15-P2-HB 28.70 de 3.27 abc 18.72 ab 4.05 ab
WAB 905-B-2A 1.1 23.90 e 3.46 a 15.07 ab 3.59 ab
WAB 801-23-2-2-HB 55.80 cd 3.05 bcd 17.29 ab 3.76 ab
WAB 788-58-2-1-HB 53.16 cd 2.70 d 24.23 a 3.29 b
WAB 709-26-4-1-HB 63.28 bc 3.10 bcd 15.03 ab 3.65 ab
Bukit Garam AN1334 38.88 cde 3.27 abc 11.27 ab 3.61 ab
Bukit Garam AN 753 51.33 cde 2.99 cd 15.96 ab 3.99 ab
Bukit Garam AN 582 35.02 de 3.47 a 11.68 ab 3.75 ab
Kinabatanagn AN1084 29.82 de 3.27 abc 12.26 ab 3.79 ab
Bertih 88.27 ab 2.84 d 14.41 ab 4.21 a
Satang 111.96 a 3.31 abc 7.04 b 3.73 ab
Data within a column followed by the same lower case letter are not significantly different (P<0.05)

Straw DW was highest in Satang while grain DW was highest in WAB 788-58-2-1-
HB. This experiment showed a higher N concentration due to higher rate of N applied. The
15
N enrichment of tissue, % N derived from the fertilizer and % fertilizer use efficiency are
given in Table 6.

TABLE 6. 15N ENRICHMENT OF TISSUE, PERCENTAGE N DERIVED FROM FERTILIZER

AND PERCENTAGE N USE EFFICIENCY

Genotypes Straw Grain Straw + grain

15
N (atom % excess) NdfF (%) NUE (%)
WAB 878 SG 43 2.473 2.193 ab 26.6 a 12.0
WAB 748-11-2-HB 2.475 2.073 ab 20.2 a 19.5
WAB 880-1-38-20-15-P2-HB 2.325 1.943 b 31.7 a 16.5
WAB 905-B-2A 1.1 2.320 2.260 ab 32.7 a 12.4
WAB 801-23-2-2-HB 2.523 2.245 ab 26.7 a 15.8
WAB 788-58-2-1-HB 2.435 2.243 ab 31.6 a 18.6
WAB 709-26-4-1-HB 2.103 2.278 ab 18.7 ab 11.9
Bukit Garam AN1334 2.115 2.078 ab 19.6 ab 9.3
Bukit Garam AN 753 2.313 2.395 a 26.7 a 12.4
Bukit Garam AN 582 2.000 2.185 ab 21.4 ab 9.8
Kinabatangan AN1084 2.225 2.330 ab 27.6 a 11.0
Bertih 2.053 2.075 ab 16.3 ab 14.5
Satang 2.383 2.208 ab 6.2 b 6.4
ns ns
Data within a column followed by the same lower case letter are not significantly different (P<0.05); ns, not
significant (P<0.05)

293
 The percentages of N derived from fertilizer were higher in this experiment, with up to
32% of the plant N from the fertilizer. Phenotypic differences among the genotypes are given
in Table 7.

TABLE 7. AGRONOMIC CHARACTERISTICS OF THE RICE PLANTS

Genotypes Plant Numbers of 1000 grain

height Tillers Panicles Spikelets Grains weight (g)
(cm) pot-1 panicle-1 panicle-1
WAB 878 SG 43 104.3 a 14.3 ab 17.0 ab 292.7 a 116.6 a 35.62 ab
WAB 748-11-2-HB 118.3 a 9.3 bc 13.3 ab 174.5 a 174.9 a 33.93 ab
WAB880-1-38-20-15-P2- 113.7 a 11.5 12.0 b 102.2 a 122.7 a 38.48 a
HB 100.3 a abc 9.8 b 218.4 a 128.1 a 24.63 abc
WAB 905-B-2A 1.1 115.0 a 5.3 bc 12.8 ab 425.8 a 114.4 a 10.20 c
WAB 801-23-2-2-HB 103.1 a 10.3 bc 21.0 a 113.5 a 198.5 a 27.12 abc
WAB 788-58-2-1-HB 118.9 a 19.8 a 15.0 ab 112.3 a 93.1 a 31.27 ab
WAB 709-26-4-1-HB 151.7 a 10.5 bc 12.5 ab 104.0 a 119.6 a 16.20 bc
Bukit Garam AN1334 151.0 a 9.3 bc 12.5 ab 124.0 a 44.2 a 25.45 abc
Bukit Garam AN 753 151.8 a 8.3 bc 10.3 b 254.1 a 142.2 a 23.15 abc
Bukit Garam AN 582 133.4 a 9.8 bc 13.0 ab 251.9 a 177.4 a 21.61 abc
Kinabatangan AN1084 134.9 a 9.3 bc 10.8 b 136.4 a 136.2 a 24.96 abc
Bertih 142.3 a 7.8 bc 8.3 b 361.0 a 159.4 a 29.03 abc
Satang 4.3 c
Data within a column followed by the same lower case letter are not significantly different (P<0.05)

The average height of the rice plants was between 1 m and 1.5 m. The highest number
of tillers produced was observed from the lines supplied by WARDA. The local landraces
were quite tall (between 1.3-1.5 m), but the tillers produced were not as many as the West
African lines. The number of panicles produced per plant varied among the lines, with the
African lines showing the highest number. The weight per1000 grains showed almost similar
values amongst the 13 lines tested. Correlations carried out between the agronomic
characteristics showed positive correlation between grain weight and %NdfF, with Pearson
correlation coefficient (r) of 0.607. Correlations on log of root length and all the agronomic
characteristics did not show any significant relationships.

3.2.3. Field experiment

The seed dry weight ranged from 204 g to 816 g plot -1 and the highest seed dry weight
was obtained from WARDA99 to 816 g plot-1, the lowest seed dry weight was obtained from
Merah with 204 g plot-1 (Table 8). The highest plant dry weight was obtained from Merah
with 2735 g plot-1 while the lowest plant dry matter yield was obtained from WARDA20 with
794 g plot-1. The plant height ranged from 130.5 to 194.5 cm. The highest plant height was
obtained from Merah with 194.5 cm, with WARDA20 being the shortest at 130.5 cm. The
length of panicle varied from 23.6 to 29.3 cm. Pulut Petai had the longest panicle with 29.3
cm, while Nabawan had the shortest at 23.6 cm. Seed lengths (1.0 cm) were not significantly
different in all genotypes tested.

294
TABLE 8. RICE PLANT CHARACTERISTICS AT HARVEST

Genotypes Grain DW DMY Panicle length No. tillers hill-1 Plant height
(g plot-1)† (g plot-1)† (cm) (cm)
Nabawan 518 bc 949 de 23.6 e 9.5 ab 155.5 cd
Tenom 665 ab 863 de 24.8 de 6.4 b 139.6 de
WARDA20 729 ab 794 e 27.7 ab 7.8 b 130.5 e
WARDA99 816 a 1199 cd 27.8 ab 6.9 b 144.9 de
Sintok 727 ab 1492 c 25.7 cd 7.4 b 166.0 bc
Pulut Petai 806 a 2271b 29.3 a 7.1 b 190.3 a
Merah 204 d 2735 a 26.4 bc 11.8 a 194.5 a
Kuku Belang 371 cd 2170 b 28.7 a 8.1 b 178.5 ab
†DW, dry weight; DMY, dry matter yield; plot size was 2.25 m2; Data within a column followed by the same
lower case letter are not significantly different (P<0.05)

There was no significant difference (P<0.05) in N concentration in the plant between

the upland rice genotypes used (Table 9). N concentration ranged between 2.59 to 3.05%.
The P concentration ranged from 0.31 to 0.43%. Kuku Belang showed the highest P
concentration of 0.43%. WR20 showed significantly higher K concentration (3.76%) than
WR99, Tenom, Nabawan, Sintok, Merah, Kuku Belang and Pulut Petai, while Kuku Belang
and Pulut Petai showed similar K concentrations. Ca concentration in plants ranged from 443
to 653 mg kg-1 and Mg concentration ranged between 270 to 480 mg kg-1. The Cu
concentration ranged from 98 to 106 mg kg-1; Fe ranged from 255 to 379 mg kg-1; Mn from
804 to 946 mg kg-1; and Zn from155 to 186 mg kg-1 (Table 9).

TABLE 9. NUTRIENT CONCENTRATIONS IN RICE PLANTS

Concentration (%) Nutrient concentration (mg kg-1)

Variety
N P K Ca Mg Cu Fe Mn Zn
0.37
Nabawan 2.96 a 2.93 bc 443 b 372 bc 101 a 378 a 905 a 168 abc
abc
Tenom 2.65 a 0.31 c 2.94 bc 489 ab 300 cd 103 a 365 a 930 a 166 abc
WARDA20 2.99 a 0.35 bc 3.76 a 626 a 480 a 106 a 365 a 935 a 168 abc
WARDA99 3.02 a 0.33 c 3.12 b 493 ab 343 bcd 101 a 379 a 946 a 167 abc
Sintok 3.05 a 0.33 c 2.84 bc 550 ab 376 bc 98 a 261 a 826 a 155 c
Pulut Petai 2.59 a 0.35 bc 2.41 c 574 ab 298 cd 97 a 255 a 918 a 161 bc
Merah 2.90 a 0.41 ab 2.67 bc 633 a 388 b 100 a 285 a 804 a 182 ab
Kuku
2.83 a 0.43 a 2.47 c 595 ab 270 d 103 a 308 a 936 a 186 a
Belang
Data within a column followed by the same lower case letter are not significantly different (P<0.05)

The 15N enrichment in plants ranged between 0.629 to 0.753 atom % excess (Table
10), while the 15N enrichment in upland rice seeds ranged from 0.553 to 0.757 atom % excess.
The total NdfF in upland rice ranged from 61.0 to 121.0 kg ha-1 (Table 10). Merah showed the
highest NdfF with 121.0 kg ha-1 and the lowest was obtained from variety Tenom with 61.0
kg ha-1. Variety Merah showed the highest nitrogen use efficiency in upland rice with 80.7%
and the lowest nitrogen use efficiency in upland rice was obtained from variety Tenom with
40.7%.

295
TABLE 10. 15N ENRICHMENT, AMOUNT OF NITROGEN DERIVED FROM FERTILIZER AND
PERCENTAGE NITROGEN USE EFFICIENCY OF UPLAND RICE
15
Variety N (atom % excess) N derived from fertilizer (kg ha-1) % N use
Plant Grain Plant Grain Total efficiency
Nabawan 0.671 a 0.553 a 40.2 de 23.0 bc 63.2 bc 42.1 bc
Tenom 0.721 a 0.614 a 31.3 e 29.7 abc 61.0 c 40.7 c
WRDA 20 0.629 a 0.684 a 27.7 e 41.5 ab 69.2 bc 46.2 bc
WRDA 99 0.731 a 0.696 a 48.3 cde 51.8 a 100.1 abc 66.7 abc
Sintok 0.711 a 0.654 a 64.0 bcd 38.8 ab 102.7 ab 68.5 ab
Pulut Petai 0.770 a 0.758 a 70.6 bc 43.0 ab 113.6 a 75.7 a
Merah 0.729 a 0.714 a 109.3 a 11.7 c 121.0 a 80.7 a
Kuku Belang 0.753 a 0.685 a 78.6 b 21.1 bc 99.7 abc 66.4 abc
Data within a column followed by the same lower case letter are not significantly different (P<0.05)

3.3. P uptake efficiency of upland rice

Out of the 10 genotypes used, only 5 genotypes germinated and survived. The five
were Tenom, Nabawan, Keningau, Kinabatangan, and BG1334A. Most of the fertilizer P
was found in the grain (Table 11) and between 3-8% of the applied P fertilizer was utilized by
the rice plants (Table 12).

TABLE 11. DRY MATTER YIELD, P UPTAKE AND PERCENTAGE OF P DERIVED FROM
FERTILIZER IN STRAW AND GRAIN OF RICE PLANTS

Genotypes Dry matter (g pot-1) P (mg pot-1) PdfF (%

Straw Grain In straw In grain In straw In grain
Tenom 36.04 17.15 42.4 a 47.24 a 0.00 b 39.40 b
Nabawan 24.26 15.48 23.2 c 43.03 a 9.90 b 72.69 a
Keningau 40.97 23.12 37.6 b 45.05 a 0.00 b 22.08 c
Kinabatangan 40.02 19.52 46.8 a 44.00 a 22.64 a 40.98 b
BG1334A 29.95 10.07 39.3 ab 24.24 b 35.51 a 47.47 b
ns ns
Data within a column followed by the same lower case letter are not significantly different (P<0.05), ns, not
significant (P<0.05)

TABLE 12. AMOUNT OF P DERIVED FROM FERTILIZER IN STRAW AND GRAIN OF RICE
AND PERCENTAGE OF FERTILIZER P UTILIZED BY THE RICE PLANTS

Genotypes PdfF (mg pot-1) P use efficiency

In straw In grain Total (%)
Tenom 0.0 b 18.6 b 18.6 b 7.7 ab
Nabawan 2.4 b 31.3 a 33.7 a 8.0 a
Keningau 0.0 b 10.0 c 10.0 c 2.4 b
Kinabatangan 10.6 a 18.0 b 28.6 a 6.8 a
BG1334A 14.0 a 11.5 bc 25.5 a 6.1 a
Data within a column followed by the same lower case letter are not significantly different (P<0.05)

No significant difference in straw yield between genotypes were measured (Table 13).
WARDA 100 gave the highest seed yield of 13.28 g polybag -1 (Table 13). Most of the P
taken up was found in the straw. The highest percentage and amount of P derived from
fertilizer was obtained from grains of WARDA100 (Table 14). No difference in percentage P
use efficiency was observed.

296
TABLE 13. DRY MATTER YIELD, P UPTAKE AND PERCENTAGE OF P DERIVED FROM
FERTILIZER IN STRAW AND GRAIN OF RICE PLANTS GROWN UNDER FIELD
CONDITIONS

Genotypes Dry matter (g polybag-1) P (mg pot-1) PdfF (%)

Straw Grain Straw Grain Straw Grain
BG1334A 63.94 a 9.10 b 60.4 a 12.1 bc 31.4 ab 17.3 b
WR100 51.66 a 13.28 a 47.7 ab 21.6 a 42.6 ab 51.9 ab
BG582 55.61 a 7.46 b 50.2 ab 11.6 bc 31.5 ab 30.4 ab
NB773 57.90 a 8.34 b 49.6 ab 12.9 bc 42.1 ab 34.1 ab
KB1084 55.93 a 9.55 b 46.3 ab 15.0 b 40.9 ab 42.1 ab
KN749 48.37 a 8.15 b 41.7 ab 12.3 bc 21.5 b 30.7 ab
TE1211 57.88 a 9.63 b 45.6 ab 12.7 bc 42.3 ab 49.8 ab
BG753 53.83 a 7.72 b 42.6 ab 8.9 c 51.8 ab 30.3 ab
BG1249 54.66 a 8.30 b 33.7 b 10.6 bc 60.6 a 58.7 a
Data within a column followed by the same lower case letter are not significantly different (P<0.05)

TABLE 14. AMOUNT OF P DERIVED FROM FERTILIZER IN STRAW AND GRAIN AND P
USE EFFICIENCY OF RICE PLANTS GRPOWN UNDER FIELD CONDITIONS

Genotypes PdfF (mg pot-1) P use efficiency

Straw Grain Total (%)
BG1334A 17.2 a 2.4 b 20.4 a 2.45 a
WR100 19.5 a 11.4 a 30.9 a 3.71 a
BG582 173 a 3.6 b 20.9 a 2.50 a
NB773 23.8 a 4.4 b 29.0 a 3.49 a
KB1084 19.3 a 6.8 ab 26.2 a 3.14 a
KN749 12.1 a 5.0 b 17.1 a 2.05 a
TE1211 26.1 a 7.1 ab 33.2 a 4.00 a
BG753 28.3 a 2.0 b 30.8 a 3.70 a
BG1249 28.6 a 5.7 b 34.3 a 4.12 a
Data within a column followed by the same lower case letter are not significantly different (P<0.05)

4. DISCUSSION

Plant root systems are highly plastic in their development and can adapt their
architecture in response to prevailing environmental conditions [10, 11]. In Arabidopsis, it
has been shown that uniformly high nitrate (10 mM) suppresses lateral development, while
plants grown at low levels of nitrate (10 µM), and when a section of the primary root was
exposed to high nitrate levels, lateral root production was stimulated specifically in that area.
The main effect of nitrate appears to be on the rate of lateral root elongation rather than on
lateral root initiation, while the elongation rate of the primary root is identical on 10 µM and
10 mM nitrate, and the metabolism of nitrate is apparently not necessary for the architectural
changes [10]. Since metabolism of nitrate is not required for the root architectural changes,
the differences observed in the root length of the upland rice landraces shown must be due to
their genetic differences.

The N content in grain and straw obtained in our studies were similar to those
obtained in upland rice varieties in Brazil [12]. Harvest indices (HI) (grain yield / total
biomass yield) were also similar to the traditional upland rice variety from Northern Laos
[13]. Fertilizer nitrogen use efficiency for these varieties tested ranged between 40 and 80%,
which is very high, considering that efficiency value between 23-30% was reported for upland
rice when 300 kg−N ha-1 was applied [14].

297
 It has been mentioned that rooting depth is one of the root characteristics that
determines the ability of a crop to intercept N, particularly NO3 - during periods of leaching
[15]. Since upland rice is grown under aerobic conditions, most of the N will be in the form
of NO3-, and is liable to be leached under heavy rainfall conditions of the tropics. Thus
varieties capable of producing long seminal roots will be potentially capable of producing
deep roots. In this study, we did not find any positive relationships between root
characteristics taken at two weeks of age with all the agronomic characteristics of the plants
grown in the greenhouse or in the field.

5. CONCLUSION

WARDA lines showed significantly longer root length compared to local upland land
races at 2 weeks old using the cigar-roll method, but no relationships were found between
these characteristics and the agronomic characteristics of the plants when they were grown in
the greenhouse or field conditions. Nitrogen fertilizer use was high in the lines tested, while P
use was similar to other crops.

REFERENCES

[1] INTERNATIONAL RICE RESEARCH INSTITUTE, IRRI Program Report, Los

Baños, Philippines (1998) 54–69.
[2] HEDLEY, M.J., KIRK, G.J.D., SANTOS, M.B., Phosphorus efficiency and the forms
of soil phosphorus utilized by upland rice cultivars, Plant Soil 158 (1994) 53–62.
[3] LAFITTE, R.H., COURTOIS, B., ARRAUDEAU, M., Genetic improvement of rice in
aerobic systems: progress from yield to genes, Field Crops Res. 75 (2002) 171–190.
[4] GEORGE, T., et al., Upland rice response to phosphorus fertilization in Asia, Agron. J.
93 (2001) 1362–1370.
[5] BARRACLOUGH, P.B., The growth and activity of winter wheat roots in the field –
Root growth of high yielding crops in relation to shoot growth, J. Agric. Sci. 103
(1984) 439–442.
[6] LYNCH, J.P., Root architecture and plant productivity, Plant Physiol. 109 (1995) 7–
13.
[7] HOAD, S.P., et al., The management of wheat, barley and oat root systems, Adv.
Agron. 74 (2001) 193–246.
[8] BHADRARAY, S., et al., Phosphorus mobilization in hybrid rice rhizophere
compared to high yielding varieties under integrated nutrient management, Biol. Fertil.
Soils 35 (2002) 73–78.
[9] HILL, J.O., et al., Morphology and responses of roots of pasture species to phosphorus
and nitrogen nutrition, Plant Soil 286 (2006) 7–19.
[10] ZHANG, H.M., FORDE, B.G., Regulation of Arabidopsis root development by nitrate
availability, J. Exp. Bot. 51 (2000) 51–59.
[11] ZHANG, H., et al., Dual pathways for regulation of root branching by nitrate, Proc.
Nat. Acad. Sci. USA 96 (1999) 6529–6534.
[12] ARF, O., et al., Soil management and nitrogen fertilization for sprinkler-irrigated
upland rice cultivars, Scientia Agricola 60 (2003) 345–352.
[13] SAITO, K, et al., Response of traditional and improved upland rice cultivars to N and
P fertilizer in northern Laos, Field Crop Res. 96 (2006) 216–223.
[14] WANG, Y., et al., Effects of nitrogen fertilization on upland rice based on pot
experiments, Commun. Soil Sci. Plant Anal. 39 (2008) 1733–1749.
[15] GASTAL, F., LEMAIRE, G., N uptake and distribution in crops: An agronomical and
ecophysiological perspective, J. Exp. Bot. 53 (2002) 789–799.

298
 BIOLOGICAL NITROGEN FIXATION EFFICIENCY IN BRAZILIAN
COMMON BEAN GENOTYPES AS MEASURED BY 15N METHODOLOGY

V.I. FRANZINI, F.L. MENDES

Brazilian Agricultural Research Corporation,
EMBRAPA-Amazonia Oriental,
Belém, PA,

T. MURAOKA, A.R. TREVISAM

Center for Nuclear Energy in Agriculture,
University of São Paulo,
Piracicaba, SP,

Brazil

J.J. ADU-GYAMFI
Soil and Water Management & Crop Nutrition Laboratory,
International Atomic Energy Agency,
Seibersdorf,
Austria

Abstract

Common bean (Phaseolus vulgaris L.) represents the main source of protein for the Brazilian
and other Latin-American populations. Unlike soybean, which is very efficient in fixing atmospheric
N2 symbiotically, common bean does not dispense with the need for N fertilizer application, as the
biologically fixed N (BNF) seems incapable to supplement the total N required by the crop. A
experiment under controlled conditions was conducted in Piracicaba, Brazil, to assess N2 fixation of
25 genotypes of common bean (Phaseolus vulgaris L.). BNF was measured by 15N isotope dilution
using a non-N2 fixing bean genotype as a reference crop. The common bean genotypes were grown in
low (2.2 mg N kg-1 soil) or high N content soil (200 mg N kg-1 soil), through N fertilizer application,
as urea-15N (31.20 and 1.4 atom % 15N, respectively). The bean seeds were inoculated with Rhizobium
tropici CIAT 899 strain and the plants were harvested at grain maturity stage. The contribution of
BNF was on average 75% of total plant N content, and there were differences in N fixing capacity
among the bean genotypes. The most efficient genotypes were Horizonte, Roxo 90, Grafite, Aporé
and Vereda, when grown in high N soil. None of the genotypes grown in low N soil was efficient in
producing grains compared to those grown in high N soil, and therefore the BNF was not able to
supply the total N demand of the bean crop.

1. INTRODUCTION

Common bean (Phaseolus vulgaris L.) is the main source of protein for the Brazilian
population. Nitrogen (N) is the nutrient taken up in larger amounts by the bean plant, and N
supplied as fertilizer is expensive and easily lost by volatilization or leaching [1, 2].
Approximately 50% of total N uptake is exported in the grain and the remainder stays in the
soil in the form of crop residues [1, 3].

Maximizing the use of N by bean is important because of the economic and

environmental aspects, as this nutrient presents risk to the environment as potentially
contaminating groundwater [4] due to leaching of nitrate. Moreover, it is observed in field
that it is possible to achieve bean yields above 2500 kg ha-1 according to the BNF process
without N addition [5, 6]. However, inoculants are used in only 2–3% of the acreage of
common bean [7]. As the ability for nodulation varies with the common bean genotypes and

299
Rhizobium strain, the nodulation efficiency is highly dependent on the genotype of common
bean [8].

The objective of this study was to compare the common bean’s ability for effective
BNF, evaluated using the 15N dilution technique, for 25 common bean genotypes grown under
controlled conditions with no limitation to BNF (light, temperature, water, nutrients, and
efficient rhizobium strain association with common bean).

2. MATERIAL AND METHODS

2.1. Experimental

The experiment was conducted in the greenhouse at the Center for Nuclear Energy in
Agriculture (CENA / USP), located at latitude 22°42'30'' S, longitude 47°38'01'' W and 554 m
altitude, in Piracicaba, Sao Paulo, Brazil.

The study were performed in 3.0 l plastic pots, containing 2.5 kg of air-dried soil,
collected from the 0 to 0.20 m layer of a dystrophic Typic Haplustox [9]. The soil had 280, 70
and 650 g kg-1content of clay, silt and sand, respectively, and the following chemical
characteristics: pH (0.01 mol l-1 CaCl2), 4.5; organic matter, 18.0 g dm-3 ; P extracted by resin,
5 mg dm-3 ; K, 0.6 mmolc dm-3 ; Ca, 11.5 mmolc dm-3; Mg, 5.2 mmolc dm-3; H + Al, 35.4
mmolc dm-3; CEC, 52.7 mmolc dm-3; sum of bases, 17.3 mmolc dm-3; base saturation, 32.8%,
according to methodology described by [10] and P by Mehlich-1, 3 mg dm-3 [11].

After application of lime (calcium carbonate equivalent = 110%) to raise the base
saturation to 70% for the common bean, according to the official recommendation of the
Bulletin 100 [12], the soil was incubated for 30 days and the moisture content was maintained
at approximately 70% of water holding capacity.

The experimental design was a randomized complete block with four replications. The
experiment consisted of 52 treatments, which were divided into two groups. Each group had
26 treatments, which consisted of 25 common bean genotypes (Aporé, BRS Grafite, BRS
Horizonte, BRS Pitanga, BRS Vereda, BRSMG Pioneiro, Carioca 80, CNF 10, FT Nobre,
IAC Tybatã, IAC UNA, Jalo precoce, LP 01-38, Ônix, Ouro Negro, Pontal, Rosinha, Roxão
EEP, Roxinho, Roxo 90, Rubi, Rudá, Sangue de Boi, Thayú e Timbó) and a control plant
(non-nodulating common bean).

The first treatment (group of 25 common bean genotypes and one standard plant
species) was carried out to compare the BNF between genotypes in the absence of mineral-N.
As the 15N isotopic dilution method requires the labeling of soil N, a low rate of mineral N
(2.20 mg kg-1 soil as 15N-urea with high enrichment of 31.20 atom % 15N was used. The
second treatment was to compare the BNF by the same common bean genotypes used in the
first group in the presence of mineral-N. N was applied at 200 mg kg -1 soil as 15N-urea (1.41
atom % 15N). To avoid inhibiting the nodulation of plants, the mineral N dose was split in two
applications, 1/3 at sowing and the remaining portion at the V4 stage (issuance of the third
trifoliate leaf).

The P and K fertilization was performed with application of 200 mg kg -1 soil of P as

triple superphosphate and 200 mg kg-1 soil of K as potassium sulfate in all pots. Micronutrient
fertilization was performed by application of a nutrient solution in all treatments at rates of
0.5 mg kg-1 of B, 1.5 mg kg-1 of Cu, 3.0 mg kg-1 of Fe, 2.0 mg kg-1 of Mn, 3.0 mg kg-1 of Zn
and 0.1 mg kg-1 of Mo.

300
 The experiment was conducted in summer under conditions of high intensity of light
and temperature, in order not to limit the BNF. The seeds of the bean genotypes were
inoculated with the strain of Rhizobium tropici CIAT 899 (SEMIA 4077), which is
recommended commercially for common bean in Brazil. The inoculation was done three
hours before sowing, at a rate of 500 grams of inoculants (109 viable cells g -1 peat) to 50 kg
of seed plus 300 ml of 10% sugar solution (w: v) to improve its adherence to the seeds.

Five common bean seeds from each genotype and standard plant (non-nodulating bean
genotype - NORH 54) were sown in each experimental, thinned to one plant per pot. Soil
moisture was maintained at approximately 70% of water retention capacity during the
experiment. Plants were harvested after the pod maturity stage and separated into roots, shoots
(stems, leaves and bark of legumes) and grain. At harvest, the number of nodules per plant
and fresh weight of nodules per plant were measured.

Plant samples were dried and ground in Wiley-type mill, passed through 10 mesh
sieve and weighed on a precision analytical balance (five decimal) for 15N isotope (atom %)
and total N determination in a mass spectrometer (IRMS) interfaced with an elemental
analyzer, according to methodology described by [13].

2.1. N calculations

N uptake was calculated according to the following equation:

TN = N x Sdm
Where:
TN = plant total N content (mg plant-1);
N = plant N concentration (g kg-1);
Sdm = shoot dry mater (g plant-1).

The equation of symbiotic fixed N is presented below [14].

 (%A 15Nexcess ) genotype 

%Npdfix = 1 -  15 
 (%A Nexcess ) control 

Where:

% Npdfix: % N in the plant derived from the symbiotic fixation;

(%A 15N excess)genotype: atom % 15N excess in the nodulated common bean genotype;
(%A 15N excess)control: atom % 15N excess in the control plant (non-nodulating bean);

With the total N content in the SDM of common bean genotypes (TN, mg plant -1), the
amount of N in the plant derived from symbiotic fixation (QNpdfix) was calculated.

(%Npdfix) . (NA)genotype .
QNpdfix =
100
Where:

QNpdfix: amount of N in the plant derived from symbiotic fixation (mg);

% Npdfix: % of N in the plant derived from symbiotic fixation;
(TN)genotype: total N content in the Sdm of the common bean genotype (mg).

301
2.2. Statistical analysis

Cluster analysis of common bean genotypes was carried out with the SAS 9.1 -
“Statistical Analysis System” [15] and SYSTAT version 10.2 software programs, using the
UPGMA (unweighted pair group arithmetic average clustering). The cluster analysis was
preceded by the standardization of data before the Euclidian distances calculation, as the
studied variables presented different scales. After standardization all the variables were
equally important in the determination of these distances. Final results of the groups were
presented as dendrograms. Within the text, the symbol * preceding a genotype denotes the
treatment fertilized with the higher rate of urea.

3. RESULTS AND DISCUSSION

With high urea application, the number of nodules = 0; shoot DM = 54.24 g plant -1; N
uptake in shoots = 307 mg plant-1 ; N derived from fertilizer in shoots = 69.0%; N derived
from BNF in shoots = 0%; N derived from soil in shoots = 31.0%; Recovery of urea N in
shoots = 42.4%; Root DM = 15.76 g plant -1 ; Grain DM = 3.83 g plant-1; N uptake in grain =
141 mg plant-1 ; number of pods = 20.5; number of grains = 40.3; N derived from fertilizer in
grain = 69.4%; N derived from BNF in grain = 0%; N derived from soil in grain = 30.6%;
Recovery of urea-N in grain = 19.7%; total shoot DM = Shoot DM + Grain DM = 58.07 g
plant-1 ; N uptake by total shoot DM = 449 mg plant-1 ; N derived from BNF in total shoot DM
= 0%; N derived from fertilizer in total shoot DM = 69.1%; Recovery of urea-N in total shoot
DM = 62.1%; and N derived from soil in total shoot DM = 30.9%.

With low urea application, the number of nodules = 0; shoot DM = 12.40 g plant -1; N
uptake in shoots = 80.5 mg plant-1 ; N derived from fertilizer in shoots = 2.5%; N derived from
BNF in shoots = 0%; N derived from soil in shoots = 97.5%; Recovery of urea N in shoots =
36.4%; Root DM = 5.37 g plant-1 ; Grain DM = 0.96 g plant-1 ; N uptake in grain = 31.6 mg
plant-1 ; number of pods = 9.0; number of grains = 16,5; N derived from fertilizer in grain =
1.6%; N derived from BNF in grain = 0%; N derived from soil in grain = 98.4%; Recovery of
urea-N in grain = 9.4%; total shoot DM = Shoot DM + Grain DM = 13.36 g plant -1; N uptake
by total shoot DM = 112 mg plant-1 ; N derived from BNF in total shoot DM = 0%; N derived
from fertilizer in total shoot DM = 0.5%; Recovery of urea-N in total shoot DM = 45.9%; and
N derived from soil in total shoot DM = 99.5%.

The N uptake from BNF in total shoot DM of 25 common bean genotypes at low or
high application of urea, correlated significantly and positively with shoot DM (0.656 ***),
grain DM (0.699 ***) and root DM (0.493 ***). This increase in production of DM of shoots,
roots and grain indicates that plants responded both in grain production as well as shoots and
roots biomass to N supplied by the BNF. Considering these three variables, the level of
homogeneity of 1.0 in the Euclidean distance, we observed the formation of the following
seven distinct homogeneous groups as shown in Fig. 1.

1st: Jalo Precoce, Grafite, Roxão EEP and Rubi;

2nd: *Roxão EEP and *Jalo Precoce;
3rd: Sangue de Boi, CNF 10, Ônix, Pontal, Roxo 90, Horizonte, Thayú, Pitanga, IAC
UNA, FT Nobre, Aporé, LP 01-38, Ouro Negro, Pioneiro, Rudá, Carioca 80, Rosinha,
Tybatã, Roxinho and Vereda;
4th: Timbó;
5th: *Carioca 80, *FT Nobre, *Roxo 90, *Ouro Negro, *Rosinha, *Pontal, *Pioneiro,
*CNF 10, *Rudá, *Rubi, *Pitanga, *Aporé, *Horizonte, *Grafite, *Sangue de Boi,
*LP01-38, *Ônix and *Tybatã;

302
6th: Vereda, *Roxinho, *IAC UNA and *Thayú; and
7th: *Timbó.
*Timb ó
* Th ayú
*IAC UNA
* Ro x inh o
*Vereda
* Ty batã
* Ôn ix
*LP0 1-3 8
*Sang u e B oi
*Grafite
* Ho rizon te
* Ap oré
*Pitang a
* Ru b i
*R ud á
* CNF 10
*Pion eiro
*Pon tal
*R osinh a
* Ou ro Negro
*R ox o 90
*FT No bre
* Cario ca 80
Timb ó
Vereda
Ro xin h o
Ty batã
R osinh a
C ario ca 8 0
R ud á
Pion eiro
Ou ro Negro
LP0 1-3 8
Apo ré
FT Nob re
IAC UNA
Pitanga
Thayú
Horizo nte
R ox o 9 0
Pon tal
Ôn ix
CNF 10
San gu eB oi
*R ox ão EEP
*Jalo Precoce
Ru bi
R ox ão EEP
Grafite
Jalo p recoce

0.0 0.5 1.0 1.5 2.0 2.5

Euclidean distance

FIG. 1. Dendrogram resulting from hierarchical cluster analysis of 25 genotypes of common bean,
based on DM (g) of shoots, roots and grain and N in total shoot DM derived from BNF. * denotes
genotypes that received high urea fertilization.

There was a response in grain production, and shoot and root DM to BNF, even in
genotypes fertilized with the higher rate of urea, and this productive response varied among
genotypes (Fig. 1). The genotypes, in association with rhizobium, which most benefited from
BNF and produced more DM of shoots roots and grain were Vereda, Roxinho, IAC UNA and
Thayú, in the treatment with higher rate of N fertilization (6th group). The BNF contributed
around 70% of total N in these genotypes, even with the higher addition of urea-N; it shows
that, depending on the bean genotype, BNF can contribute with high amounts of N2 fixation,
even in N fertilized plants. Moreover, considering the seven groups of genotypes, there was
no genotype in the treatments with low and high fertilizer N classified in the same group.
Therefore, the genotypes that depended only on soil N and BNF as N sources for development
were not as efficient in terms of shoot, root and grain DM production, as those receiving a
higher rate of N fertilizer. On average, considering the 25 bean genotypes, fertilized with low
or high urea, BNF contributed approximately 75% of total N absorbed by the plants.

303
 The number of plant nodules correlated positively with the fresh weight of nodules
(0.704 ***), the total N content in shoot DM (0.367***) and with the total plant N derived
from BNF (0.350***). Although significant, the correlation coefficients of the number and
fresh weight of nodules with total plant N and N uptake in the plant from BNF were relatively
low. The number and fresh weight of nodules had coefficients of variation of 18.2 and 20.2%,
respectively (Table 1).

TABLE 1. NODULE FRESH WEIGHT AND NUMBER, N UPTAKE OF SHOOTS AND N

DERIVED FROM BNF OF 25 GENOTYPES OF COMMON BEAN AT LOW (+) OR HIGH (++)
UREA APPLICATION

Genotype Nodule fresh weight Nodule number N uptake (mg)

(mg plant-1) Total in shoots Derived from BNF
+ urea ++ urea + urea ++ urea + urea ++ urea + urea ++ urea
Carioca 80 0.59 0.52 162 173 968 1412 855 1048
Rudá 1.03 0.66 171 107 955 1271 840 859
Aporé 1.34 2.22 139 227 853 1405 755 914
Pontal 1.82 1.28 195 199 855 1358 729 885
Horizonte 1.83 3.09 191 242 844 1403 720 907
Pioneiro 0.76 0.95 114 130 909 1401 792 889
Rosinha 0.84 1.05 155 212 954 1358 854 909
Rubi 0.95 1.62 136 163 680 1321 565 828
Vereda 1.51 2.11 173 248 1158 1560 1028 1129
Tybatã 1.40 1.53 158 155 944 1456 837 936
CNF 10 1.04 0.81 157 145 894 1276 771 742
Roxão EEP 1.03 0.97 141 113 571 1005 446 539
Sangue de Boi 0.84 0.73 126 115 700 1332 584 838
Roxinho 1.25 0.86 135 135 1096 1628 967 1097
Timbó 0.66 0.68 108 102 1048 1387 935 969
Roxo 90 1.49 2.02 147 301 837 1399 723 906
Pitanga 1.47 1.43 158 149 1050 1369 928 925
Ouro Negro 1.53 1.30 171 149 969 1424 838 905
Ônix 1.41 1.08 142 137 851 1415 736 944
IAC UNA 0.80 1.06 153 168 1045 1628 888 1159
Jalo precoce 0.77 1.19 133 171 404 1238 256 678
Grafite 1.21 1.93 139 246 540 1251 433 834
Thayú 1.27 1.72 157 268 1128 1821 1006 1351
FT Nobre 1.36 1.14 124 145 930 1532 772 1015
LP01-38 1.45 1.56 154 157 873 1273 740 741
Average 1.26 162 1140 839
CV (%) 20.2 18.2 12.7 13.6

Considering these four variables, the level of homogeneity in the Euclidean distance of
1.0 (Fig. 2), seven homogeneous groups were obtained:

1st: Jalo Precoce, Grafite, Roxão EEP, Rubi, Sangue de Boi and *Roxão EEP;
2nd: Horizonte and Pontal;
3rd: Pitanga, Tybatã, Ouro Negro, LP01-38, Roxo 90, Ônix, Aporé, FT Nobre, CNF 10,
Rudá, IAC UNA, Rosinha, Carioca 80, *CNF 10, *Jalo Precoce, *Rosinha, *Pontal,
*LP01-38, *Rubi, Roxinho, Thayú, Vereda, *Tybatã, *Pitanga, *Ouro Negro, *FT
Nobre, *Ônix, *Pioneiro, *Timbó, *Rudá, *Sangue de Boi, Timbó and Pioneiro;
4th: *Carioca 80, *Roxinho and *IAC UNA;
5th: *Thayú;
6th: *Roxo 90, *Grafite, *Aporé and *Vereda; and

304
7th: *Horizonte.

Among these groups, the 6th and 7th groups, formed by the genotypes *Horizonte,
*Roxo 90, *Grafite, *Aporé and *Vereda stood out, especially on nodulation and N
accumulation (Table 1). These five common bean genotypes with highest nodulation were
grown with the high N rate.
*Rosinha
*Pioneiro
*FT Nobre
*Aporé
*Ônix
*Rubi
*Vereda
*Tybatã
*Thayú
*Sangue Boi
*Roxinho
*Roxo 90
*Pontal
*Ouro Negro
*IAC UNA
*Carioca 80
*Rudá
*Pitanga
*CNF 10
Rosinha
*Horizonte
Pioneiro
Rudá
Carioca 80
*LP01-38
Ônix
Tybatã
SangueBoi
FT Nobre
LP01-38
CNF 10
Roxinho
IAC UNA
Aporé
*Grafite
Vereda
Ouro Negro
Thayú
Pitanga
*Timbó
Timbó
*Roxão EEP
*JaloPrecoce
Roxo 90
Pontal
Horizonte
Rubi
Roxão EEP
Grafite
Jalo precoce

0.0 0.5 1.0 1.5 2.0

Euclidean distance

FIG. 2. Dendrogram resulting from hierarchical cluster analysis of 25 genotypes of common bean
based on number and fresh weight of nodules, accumulation of N (mg) in total shoot dry matter and N
in total shoot DM derived from BNF. * denotes genotypes that received high urea fertilization.

Considering the edible portion and economic interest of the crop, it was observed that
the grain DM of common bean genotypes, fertilized with a low or high rate of urea-N,
correlated significantly and positively with the number of pods (0.824***) and the number of
grains (0.878***). The number of pods and seeds per plant were highly and positively
correlated with grain production. The number of pods is one of the most important in
increasing the production of common bean [16].

305
1st: (low productive group) formed by treatment Jalo Precoce;
2nd: (moderately productive group) by treatments Grafite, Roxão EEP, Rubi, Horizonte,
Pontal, Roxo 90 and *Jalo Precoce;
3rd: (moderately to very productive group) by treatments *Roxão EEP, Timbó, *Timbó,
Pitanga, Thayú, Ouro Negro, Vereda, *Grafite, Aporé, IAC UNA, Roxinho, CNF 10,
LP01-38, FT Nobre, Sangue de Boi, Tybatã, Ônix, *LP01-38, Carioca 80, Rudá,
Pioneiro, *Horizonte, Rosinha, *CNF10, *Pitanga and *Rudá; and the
4th: (very productive group) by the treatments *Carioca 80, *IAC UNA, *Ouro Negro,
*Pontal, *Roxo 90, *Roxinho, *Sangue de Boi, *Thayú, *Tybatã, *Vereda, *Rubi,
*Onix, *Aporé, *FT Nobre, *Pioneiro and *Rosinha.

From grain DM, number of pods and number of grains (Table 2), in the level of
homogeneity of 1.0 in the Euclidean distance dendrogram (Fig. 3), we observed the formation
of four homogeneous groups:

TABLE 2. MEAN DRY MATTER YIELD OF SHOOTS, ROOTS AND GRAIN, NUMBER OF
PODS AND GRAINS OF 25 GENOTYPES OF COMMON BEAN AT LOW (+) OR HIGH (++)
UREA APPLICATION

Genotype Dry matter yield (g plant-1) Number plant-1

Shoot Root Grain Pods Grain
+ urea ++ urea + urea ++ urea + urea ++ urea + urea ++ urea + urea ++ urea
Carioca 80 20.0 29.4 7.3 9.6 28.3 32.7 20.0 24.3 108.0 122.3
Rudá 18.7 26.1 6.5 8.3 27.8 29.4 19.0 23.3 115.7 150.3
Aporé 17.6 35.5 5.4 11.1 24.5 28.1 19.0 27.7 95.7 172.7
Pontal 17.0 35.3 7.8 9.5 20.1 33.7 11.7 24.0 74.7 160.7
Horizonte 20.0 38.9 6.0 9.2 17.3 25.5 12.0 21.3 66.7 132.3
Pioneiro 18.8 34.7 5.3 8.8 28.4 35.2 20.0 31.3 121.3 174.3
Rosinha 20.8 34.5 6.6 11.3 29.3 35.9 22.7 32.3 127.7 189.0
Rubi 14.6 32.8 4.8 9.5 15.0 29.7 11.0 28.7 59.0 158.7
Vereda 23.7 42.8 10.6 16.4 24.8 30.1 18.0 27.0 105.7 160.0
Tybatã 21.2 45.2 9.3 12.4 22.4 29.5 17.7 26.3 104.0 154.7
CNF 10 16.8 26.3 7.2 8.0 23.4 28.9 16.0 22.7 94.3 131.7
Roxão EEP 13.2 27.2 5.3 10.8 13.3 21.0 11.7 26.7 50.7 117.7
Sangue de Boi 14.5 30.6 7.1 14.1 20.7 31.0 17.3 30.0 103.3 161.0
Roxinho 23.2 36.3 10.5 15.3 23.7 32.2 16.7 30.3 99.7 172.0
Timbó 31.9 43.8 9.3 21.2 15.7 19.8 22.3 21.7 81.7 100.3
Roxo 90 19.8 33.5 5.6 13.2 19.2 33.2 14.3 27.0 72.7 161.7
Pitanga 21.3 30.7 5.6 9.6 22.6 27.6 21.7 24.7 106.3 136.0
Ouro Negro 18.9 30.1 3.8 11.9 26.2 35.5 17.0 26.7 87.7 135.0
Ônix 19.6 41.9 6.9 12.2 21.8 28.8 13.3 28.7 94.3 159.3
IAC UNA 21.4 35.6 6.1 12.6 24.0 32.3 17.3 24.7 92.0 138.7
Jalo precoce 10.3 33.8 5.3 16.4 7.0 18.2 5.3 15.7 21.3 56.0
Grafite 14.8 35.4 8.1 14.6 10.4 24.9 8.7 18.3 48.0 102.0
Thayú 21.7 38.5 6.9 14.4 24.9 30.5 21.7 27.0 107.3 146.3
FT Nobre 19.6 33.4 5.4 12.7 21.9 33.5 15.7 29.3 93.0 187.3
LP01-38 17.9 30.9 4.7 11.9 24.5 29.2 15.3 18.7 93.0 99.0
Average 26.8 9.4 25.5 20.9 116.1
CV (%) 11.9 15.2 12.9 13.7 14.5

306
 *Horizonte
*Vereda
*Aporé
*Grafite
*Roxo 90
*Thayú
*IAC UNA
*Roxinho
*Carioca 80
Pioneiro
Timbó
*Sangue Boi
*Rudá
*Timbó
*Pioneiro
*Ônix
*FT Nobre
*Ouro Negro
*Pitanga
*Tybatã
Vereda
Thayú
Roxinho
*Rubi
*LP01-38
*Pontal
*Rosinha
*JaloPrecoce
*CNF 10
Carioca 80
Rosinha
IAC UNA
Rudá
CNF 10
FT Nobre
Aporé
Ônix
Roxo 90
LP01-38
Ouro Negro
Tybatã
Pitanga
Pontal
Horizonte
*Roxão EEP
SangueBoi
Rubi
Roxão EEP
Grafite
Jalo precoce

0.0 0.5 1.0 1.5 2.0 2.5

Euclidean distance

FIG. 3. Dendrogram resulting from hierarchical cluster analysis of 25 genotypes of common bean
based on grain dry matter, number of pods and number of grains. * denotes genotypes that received
high urea fertilization.

Among the 25 common bean genotypes in the treatment with low N fertilizer, none
were classified in the group of very grain productive. This indicates that BNF was not able to
meet all the N demands of the plant N to achieve equivalent grain yields of the same
genotypes that received urea-N fertilization. Common bean is considered to be a species with
a low capacity for nodulation and BNF compared to other grain legumes [17]. Nevertheless,
the BNF has contributed up to 90 kg−N ha-1 in various bean crops, which represented 40 to
50% of the demand of this crop [18]; in seven field experiments the observed average and
maximum values were, respectively, 35 and 70% of N in the plant from the atmosphere in
common bean genotypes [19].

The efficiency of bean BNF depends, among other factors, on the genotype. The
selection of genotypes more efficient in symbiosis with rhizobia is an alternative to reduce N
fertilization. For example, in field studies, fertilization with 20 kg−N ha-1, together with
inoculant strain of R. tropici CIAT 899 permitted common bean to yield higher than 3000 kg
ha-1, equivalent to the application of 160 kg−N ha-1 [6]. In addition, one must consider that the
success of inoculation with strains of bean rhizobia with high efficiency is associated with

307
competitive ability of such strains and adaptation to environmental conditions [20]. Under
appropriate environmental conditions, the atmospheric N2 fixed by symbiosis can meet most
of the N needs of common bean [21]. However, some soil conditions such as low pH and high
concentrations of Al often limit all stages of root infection, nodule formation and assimilation
of N by the plant [22].

Studies under field conditions have shown that it is possible to achieve bean yields
above 2500 kg ha-1 by the BNF process without addition of mineral N fertilizer [5, 6]. It is
noteworthy, however, that the energy used in BNF is ATP; the photosynthates are important
to the process of BNF for the N2-fixing organisms, because they generate reducing power and
ATP for the nitrogenase system, are substrates for growth and maintenance of microbial cells
and supply carbon skeletons, ATP and reducing power for the assimilation of NH3 [23].
Therefore, environmental conditions or management practices that influence the availability
of photosynthates also affect BNF. The bean growing season (winter/summer), for example,
influences the availability of photosynthates, and thus the BNF.

4. CONCLUSIONS

− Common bean genotypes differ in their ability to fix N2 from the atmosphere
through BNF.

− Among 25 common genotypes, Horizonte, Roxo 90, Grafite, Aporé and Vereda
were the most efficient in BNF, when grown in the presence of urea fertilizer;

− Under controlled conditions (no limitation of temperature, water and nutrients), only
with BNF and soil as sources of N for plants (without fertilization with urea),
common bean genotypes cannot produce grain, shoot or root dry matter equal to those
fertilized with urea.

ACKNOWLEDGMENTS

The work was supported by IAEA (International Atomic Energy Agency) - research
contract 13779. V. Franzini acknowledges a graduate fellowship from CAPES (Coordenação
de Aperfeiçoamento de Pessoal de Nível Superior, Brazil).

REFERENCES

[1] DE OLIVEIRA, I.P.., ARAUJO, R.S., DUTRA, L.G., “Nutrição mineral e fixação
biológica de nitrogênio”, Cultura do Feijoeiro Comum no Brasil, (Araujo, R.S., et al.,
Eds), Potafos, Piracicaba (1996) 169–221.
[2] CANTARELLA, H., “Nitrogênio”, Fertilidade do Solo, (NOVAIS, R.F., et al., Eds),
Sociedade Brasileira de Ciência do Solo, Viçosa (2007) 375–470.
[3] FAGERIA, N.K.,SANT'ANA, E.P., MORAIS, O.P., Maximização da eficiencia de
produção das culturas.Brasilia: EMBRAPA-SCT/EMBRAPA-CNPAF, (1999) 294 pp.
[4] DOS SANTOS, A.B.et al., Resposta do feijoeiro ao manejo de nitrogenio em varzeas
tropicias. Pesq. Agropec. Bras. 38 (2003) 1265-1271.
[5] HUNGRIA, M., et al., Isolation and characterization of new efficient and competitive
bean (Phaseolus vulgaris L.) rhizobia from Brazil, Soil Biol. Biochem. 32 (2000)
1515–1528.
[6] PELEGRIN, R., et al., Resposta da cultura do feijoeiro à adubação nitrogenada e à
inoculação com rizóbio, Rev. Bras. Ciência Solo 33 (2009) 219–226.

308
[7] LOPES, E.S., “Fixação biológica do nitrogênio no sistema solo-planta”, Simpósio
Sobre Nitrogênio e Enxofre na Agricultura Brasileira, Piracicaba (YAMADA, T.,
ABDALLA, S.R.S., VITTI, G.C., Eds), International Plant Nutrition Institute, 2 (2007)
43–72.
[8] FRANCO, M.C., et al., Nodulação em cultivares de feijão dos conjuntos gênicos
andino e meso-americano, Pesq. Agropecuária Brasil. 37 (2002) 1145–1150.
[9] DOS Santos, H.G., et al., Sistema Brasileiro de Classificação de Solos, 2nd Edn,
EMBRAPA-Solos, Rio de Janeiro (2006) 306 p.
[10] VAN RAIJ, B., et al., Análise Química para Avaliação da Fertilidade de Solos
Tropicais, Instituto Agronômico, Campinas (2001) 285 p.
[11] EMBRAPA-SOLOS, Manual de Métodos de Análises de Solos, 2nd Edn, Rio de
Janeiro, (1997) 212 p.
[12] VAN RAIJ, B., et al., Recomendação de adubação e calagem para o Estado de São
Paulo, 2nd Edn, Instituto Agronômico e Fundação IAC, Campinas (Boletim Técnico,
100) (1997) 285 p.
[13] BARRIE, A., PROSSER, S.J., “Automated analysis of light-element stable isotopes by
isotope ratio mass spectrometry”, Mass Spectrometry of Soils (BOUTTON, T.W.,
YAMASAKI, S., Eds), Marcel Dekker, New York (1996) 1–46.
[14] MCAULIFFE, C., et al., Influence of inorganic nitrogen on nitrogen fixation by
legumes as revealed by 15N, Agron. J. 50 (1958) 334–337.
[15] SAS Institute, SAS User’s Guide: Statistics, vs. 8.2., SAS Institute, Cary (2001).
[16] FAGERIA, N.K., et al., “Nutrição de fósforo na produção de feijoeiro”, Simpósio
Sobre Fósforo na Agricultura Brasileira, Associação Brasileira para Pesquisa da
Potassa e do Fosfato, Piracicaba 17 (2004) 435–455.
[17] GRAHAM, P.H., et al., Addressing edaphic constraints to bean production: The bean /
cowpea CRSP project in perspective, Field Crops Res. 82 (2003) 179–192.
[18] WESTERMANN, D.T., et al., Nitrogen sources for bean seed production, Agron. J. 73
(1981) 660–664.
[19] HARDARSON, G., et al., Genotypic variation in biological nitrogen fixation by
common bean, Plant Soil 152 (1993) 59–70.
[20] STRALIOTTO, R., TEIXEIRA, M.G., MERCANTE, F.M., “Fixação biológica de
nitrogênio”, Produção de Feijoeiro Comum em Várzeas Tropicais, (AIDAR, H.,
KLUTHCOUSKI, J., STONE, L.F., Eds), EMBRAPA Arroz e Feijão, Santo Antônio
de Goiás (2002) 122–153.
[21] HUNGRIA, M., NEVES, M.C.P., VICTORIA, R.L., Assimilação do nitrogênio pelo
feijoeiro; II. Absorção e translocação do N mineral e do N2 fixado, Rev. Bras. Ciência
Solo 9 (1985) 202–220.
[22] MARTÍNEZ-ROMERO, E., et al., Rhizobium tropici, a novel species nodulating
Phaseolus vulgaris L. beans and Leucaena sp. trees, Int. J. System. Bacteriol. 41 (1991)
417–426.
[23] MOREIRA, F.M.S., SIQUEIRA, J.O., Microbiologia e Bioquímica do Solo, Editora
UFLA, Lavras (2006) 729 p.

309
 GENOTYPIC VARIATION OF EARLY MATURING SOYBEAN
GENOTYPES FOR PHOSPHORUS UTILIZATION EFFICIENCY UNDER
FIELD GROWN CONDITIONS

R.C. ABAIDOO
Kwame Nkrumah University of Technology,
Kumasi,
Ghana

and

International Institute of Tropical Agriculture,

Ibadan,
Nigeria

A. OPOKU, S. BOAHEN
Kwame Nkrumah University of Technology,
Kumasi,
Ghana

M.O. DARE
Federal University of Agriculture,
Abeokuta,
Nigeria

Abstract

Variability in the utilization of phosphorus (P) by 64 early-maturing soybean (Glycine max L.

Merr.) genotypes under low-P soil conditions were evaluated in 2009 and 2010 at Shika, Nigeria.
Fifteen phenotypic variables; number of nodules, nodule dry weight, grain yield, plant biomass, total
biomass, biomass N and P content, Phosphorus Utilization Index (PUI), shoot P Utilization efficiency
(PUIS), grain P Utilization efficiency (PUIG), Harvest Index (HI), Biological N fixed (BNF), total N
fixed and N and P uptake were measured. The four clusters revealed by cluster analysis were basically
divided along (1) plant biomass and uptake, (2) nutrient acquisition and utilization and (3) nodulation
components. Three early maturing genotypes, TGx1842-14E, TGx1912-11F and TGx1913-5F, were
identified as having high P utilization index and low P uptake. These genotypes could be a potential
source for breeding for P use efficiency in early maturing soybean genotypes.

1. INTRODUCTION

Soybean (Glycine max L. Merr.) is an important source of protein to many households

in West Africa. It is an emerging grain legume in the farming systems in the West African
savanna region, and is considered to contribute to the N economy of the soil through
biological N2 fixation (BNF) with indigenous rhizobia populations. It has been stated that
more than 25% of its total shoot N content can come from BNF [1]. Critical to the optimal
growth of soybean and BNF is the adequate supply of P to the crop, but P is one of the most
limiting nutrients in the West African savanna region because of the kaolinitic soil properties
of the area. This problem results in reduced plant growth and final grain yield [2]. It has been
observed that soybean response to low P condition varies with different genotypes [3], and
some mechanisms have been attributed to the differential response exhibited by various
genotypes. In alleviating the problem of the low P condition in the soil, maximum potential of

311
adaptation of the plants with a larger relative capacity in the absorption and use efficiency of
the P will be a good alternative.

Most breeding programmes for P efficiency have had their major focus on P
acquisition efficiency. Most crops have relatively efficient P uptake capacity but low P
translocation and remobilization, and hence PUE becomes a significant bottleneck for further
improvements in crop P efficiency [4]. It is therefore imperative that research programmes are
initiated and implemented with the objective of identifying P utilization efficient genotypes
within existing gemplasm.

Because of the problem associated with P availability in the soils of the tropical region
and the high cost of P fertilizers, it is pertinent to develop varieties that are early maturing and
are able to efficiently use P even under low P soil conditions. Screening of large accessions of
soybean for P use efficiency is one of the activities of the IITA in Nigeria. The screening can
help to identify various genotypes that have better P use efficiency and use them in the
breeding program. The objective of this study was to evaluate the genetic variation of 65 early
maturing soybean genotypes for P utilization efficiency.

2. MATERIALS AND METHODS

2.1. Location

Field experiments were conducted at Shika (11o 13’ N, 7o 12’E) in 2009 and 2010.
Shika is located in the Northern Guinea savanna and has a unimodal rainfall pattern with
about 1100 mm rainfall per annum. The soil type is a ferric lixisol [5]. Fields were cleared,
ploughed, and ridges spaced 75 cm apart were constructed with a tractor-mounted ridge. The
plot size was 2.25 m2 (three rows per plot, a 1.0 m alley between plots, 1.5 m between
replications). The plots were supplied with P at the rate of 20 kg P ha -1 TSP which was
applied by broadcasting, and incorporated into the soil with hand hoes before planting of the
64 soybean genotypes (Table 1) two days after in both years. The experimental design was
Randomized Complete Block (RCB). Each soybean genotype treatment was replicated three
times. The soybean genotypes were obtained from the germplasm collection of IITA, Nigeria.

2.2. Planting

Soybean was sown along the ridges and thinned to two plants per hole at a spacing of
12.5 cm with 0.75 cm between rows, 2 weeks after planting (WAP). The plots were weeded
manually using hoes at 3, 6 and 10 WAP. Sampling was carried out at 50% flowering for
BNF, at pod-filling for P-use efficiency, and at harvest maturity for grain yield. At the first
sampling (50% flowering), four soybean plants were carefully dug out from the 0.5 m portion
of the two central rows. To minimize damage to the root system, the soil around the plant was
loosened using forks. Plant samples were processed for number and dry weight of nodules,
and measurement of N2-fixation using the hot water extraction method [6–8]. The ground
stem and petiole of the shoot was passed through 1 mm sieve, after which the sample was
extracted with 25 ml boiling water for 2 min. The extract was filtered, made up to volume (50
ml), and stored at –15oC to be analysed later for ureides and % N derived from the
atmosphere (% Ndfa).

Harvested plant samples were chopped into 10- to 20-mm pieces and sub-sampled
with about 200 g fresh weight being oven-dried at 70 C before grinding to pass through a 0.5
mm sieve for measurement of N and P concentrations. For the last harvest, plant samples were
separated into reproductive (grains) and vegetative parts (shoots) after oven drying. The

312
grains were threshed from the pods and further dried. Plant shoots, nodules and straw were
dried in the oven at 70 ºC to a constant weight before weighing. Total N in the grains, shoots
and straw was determined by the Kjeldahl procedure [9]. Phosphorus concentration in tissues
was determined using the procedures described by [10].

2.3. P utilization efficiency

Phosphorus Utilization Index (PUI) was determined as the inverse of nutrient (P)
concentration in the biomass. Grain P utilization efficiency (PUIG) was determined as
expressed by [11, 12), while shoot P utilization efficiency (PUIS) were determined as
expressed by [13].

PUIG = Grain yield x PUI

PUIS = Harvest index x PUI

2.4. Statistical analysis

Combined analyses of variance were conducted for eleven variables measured in 2009
and 2010 using the Generalized Linear Model (GLM) Procedure of the statistical analysis
system (14). The cluster analysis was performed by the CLUSTER analysis of the GENSTAT
package using the standardized data of the complete linkage method, which explained a
greater proportion of the variation in the set of data than other methods. Data were subjected
to Genotype × Trait (GT) analysis [14, 15] to identify cultivars that were superior with respect
to selected traits. The data were not transformed (‘Transform = 0’), standardized (Scale = 1)
and were trait centered (Centering = 2). The trait values were standard deviation-standardized
because traits were measured in different units. The polygon and the vector views of the GT
biplot were constructed using all measured traits, and were based on genotype-focused
singular partitioning (‘SVP = 2), which rendered them appropriate for visualizing the
relationship among traits. The GT biplot analysis was carried out using GGE Biplot, a
Windows application that fully automates biplot analysis [15, www.ggebiplot.com). The GGE
bipplot model 2 equation used is

(Ŷij – µ – βj)/ dj = λ1gi1e1j + λ2gi2e2j + εij

where:

Yij is the genetic value of the combination between inbred i and trait j;
µ is the mean of all combinations involving trait j;
βj is the main effect of trait j;
λ1and λ2 are the singular values for PC1 and PC2;
gi1 and gi2 are the PC1 and PC2 eigenvectors, respectively, for inbred i;
e1j and e2j are the PC1 and PC2 eigenvectors, respectively, for trait j;
dj is the phenotypic standard deviation; and
εij is the residual of the model associated with the combination of inbred i and trait j.

313
TABLE 1. THE 64 GENOTYPES OF SOYBEAN EVALUATED AT SHIKA
BETWEEN 2009 AND 2010

Genotype Genotype
TGx1019-2EN TGx1911-9F
TGx1485-1D TGx1912-11F
TGx1740-2F TGx1912-12F
TGx1805-8F TGx1912-1F
TGx1830-20E TGx1912-2F
TGx1834-1E TGx1912-3F
TGx1835-10E TGx1912-6F
TGx1842-14E TGx1912-9F
TGx1871-12E TGx1913-5F
TGx1876-4E TGx1914-11F
TGx1878-12E TGx1914-17F
TGx1880-3F TGx1914-2F
TGx1888-29F TGx1914-4F
TGx1892-10F TGx1917-1F
TGx1893-10F TGx1918-1F
TGx1895-19F TGx1918-2F
TGx1895-22F TGx1918-3F
TGx1895-23F TGx1918-5F
TGx1895-33F TGx1920-1F
TGx1895-50F TGx1921-2F
TGx1895-6F TGx1921-6F
TGx1902-1F TGx1921-7F
TGx1903-11F TGx1922-1F
TGx1903-13F TGx1923-4F
TGx1903-1F TGx1904-4F
TGx1903-2F TGx1904-5F
TGx1903-3F TGx1904-6F
TGx1903-4F TGx1908-6F
TGx1903-7F TGx1909-2F
TGx1903-8F TGx1909-3F
TGx1911-7F TGx1911-2F
TGx1911-8F TGx1911-3F

3. RESULTS

Significant (P<0.05) genotypic variation was observed in grain yield, HI, total
biomass at harvest, PUIG and PUIS (Table 2). Plant biomass, biomass P concentration, P
uptake and PUI were, however, not significantly affected by the genotype. Grain yield ranged
between 377 kg ha-1 in TGx 1985-19F and 1798 kg ha-1 in TGx 1880-3F. The PUI, PUIG,
PUIS and HI ranged between 3.54 in TGx 1895-33 and 5.97 in TGx 1913-5F; 1561.35 in TGx
1895-19F and 8197.32 in TGx 1880-3F: 0.89 in TGx 1888-29F and 2.01 in TGx 1913-5F and
0.20 in TGx 1871-12E and 0.42 in TGx 1922-1F, respectively (Table 2). TGx 1880-3F was
significantly (P<0.05) higher in grain yield and PUIG than 75% of other genotypes.

314
TABLE 2. SUMMARY OF THE STATISTICAL ANALYSIS OF BIOMASS AND GRAIN YIELD,
P ACCUMULATION AND USE EFFICIENCY OF 64 GENOTYPES OF SOYBEANS
EVALUATED AT SHIKA BETWEEN 2007 AND 2008

Statistics Plant Grain Total HI P P uptake PUI PUIG PUIS

biomass yield biomass conc. (kg ha-1 )
(kg ha-1 ) (kg ha-1 ) (kg ha-1 ) (%)
Mean 5300.1 1001.7 6301.7 0.31 0.24 13.0 4.35 4327.1 1.38
Minimum 3303.7 377.7 3987.1 0.20 0.19 7.40 3.54 1561.4 0.89
Maximum 7816.1 1797.9 8736.1 0.42 0.29 21.7 5.97 8197.3 2.01
SE 2022 247.6 2075 0.1 0.04 6.4 0.74 1282 0.39
F statistics for Genotype (G) ns *** * *** ns ns ns *** ***
F statistics for Year (Y) *** *** *** *** ns *** ns *** ***
F statistics for G × Y ns *** ns *** ns ns ns ns ns
Symbols *, **, *** denote P<0.05, P<0.01, P< 0.001, respectively; ns, not significant

The soybean genotypes did not significantly (P<0.05) vary in % Ndfa (Table 3).
However, number of nodules, nodule dry weight and total N fixed were significantly (P<
0.05) influenced by the genotype. The %Ndfa ranged from 51.4% in TGx 1918-2F to 62.2%
in TGx 1805-8F (Table 3). TGx 1740-2F had significantly (P<0.05) higher number of nodules
than other genotypes except TGx 1911-9F, 1911-3F and 1903-3F. TGx 1911-9F, 1917-1F,
1921-2F and 1912-9F were the top four genotypes for total N fixed and N uptake among the
64 genotypes (Table 3).

TABLE 3. SUMMARY OF THE STATISTICAL ANALYSIS OF NODULATION, N

ACCUMULATION AND FIXATION OF THE 64 GENOTYPES OF SOYBEAN EVALUATED AT
SHIKA BETWEEN 2009 AND 2010

Statistics No. of Nodule Biomas %Ndfa Total N N uptake

nodules dry weight N conc fixed (kg ha-1)
(×107) (kg ha-1) (%) (kg ha-1)
Mean 1.24 97.0 2.82 55.9 82.7 144.5
Minimum 0.38 33.7 2.16 51.4 48.0 88.2
Maximum 3.09 158.9 3.34 62.2 135.5 224.1
SE 0.44 33.4 0.32 3.6 31.7 54.3
F statistics for Genotype (G) *** *** ** ns * *
F statistics for Year (Y) ns *** *** *** *** ***
F statistics for G × Y *** *** ns ns ns ns
Symbols *, **, *** denote P<0.05, P<0.01, P< 0.001, respectively; ns, not significant

The G × T biplot with a polygon view was used to identify genotypes that were
superior with respect to some traits. The genotypes at each vertex of the polygon (vertex
cultivar) possessed the highest values for traits found within its sector. The principal
components PC1 and PC2 accounted for 58.8% of the total variation among the measured
traits of the genotypes (Fig. 1). In the biplot view, TGx1918-5F was the vertex genotype in
the sector that contained traits HI, grain yield, PUI, PUIS, PUIG and BNF, indicating that this
genotype had the highest values for these traits among the 64 genotypes (Fig. 1). The
genotype, TGx 1921-2F had the highest values for the number of nodules, total biomass and
plant biomass, while TGx 1912-9F was the vertex genotypes for the sector that had nodule
dry weight, N uptake and total N fixed. The plot showed that TGX1903-2F was the best
cultivar for biomass P content and TGX 1892-10F for biomass N content (Fig. 1).

315
FIG. 1. Genotype × trait biplot of all traits of 64 early maturing soybean genoypes evaluated in
Nigeria between 2009 and 2010. PUI = Phosphorus Utilization Index; PUIG = Grain P utilization
efficiency; PUIS = shoot P utilization efficiency.

The interrelationships among the measured traits were displayed in a vector view
biplot (Fig. 2), the rays connecting the traits to the biplot origin are referred to as trait vectors
and the cosine of the angle between the vectors of any two traits measures the similarity
between them. However, traits with shorter vectors are considered not strongly correlated
with those with longer vectors and also were probably not strongly correlated with other short
vector traits [14]. Therefore, even though BNF had acute angle (<90°) with HI, PUIS, PUI,
PUIG and grain yield, its short vector suggested that it had weak positive correlation with
those traits (HI, PUIS, PUI, PUIG and grain yield). The vector lengths of number of nodule
and nodule dry weight were also relatively shorter than those of total biomass, plant biomass,
total N fixed, N and P uptake and grain yield which had acute angles with it. Biomass N and P
contents had longer vectors and were at angles >90° to PUI, PUIS, PUIG, HI and grain yield,
indicating that they were negatively correlated with PUI and grain yield. Grain yield, PUIG
and PUIS had acute angles between them with longer vectors, indicating they were positively
correlated. Similarly, total biomass, plant biomass, N and P uptake and total N fixed were also
positively correlated, except that PUIG, PUI and PUIS had angles approximately 90° with
total N fixed, indicating a non significant correlation between PUI and total N fixed. PUI and
biomass P content had an angle that is approximately 1800 between their vectors, indicating a
high negative correlation between the two variables. Similar high negative correlations were
observed between the following trait pairs; Grain yield and biomass N concentration, PUIG
and biomass P concentration, and PUIS and biomass P concentration. In general, the pattern
of relationships among these traits is such that it is easy to classify them into three groups;
Total biomass, plant biomass, N uptake, P uptake, and total N fixed all in one group; Biomass
N concentration and biomass P concentration in the second group; and grain yield, PUI, PUIS

316
and PUIG in the third trait group. Other traits could be grouped into any other of the three
groups because of the relatively short vectors, indicating that they are unique in characterizing
the soybean genotypes.

FIG. 2. A vector view of the genotype × trait biplot showing interrelationship among all traits of the
64 early maturing genotypes evaluated in Nigeria between 2009 and 2010. PUI = Phosphorus
Utilization Index; PUIG = Grain P utilization efficiency; PUIS = shoot P utilization efficiency.

Clustering of the soybean genotypes based on the eleven variables produced four
major clusters at a reversed distance of 0.8 which showed a remarkable correspondence with
the PCA axes (Fig. 3). Cluster one had nine genotypes and were loaded with genotypes that
had high plant biomass N and P uptake and total N fixed (e.g. TGx1911-9F, TGx 1914-2F
and TGx 1921-2F). Cluster two had 25 genotypes that included genotypes TGx 1893-10F and
1895-33F, which were high in biomass N and P content. Cluster three was composed of three
genotypes, TGX1842-14E, TGX1912-11F and TGX1913-5F, which generally had high P
utilization efficiency but low P uptake and N fixation. TGx1912-11F and TGx1913-5F had
the highest P utilization index among the 64 genotypes. Cluster four comprised 27 genotypes.
The grain yield, BNF and nodule dry weight were high for most of the genotypes in this
cluster. Among the genotypes found in this cluster were TGX1880-3F, TGX1903-7E and
TGX1917-1F, TGX1911-3F and TGX 1805-8F.

317
TGX1019-2EN
TGX1903-11F
TGX1912-1F
TGX1912-6F
TGX1880-3F
TGX1895-23F
TGX1909-3F
TGX1903-8F
TGX1911-7F
TGX1904-4F
TGX1911-2F
TGX1485-1D
TGX1805-8F
TGX1740-2F
TGX1835-10E
TGX1918-1F
TGX1903-3F
TGX1912-2F
TGX1871-12E
TGX1902-1F
TGX1920-1F
TGX1918-2F
TGX1921-7F
TGX1903-1F
TGX1914-17F
TGX1903-7F
TGX1912-12F
TGX1842-14E
TGX1912-11F
TGX1913-5F
TGX1830-20E
TGX1834-1E
TGX1876-4E
TGX1895-22F
TGX1918-5F
TGX1908-6F
TGX1911-8F
TGX1914-11F
TGX1895-50F
TGX1918-3F
TGX1904-5F
TGX1893-10F
TGX1895-33F
TGX1878-12E
TGX1909-2F
TGX1903-13F
TGX1912-3F
TGX1903-4F
TGX1904-6F
TGX1922-1F
TGX1914-4F
TGX1892-10F
TGX1895-19F
TGX1895-6F
TGX1921-6F
TGX1888-29F
TGX1903-2F
TGX1911-3F
TGX1911-9F
TGX1912-9F
TGX1914-2F
TGX1917-1F
TGX1921-2F
TGX1923-4F

1.0 0.9 0.8 0.7 0.6 0.5

FIG. 3. Dendrogram of 64 early maturing soybean genotypes derived from quantitative traits using
first four principal components and complete linkage cluster analysis.

4. DISCUSSION

The significant positive and negative correlation of grain yield with PUI and biomass
N concentration, respectively, is an indication that P concentration of the plant may not
increase the grain yield of soybean, but how efficiently the P accumulated in the biomass is

318
utilized. There had been a contention on whether P acquisition or utilization should be the
main focus of P use efficiency in breeding programmes. Breeding P efficient genotypes has
focussed more on P acquisition efficiency (PAE) [17]. The result of the present study is an
indication that PAE may not only be the best solution but also the P utilization efficiency. It
has been speculated from soybean studies that enhancement of P utilization efficiency might
become a potentially powerful strategy for increasing P efficiency in modern crops grown in
intensive cropping systems [5].

Although the P Utilization Index of the 64 early maturing genotypes was not
significantly affected by genotype, the P utilization for grain development measured as PUIG,
and for harvest index measured as PUIS, was significantly influenced by genotype. This was
probably due to the significant genotypic variation for grain production which could have
influenced the P sink during grain formation. Therefore breeding of genotypes for grain
production on low P soil can be enhanced by selecting those genotypes with high PUIG and
PUIS. Genotypes identified with high PUIG and PUIS in this study included TGx 1880-3F,
1913-5F and 1485-1D. Screening for the P utilization efficient early maturing soybean
genotypes in this study as revealed by the GT biplot produced a result that was probably more
desirable to farmers in low P soil. The sector that had the trait PUI, PUIG and PUIS also had
the grain yield and HI. The indication for this was that the genotypes that fell into this sector
were the best for the traits in the sector. Therefore, selection of a genotype within this sector
for further breeding programmes can produce high P utilization efficient and high grain yield
genotypes for low P soil.

The cluster analysis was consistent with the GT biplot analysis. Like the GT biplot,
the four clusters can be grouped along three major components which were biomass
accumulation, nutrient accumulation and nodulation components. Cluster 1 was defined more
by high plant biomass, N and P uptake and total N fixed. Cluster 2 was defined more by the
high nutrient concentrations (N and P) in the biomass whereas cluster 3 was defined by their
high PUI and low P uptake. Cluster 4 was defined by the nodulation variables and grain yield.
Phosphorus utilization efficiency (PUE) is the ability to produce biomass or yield using the
acquired P [17].

Most crops have relatively efficient P uptake capacity but low P translocation and
remobilization, and hence PUE becomes a significant bottleneck for further improvements in
crop P efficiency [5]. The genotypes in cluster 2 displayed a higher propensity to accumulate
N and P. However it was the genotypes in cluster 3 that were able to efficiently use P better
than other genotypes. The interesting finding of this study was that just three genotypes,
TGX1842-14E, TGX1912-11F and TGX1913-5F were selected for PUI which was used as a
measure for P utilization efficiency. It means that the three genotypes could be good
candidates for further breeding work.

It was expected that PUI will influence BNF because P is a component of ATP which
provides the energy that helps the symbiont to fix N in soil. Nitrogenase makes use of ATP
for the reduction of N2. However, there was no significant relationship between the PUI and
BNF. The probable explanation could be that competition from the plant for energy for
biomass accumulation and grain filling exceeded that of the symbiont that fixes N in the root
nodules of these genotypes. Therefore, more of the P in the genotypes was utilized for
biomass formation.

319
 REFERENCES

[1] VASILAS, B.I., et al., Relationship of nitrogen utilization patterns with soybean yield
and seed-fill period, Crop Sci. 35 (1995) 809–813.
[2] GAUME, A., Low Phosphorus Tolerant of Various Maize Cultivars: the
Contribution of the Root Exudation, Dissertation for the degree of Doctor of Natural
Sciences, Swiss Federal Institute of Technology, Zurich, Switzerland (2000)
[3] ABDELGADIR, A.H., The Role of Mycorrhizae in Soybean Growth in P deficient
Soil in the Humid Tropics, Ph. D thesis, Cornell University, Ithaca, NY, USA (1998)
255 p.
[4] WANG, X., et al., Acquisition or utilization, which is more critical for enhancing
phosphorus efficiency in modern crops? Plant Sci. 179 (2010) 302–306.
[5] FOOD AND AGRICULTURE ORGANIZATION, World Reference Base for Soil
Resources, FAO / ISRIC / ISSS (1998).
[6] HERRIDGE, D.F., Use of the ureide technique to describe the nitrogen economy of
field-grown soybeans, Plant Physiol. (1982) 70 (1982) 7–11.
[7] HERRIDGE, D.F., et al., Measurement of nitrogen fixation by soybean in the field
using the ureide and natural 15N abundance methods, Plant Physiol. 93 (1990)
708–716.
[8] HERRIDGE, D.F., PEOPLES, M.B., BODDEY, R.M., Global inputs of biological
nitrogen fixation in agricultural systems, Plant Soil 311 (2008) 1–18.
[9] BREMNER, J.M., MULVANEY, C.S., “Nitrogen – Total”, Methods of Soil Analysis,
Part 2, 2nd edn, (PAGE, A.L., Ed.), Agronomy Monogr. 9, ASA and SSSA, Madison,
WI (1982) 595–624.
[10] OKALEBO, J.R., GATHUA, K.W., WOOMER, P.L., “Fractionation of organic
matter by particle size”, Laboratory Methods of Soil and Plant Analysis: A Working
Manual, (OKALEBO, J.R., et al., Eds), TSBF-KARI-UNESCO, Nairobi (1993) 66–68.
[11] MANSKE, G.G.B., et al., Importance of P uptake efficiency versus P utilization for
wheat yield in acid and calcareous soils in Mexico, Eur. J. Agron. 14 (2001) 261–274.
[12] MANSKE, G.G.B., et al., Phosphorus use efficiency in tall, semi-dwarf and dwarf
near-isogenic lines of spring wheat, Euphytica 125 (2002) 113–119.
[13] ORTIZ-MONASTERIO, R.J.I., et al., Genetic progress in wheat yield and nitrogen
use efficiency under four N rates, Crop Sci. 37 (1997) 898–904.
[14] SAS., SAS / STAT guide for personal computers, version 9.1, SAS Institute Inc., Cary,
NC (2003).
[15] YAN, W., GGE biplot: A Windows application for graphical analysis of multi-
environment trial data and other types of two-way data, Agron. J. 93 (2001) 1111–
1118.
[16] YAN, W., KANG, M.S., GGE Biplot Analysis: A Graphical Tool for Breeders,
Geneticists, and Agronomists, CRC Press, Boca Raton, FL (2003).
[17] ROSE, T.J., WISSUWA, M., Rethinking internal phosphorus utilization efficiency: A
new approach is needed to improve PUE in grain crops, Adv. Agron. 116 (2012)
185–217.

320
 IAEA PUBLICATIONS ON SOIL AND WATER MANAGEMENT AND CROP

NUTRITION

2001 Use of Isotope and Radiation Methods in Soil and Water Management and Crop
Nutrition –Manual (IAEA Training Course Series No. 14).

2002 Water Balance and Fertigation for Crop Improvement in West Asia (IAEA-
TECDOC-1266).

2002 Assessment of Soil Phosphorus Status and Management of Phosphatic Fertilisers to

Optimise Crop Production (IAEA-TECDOC-1272).

2002 Nuclear Techniques in Integrated Plant Nutrient, Water and Soil Management (Proc.
Symp.Vienna, 2000) (IAEA C+S Paper Series 11/P).

2002 Irradiated Sewage Sludge for Application to Cropland (IAEA-TECDOC-1317).

2003 Neutron and Gamma Probes: Their Use in Agronomy (Second Edition) (IAEA
Training Course Series No. 16).

2003 Les Sondes à Neutrons et à Rayons Gamma: Leur Applications en Agronomie

(Deuxième Édition) (Collection Cours de Formation No.16).

2003 Las Sondas de Neutrones y Gamma: Sus Aplicaciones en Agronomía (Segunda

Edición) (Colección Cursos de Capacitación No. 16).

2003 Management of Crop Residues for Sustainable Crop Production (IAEA-TECDOC-

1354).

2003 Guidelines for the Use of Isotopes of Sulfur in Soil-Plant Studies (IAEA Training
Course Series No. 20).

2003 Maximising the use of biological nitrogen fixation in agriculture (Developments in

Plant and Soil Sciences No. 99. FAO, Rome, and Kluwer Academic Publishers,
Dordrecht).

2004 Use of phosphate rocks for sustainable agriculture. (FAO Fertilizer and Plant
Nutrition Bulletin 13. Joint FAO/IAEA Division of Nuclear Techniques in Food and
Agriculture and the Land and Water Development Division-AGL, FAO, Rome).

2004 Use of Isotope and Radiation Methods in Soil and Water Management and Crop
Nutrition (An interactive CD).

2005 Nutrient and water management practices for increasing crop production in rainfed
arid/semi-arid areas (IAEA-TECDOC-1468).

2007 Management Practices for Improving Sustainable Crop Production in Tropical Acid
Soils Results of a Coordinated Research Project organized by the Joint FAO/IAEA

321
 Programme of Nuclear Techniques in Food and Agriculture, Proceedings Series,
STI/PUB/1285.

2008 Guidelines on Nitrogen Management in Agricultural Systems, Training Course

Series No. 29: IAEA-TCS-29/CD - ISSN 1998-0973.

2008 Field Estimation of Soil Water Content, A Practical Guide to Methods,

Instrumentation and Sensor Technology, Training Course Series No. 30. Also IAEA-
TCS-30/CD - ISSN 1998-0973.

2008 Management of Agroforestry Systems for Enhancing Resource use Efficiency and
Crop Productivity, IAEA-TECDOC-1606, ISBN 978-92-0-110908-8.

2009 Management of Agroforestry Systems for Enhancing Resource use Efficiency and
Crop Productivity, IAEA-TECDOC-CD-1606, ISBN 978-92-0-150909-3.

2011 Impact of Soil Conservation measures on Erosion Control and Soil Quality. IAEA-
TECDOC-1665, Vienna. 329 p.
http://www-pub.iaea.org/MTCD/Publications/PDF/TE_1665_web.pdf

2012 Greater Agronomic Water Use Efficiency in Wheat and Rice using Carbon Isotope
Discrimination, IAEA-TECDOC-1671, Vienna. 280 p.
http://www-pub.iaea.org/MTCD/Publications/PDF/TE_1671_web.pdf

322
 ABBREVIATIONS AND ACRONYMS

ARC (WARDA) Africa Rice Center (West Africa Rice Development Association)
AM Arbuscular mycorrhizal
ARB Adventitious Root Branching
ARL Adventitious Root Length
ARN Adventitious Roots Number
BNF Biological Nitrogen Fixation
BRB Basal Root Branching
BRD Basal Root Depth
BRGA Basal Root Growth Angle
BRL Basal Root Length
BRN Basal Root Number
BRWN Basal Root Whorls number
CENA/USP Center for Nuclear Energy in Agriculture/University of Sao Paulo
CGIAR Consultative Group on International Agricultural Research
CIAT Centro Internacional de Agricultura Tropical (International Center for
Tropical Agriculture), Colombia
CIMMYT Centro Internacional de Mejoramiento de Maiz y Trigo (International
Maize and Wheat Improvement Center), Mexico
CRP Coordinated Research Project
ENR Efficient and non-responsive
ER Efficient and responsive
FAO Food and Agriculture Organization, Rome
GY Grain Yield
ICRISAT International Crops Research Institute for the Semi-Arid Tropics, India
IITA International Institute of Tropical Agriculture, Nigeria
INRAN Institut National de Recherche Agronomique du Niger (INRAN)
IRAD Institute of Agricultural Research for Development
IRRI International Rice Research Institute, Philippines
LAI Leaf Area Index
LCHL Leaf Chlorophyll
LRL Lateral Root Length
LRN Lateral Root Number
NARS National Agricultural Research Systems
NENR Non-efficient and non-responsive
NER Non-efficient and responsive,
NERICA New Rice for Africa
NODWT Nodule weight
NUE Nitrogen use efficiency
PAE Phosphorus acquisition efficiency
PCA Principal Component Analysis
PEI Phosphorus efficiency index
PLHT Plant Height
PRB Primary Root Branching
PRD Primary Root Depth
PUE Phosphorus utilization efficiency
PUI Phosphorus utilization index
RA Root Angle
RAMC Root Arbuscular Mycorrhiza Colonization
RB Root Branching

323
RCA Root cortical aerenchyma
RDIA Root Diameter
RGY Relative Grain Yield
RHLD Root Hairs Length Density
RILs recombinant inbred lines
RL Root Length
RLD Root Length Density
RSA Root Surface Area
RTV Root Volume
SA Specific activity
SARI Savannah Agricultural Research Institute
SHB Shoot Biomass
SNF Symbiotic nitrogen fixation
SRE Seminal Root Elongation
SRL Seminal Root Length
STDIA Stem Diameter
UNESCO United Nations Educational, Scientific and Cultural Organization

324
 LIST OF PARTICIPANTS

Abaidoo, R.C. International Institute of Tropical Agriculture

Oyo Road
P.M.B. 5320, Ibadan
Nigeria
Email [email protected]

Atkins, C.A. School of Plant Biology (M090)

The University of Western Australia
35 Stirling Highway, Crawley, WA 6009
Australia
Email:[email protected]
Email [email protected]

Bado, V.B. Africa Rice Center (AfricaRice)

BP 96, Saint Louis,
Senegal
E-mail [email protected]

Bayuelo Jiménez, J.S. Instituto de Investigaciones Agropecuarias y Forestales

Universidad Michoacana de San Nicolás de Hidalgo
Km. 9.5 Carr. Morelia Zinapécuaro
58880 Morelia, Michoacán
Mexico
Email [email protected]

Drevon, J.J. Institut National de la Recherche Agronomique (INRA),

UMR Eco&Sols, 2, Place Viala,
34060 Montpellier CEDEX 01
France
Email [email protected]
Email [email protected]

Gómez Jorrin, L.A Soil Institute, Autopista Costa - Costa,

Antigua Carretera de Vento
La Habana, Capdevila Boyeros 10800
Cuba
Email [email protected]

Horst, W.J. Institute for Plant Nutrition,

Leibniz University of Hannover,
30419 Hannover
Federal Republic of Germany
Email [email protected]

325
 Kugblenu, Y.O. Department of Crop Science
University of Ghana
P.O. Box 44
Legon Accra
Ghana
Email [email protected]

Lynch, J.P. Department of Horticulture

Pennsylvania State University
102 Tyson Building
University Park, PA 16802
United States of America
Email [email protected]

Miguel, M. Agricultural Research Institute of Mozambique

Ministry of Agriculture
FPLM Avenue 2698
PO Box 3658
Maputo
Mozambique
Email [email protected]
Email [email protected]

Muraoka, T. Centro de Energia Nuclear na Agricultura

Universidade de Sao Paulo (USP)
DVPROD, Soil Fertility, 4720
Avenida Centenario 303
Caixa Postal 96, 13400-970 Piracicaba, SP
Brazil
Email [email protected]
Email [email protected]
Email [email protected]

Musa, M.H. Department of Land Management

Universiti Putra Malaysia (UPM)
Please fill the Address Line 1
43400 Serdang, Selangor
Malaysia
Email [email protected]

The, C. Institut de recherche agronomique de Nkolbisson

B.P. 2067
Yaoundé
Cameroon
Email [email protected]

326
Traore, K. Institut de l'environnement et de recherches agricoles
B.P. 910
Bobo-Dioulasso 01
Burkina Faso
Email [email protected]
Email [email protected]

Yang, J. Institute of Agricultural Resources and Regional Planning

Chinese Academy of Agricultural Sciences (CAAS)
12 Southern Street of Zhong-Guan-Cun
Beijing 100081
China
Email [email protected]

Consultant Meeting
Vienna, Austria: 23–27 July 2005

Research Coordination Meetings

Vienna, Austria: 16–20 October 2006
Morelia, Mexico: 21–25 April 2008
Maputo, Mozambique: 23–27 August 2010
Vienna, Austria: 14–18 November 2011

327
@ No. 23

ORDERING LOCALLY
In the following countries, IAEA priced publications may be purchased from the sources listed below,
or from major local booksellers.
Orders for unpriced publications should be made directly to the IAEA. The contact details are given at
the end of this list.

AUSTRALIA
DA Information Services
648 Whitehorse Road, Mitcham, VIC 3132, AUSTRALIA
Telephone: +61 3 9210 7777  Fax: +61 3 9210 7788
Email: [email protected]  Web site: http://www.dadirect.com.au
BELGIUM
Jean de Lannoy
Avenue du Roi 202, 1190 Brussels, BELGIUM
Telephone: +32 2 5384 308  Fax: +61 2 5380 841
Email: [email protected]  Web site: http://www.jean-de-lannoy.be
CANADA
Renouf Publishing Co. Ltd.
Telephone: +1 613 745 2665  Fax: +1 643 745 7660
5369 Canotek Road, Ottawa, ON K1J 9J3, CANADA
Email: [email protected]  Web site: http://www.renoufbooks.com
Bernan Associates
4501 Forbes Blvd., Suite 200, Lanham, MD 20706-4391, USA
Telephone: +1 800 865 3457  Fax: +1 800 865 3450
Email: [email protected]  Web site: http://www.bernan.com
CZECH REPUBLIC
Suweco CZ, spol. S.r.o.
Klecakova 347, 180 21 Prague 9, CZECH REPUBLIC
Telephone: +420 242 459 202  Fax: +420 242 459 203
Email: [email protected]  Web site: http://www.suweco.cz
FINLAND
Akateeminen Kirjakauppa
PO Box 128 (Keskuskatu 1), 00101 Helsinki, FINLAND
Telephone: +358 9 121 41  Fax: +358 9 121 4450
Email: [email protected]  Web site: http://www.akateeminen.com
FRANCE
Form-Edit
5, rue Janssen, PO Box 25, 75921 Paris CEDEX, FRANCE
Telephone: +33 1 42 01 49 49  Fax: +33 1 42 01 90 90
Email: [email protected]  Web site: http://www.formedit.fr
Lavoisier SAS
14, rue de Provigny, 94236 Cachan CEDEX, FRANCE
Telephone: +33 1 47 40 67 00  Fax: +33 1 47 40 67 02
Email: [email protected]  Web site: http://www.lavoisier.fr
L’Appel du livre
99, rue de Charonne, 75011 Paris, FRANCE
Telephone: +33 1 43 07 50 80  Fax: +33 1 43 07 50 80
Email: [email protected]  Web site: http://www.appeldulivre.fr
GERMANY
Goethe Buchhandlung Teubig GmbH
Schweitzer Fachinformationen
Willstaetterstrasse 15, 40549 Duesseldorf, GERMANY
Telephone: +49 (0) 211 49 8740  Fax: +49 (0) 211 49
Email: [email protected]  Web site: http://www. http://www.goethebuch.de/
HUNGARY
Librotade Ltd., Book Import
PF 126, 1656 Budapest, HUNGARY
Telephone: +36 1 257 7777  Fax: +36 1 257 7472
Email: [email protected]  Web site: http://www.librotade.hu
INDIA
Allied Publishers
1st Floor, Dubash House, 15, J.N. Heredi Marg
Ballard Estate, Mumbai 400001, INDIA
Telephone: +91 22 2261 7926/27  Fax: +91 22 2261 7928
Email: [email protected]  Web site: http://www.alliedpublishers.com
Bookwell
3/79 Nirankari, Dehli 110009, INDIA
Tel.: +91 11 2760 1283  +91 11 27604536
Email: [email protected]  Web site: http://www.bookwellindia.com/
ITALY
Libreria Scientifica “AEIOU”
Via Vincenzo Maria Coronelli 6, 20146 Milan, ITALY
Tel.: +39 02 48 95 45 52  Fax: +39 02 48 95 45 48
Email: [email protected]  Web site: http://www.libreriaaeiou.eu/
JAPAN
Maruzen Co., Ltd.
1-9-18 Kaigan, Minato-ku, Tokyo 105-0022, JAPAN
Tel.: +81 3 6367 6047  Fax: +81 3 6367 6160
Email: [email protected]  Web site: http://maruzen.co.jp
NETHERLANDS
Martinus Nijhoff International
Koraalrood 50, Postbus 1853, 2700 CZ Zoetermeer, NETHERLANDS
Tel.: +31 793 684 400  Fax: +31 793 615 698
Email: [email protected]  Web site: http://www.nijhoff.nl
Swets
PO Box 26, 2300 AA Leiden
Dellaertweg 9b, 2316 WZ Leiden, NETHERLANDS
Telephone: +31 88 4679 263  Fax: +31 88 4679 388
Email: [email protected]  Web site: www.swets.com
SLOVENIA
Cankarjeva Zalozba dd
Kopitarjeva 2, 1515 Ljubljana, SLOVENIA
Tel.: +386 1 432 31 44  Fax: +386 1 230 14 35
Email: [email protected]  Web site: http://www.mladinska.com/cankarjeva_zalozba
SPAIN
Diaz de Santos, S.A.
Librerias Bookshop  Departamento de pedidos
Calle Albasanz 2, esquina Hermanos Garcia Noblejas 21, 28037 Madrid, SPAIN
Telephone: +34 917 43 48 90
Email: [email protected]  Web site: http://www.diazdesantos.es/
UNITED KINGDOM
The Stationery Office Ltd. (TSO)
PO Box 29, Norwich, Norfolk, NR3 1PD, UNITED KINGDOM
Telephone: +44 870 600 5552
Email (orders): [email protected]  (enquiries): [email protected]  Web site: http://www.tso.co.uk
On-line orders:
DELTA International Ltd.
39, Alexandra Road, Addlestone, Surrey, KT15 2PQ, UNITED KINGDOM
Email: [email protected]  Web site: http://www.profbooks.com
United Nations (UN)
300 East 42nd Street, IN-919J, New York, NY 1001, USA
Telephone: +1 212 963 8302  Fax: 1 212 963 3489
Email: [email protected]  Web site: http://www.unp.un.org
UNITED STATES OF AMERICA
Bernan Associates
4501 Forbes Blvd., Suite 200, Lanham, MD 20706-4391, USA
Tel.: +1 800 865 3457  Fax: +1 800 865 3450
Email: [email protected]  Web site: http://www.bernan.com
Renouf Publishing Co. Ltd.
812 Proctor Avenue, Ogdensburg, NY 13669, USA
Tel.: +800 551 7470 (toll free)  +800 568 8546 (toll free)
Email: [email protected]  Web site: http://www.renoufbooks.com

Orders for both priced and unpriced publications may be addressed directly to:
IAEA Publishing Section, Marketing and Sales Unit, International Atomic Energy Agency
13-00001

Vienna International Centre, PO Box 100, 1400 Vienna, Austria

Telephone: +43 1 2600 22529 or 22488 • Fax: +43 1 2600 29302
Email: [email protected] • Web site: http://www.iaea.org/books
11-48712
IAEA TECDOC 1721 Optimizing Productivity of Food Crop Genotypes in Low Nutrient Soils

International Atomic Energy Agency

ISBN 978–92–0–113113–3
ISSN 1011–4289
Vienna