Plant Analysis as a Tool to Determine

Crop Nitrogen Status

– Towards Leaf Area Based Measurements

Göran Ekbladh
Faculty of Natural Resources and Agricultural Sciences
Department of Soil Sciences
Uppsala

Doctoral thesis
Swedish University of Agricultural Sciences
Uppsala 2007

1
Acta Universitatis Agriculturae Sueciae
2007: 109

ISSN 1652-6880
ISBN 978-91-85913-08-4
© 2007 Göran Ekbladh, Uppsala
Tryck: SLU Service/Repro, Uppsala 2007

2
Abstract
Ekbladh, G. 2007. Plant analysis as a tool to determine crop nitrogen status
– towards leaf area based measurements. Doctoral thesis.
ISSN 1652-6880, ISBN: 978-91-85913-08-4

An effective plant nutrient management strategy optimises nitrogen (N) use efficiency for
minimised environmental impact, while ensuring an optimum N status of the crop for good
product quality and maximum growth. Soil or plant analysis can be used to evaluate the
strategy; however the use of plant analysis for this purpose has been limited. One reason is
lack of reliable reference values for the critical concentration needed for optimal growth.
This study builds on theories that relate ontogenetic changes in the critical N concentration
to changes in the relation between mass and surface area of the entire plant and of
individual leaves. Through the establishment of critical N concentrations on the basis of
these theories, some of the drawbacks hitherto experienced with plant analysis, such as
difficulties in defining growth stage or plant part to sample, can be avoided.
The aim of this thesis was to establish critical N concentrations for white cabbage
(Brassica oleracea L. var. capitata L. f. alba D.C.) on the basis of these theories. Multi-N-
rate and multi-harvest experiments were conducted in the field and in a climate chamber.
The results showed that the critical N concentration declined at the same rate (-0.33) as
the plant’s leaf area ratio (leaf area divided by plant mass), which is in agreement with the
2/3-Power rule or “skin-core” hypothesis. The critical N concentration (% of DM) on a
whole plant basis was estimated to 4.5 (W<1.5 t ha-1) and to 5.1W-0.33 (W>1.5 t ha-1), where
W is weight per unit area of plant dry matter exclusive of roots. Moreover, it was concluded
that the unshaded horizontally orientated leaves of cabbage can be used for leaf area based
plant analysis of individual leaves. The critical N concentration of these leaves expressed on
an area basis was found to be 3.7 g N m-2, while that for the whole plant N on a leaf area
basis was 4.7 g N m-2. The ratio of these two critical concentrations, 0.8, was similar to the
leaf N ratio (leaf N/whole plant N) of young plants before self shading occurs.
Key words: critical nitrogen concentration, ontogenetic decline, leaf area ratio, leaf nitrogen
ratio, specific leaf nitrogen
Author’s address: Göran Ekbladh, Department of Soil Sciences, Swedish University of
Agricultural Sciences (SLU), Box 7014, SE-750 07 Uppsala, Sweden.
E-mail: [email protected]

3
4
Contents
Introduction 9
Background 9
Soil and plant analysis for plant nutrient management
– problems and possibilities 9
Indicators of nutrient status 10
The soil mineral N as an indicator 10
Crop N status as an indicator 11
Leaf area - nitrogen - mass relationships; – whole plants 14
Leaf area - nitrogen - mass relationships; – individual leaves 18
Possibilities for plant analysis based on critical N concentrations
derived from leaf area or growth 19
Aims and hypothesis 20
Materials and Methods 21
Results and discussion 24
Growth – determination of growth rate and relative growth rate 24
Whole plant critical N 24
Plant N on an area basis (PNCa) 27
Ontogenetic changes in PNC and in growth components 27
Effect of self shading on the ontogenetic decline in PNC 31
Leaf N on an area basis (LNCa) 32
Applications 36
Conclusions 39
References 40
Appendices 47
Appendix A, Relationship of LAR and PNC 47
Appendix B, Determination of LNCa from LNC of Huett & Rose (1989) 48
Acknowlegdements 49

5
Preface

Papers I –IV
The thesis is based on the following papers, which are referred to by their Roman
numerals:

I. Göran Ekbladh, Ernst Witter, Tom Ericsson, 2007.

Ontogenetic decline in the nitrogen concentration of field grown white
cabbage – Relation to growth components
Scientia Horticulturae, 112, 149-155.
II. Göran Ekbladh, Tom Ericsson, Ernst Witter
Ontogenetic changes in the nitrogen concentration of white cabbage — I.
Relation to relative growth rate and its components
(Manuscript)
III. Göran Ekbladh, Tom Ericsson, Ernst Witter
Ontogenetic changes in the nitrogen concentration of white cabbage — II.
Leaf-area-based assessments on individual leaves (Manuscript)
IV. Göran Ekbladh, Ernst Witter
Determination of the critical nitrogen concentration of white cabbage
(Manuscript)

Reprint of paper I is published with the permission of Elsevier Limited.

Göran Ekbladh’s contribution to the papers included in this thesis was as follows:

I Data analysis, wrote manuscript with contributions from Witter and

Ericsson.
II Planned the experiment together with Ericsson, work in climate chambers,
data analysis, wrote manuscript with contributions from Ericsson and
Witter
III Planned the experiment together with Ericsson, work in climate chambers,
data analysis, wrote manuscript with contributions from Ericsson and
Witter
IV Planned the experiment, field work, data analysis, wrote manuscript with
contributions from Witter

6
Abbreviations
DAT Days after transplanting
DM Dry matter
GR (Absolute) growth rate [g plant-1 day-1]
LA Leaf area [m2]
LAI Leaf area index (leaf area/ground area)
LAR Leaf area ratio (LA/W) [cm2 g-1]
LN Accumulated leaf nitrogen [g N plant-1]
LNC Leaf nitrogen concentration (LN/LW, leaf weight) [Percent of DM]
LNCa Leaf nitrogen concentration on an area basis (LN/LA) [g N m-2]
LNP Leaf nitrogen productivity (GR/LN) [g plant-1 day-1 g-1]
LNR Leaf nitrogen ratio (LN/PN) [fraction]
N Nitrogen
NAR Net assimilation rate (GR/LA) [g plant-1 day-1 m-2]
NDF Neutral detergent fibre [g g-1]
NNI Nitrogen nutrition index [fraction]
NP Nitrogen productivity (GR/PN) [g plant-1 day-1 g-1]
PFD Photon flux density [mmole m-2 s-1]
PN Accumulated plant nitrogen (the amount N taken up by the plant, roots
excluded) [g N plant-1 or kg N ha-1]
PNC Plant nitrogen concentration [Percent of DM]
PNCa Plant nitrogen concentration on an area basis (PN/LA) [g N m-2]
PNR Average rate of daily accumulated plant N during the linear growth phase
(PN/DAT) [kg N ha-1 day-1]
RGR Relative growth rate [day-1]
SLA Specific leaf area (LA/LW) [g m-2]
W Weight per unit ground area of plant dry matter exclusive of roots [g
plant-1 or t ha-1]
Index c critical
Index org organic
Index m metabolic compartment
Index s structural compartment

7
8
Introduction
There has been a considerable research effort on developing techniques in soil and
plant analysis to improve the efficiency of fertilizer use for a minimised
environmental impact. The adoption of soil or plant analysis in vegetable
production has however often been low (Hartz, 2004). The reasons for limited
routine use of soil or plant analysis for appropriate plant nutrient management may
be very different, for example time-consuming sampling or difficulties in
establishing reference data against which the actual nutrient status of a crop can be
evaluated. In this thesis reference data (critical concentrations) have been derived
on the basis of a theoretical framework presented in the literature (Caloin & Yu,
1982: Caloin & Yu, 1984: Grindlay, 1997; Lemaire & Gastal, 1997; Lemaire et al.,
1997). First, the problems and possibilities hitherto encountered with soil and plant
analysis for plant nutrient management, are reviewed. The review concludes with a
suggestion for how some of the problems with plant analysis may be overcome
with leaf area based plant analysis and by referring critical concentrations to
biomass instead of time or development stage. The results are presented both on a
basis of the whole plant and of individual leaves. Plant analysis is mainly
diagnostic, as it only retrospectively reflects the plant nitrogen status, but may be a
valuable tool for evaluation of plant nutrient management strategies attempting to
more closely adjust supply to crop demand. The results are based on experiments
conducted both under field conditions and in a climate chamber with the white
cabbage cultivar SW Heckla (Brassica oleracea L. var. capitata L. f. alba D.C.).

Background

Soil and plant analysis for plant nutrient management –

problems and possibilities
The ability to identify deficiencies or excesses of plant nutrients in crops is
important in order to achieve an efficient utilisation of available nutrient resources
for crop production. Excess nutrients in the soil-crop system will increase the risk
of nutrient losses to the environment and the risk of impaired product quality
(Belec et al., 2001; Santamaria, 2006), whereas deficiencies will reduce crop
production. The risk for the former is higher when high application rates are used,
as is the case for many vegetable crops which have high nitrogen (N) demands and
leave N rich crop residues after harvest (Goulding, 2000; Neeteson & Carton,
2001). This can appear to be a minor problem because the amounts of fertilizer
used for vegetables are marginal in comparison to the total amounts used in
agriculture, since the land area used is small. In Sweden, vegetables were grown on
7047 ha in 2005 (SCB and SJV, 2006a) of the total 2.7 million ha arable land
(SCB and SJV, 2006b), while cabbage was grown on 370 ha (SCB and SJV,
2006c) compared to 113 590 ha in total in Europe (Eurostat, 2007; EU27;
Portugal, Spain and UK not included). However, although the area is small and
amounts of fertilizer used in vegetable growing are low compared to the total

9
amounts used, local effects on the environment can be considerable. Nitrate
leaching may result in groundwater concentrations that exceed the emission limits
of the EU Nitrate Directive (Monteny, 2001). The estimated average potential
nitrate leaching in Denmark 1987-92 was twice as much on vegetable farms (122
kg N ha-1) compared to farms growing standard arable crops (52 kg N ha-1) (Huus-
Bruun, 1993). Fertilizer strategies based on conventional recommendations and
measures have been shown to be insufficient to meet the EU Nitrate Directive
(Salomez et al., 2005; van Dijk & Smit, 2006). Nutrient balances from different
field vegetable production systems indicate large nutrient surpluses, suggesting that
there is a potential to improve the efficiency in fertilizer use without reductions in
yield (Neeteson et al., 2003). The attitude of the farmers is important for progress,
their motivation is necessary for a change in fertilizer strategy (Booij et al., 2003).
Evident indications of the nutrient status in their own fields and tools to measure
and evaluate the adequacy of their fertilizer programs and nutrient management
strategy may increase their motivation. This applies to both conventional and
organic production. There is a wide range of tools for evaluation of the nutrient
status of the soil or crop, ranging from simple tools that can be managed by the
farmers themselves, such as the greenness in a “fertilizer window” (Rimpau, 1984),
to methods of soil and plant analysis which require laboratory analysis.

Indicators of nutrient status

Whatever tool is used, its aim is to serve as an indicator of the actual nutrient status
of the soil-crop system. Indicators can be used to evaluate the actual plant nutrient
management strategy (diagnostic indicators) or to give predictive information such
as information on the actual fertilizer requirement for the next application
(prognostic indicators) (Lewis, 1993; Schröder et al., 2000). The use of indicators
to evaluate the actual practice implies a participatory learning process by which the
farmer’s motivation for a change is encouraged (Roling & Wagemakers, 2000).
The management of plant nutrients can be successively improved by evaluation of
the fertilizer strategy.

Generally, an ideal indicator must be reproducible (Schröder et al., 2000). For

evaluation of the nutrient status, the indicator should interpret the actual nutrient
status of the soil-crop system in the same manner over different sites and years.
The indicator can pertain to soil or to plant.

The soil mineral N as an indicator

Soil N analysis evaluates the soil mineral N supply in relation to the expected
demand of the crop and serves as a prognostic tool as it predicts the fertilizer
requirement for the remaining growth period. According to the Nmin-method
(Wehrman & Scharpf, 1979), the actual application rate of N fertilizer is estimated
from the N demand of the crop and by adjusting for the actual soil mineral N
content within the assumed rooting depth. A recommendation system, the KNS-
system, compiles data on N demand for several vegetable crops (Lorenz et al.,
1989; Feller et al., 2001). An important advantage with the method is that the

10
fertilizer requirement for any arbitrarily chosen period within the growth period
can be estimated. A similar system has been developed in the US, the pre-side soil
nitrate test (PSNT). Pre-side refers to seasonal applications beside the plants. The
nitrate content of the soil is measured (Magdoff, 1991) as, for example, in the Nmin-
method. A pre-side nitrate test critical concentration is defined as the value above
which no sidedress N is needed. Below the critical pre-side nitrate test
concentration sidedress N should be applied according to a standard rate. Pre-side
nitrate test critical concentration has been developed for different crops and also
for vegetables such as cabbage (Heckman et al., 2002). Considerable reductions in
seasonal N applications without reduction in yield were made possible by use of
the pre-side nitrate test in commercial celery and lettuce production (Hartz et al.,
2000). According to Hartz (2003) the pre-sidedress soil nitrate testing is preferable
to plant analysis for decision support for seasonal (sidedress) applications. Farmers
have been encouraged to use the Nmin method in vegetable production (Scharpf,
1991). However, adoption of the method for routine soil analysis by farmers has
been low (Hartz, 2004). Sampling is time-consuming, especially below the top soil
for deep-rooted crops. Acceptance by farmers has been better in areas where close
co-operation with a nearby laboratory is possible. In such a co-operation,
laboratory personnel were responsible for the soil sampling and provided
recommendations immediately after sampling (Ziegler et al., 1996).

Crop N status as an indicator

Plant analysis is an important tool for diagnostic evaluation of the nutrient status of
a crop (Mills & Jones, 1996; Reuter & Robinson, 1997; Kalra, 1998). It has been
widely used for identifying plant nutrition deficiencies and disturbances in crops
but only to a lesser degree for routine evaluation of the plant nutrient status for
adequate plant nutrient management.

An evaluation of the nutrient status is made possible only by relating the actual
status to a standard. The concept of critical percentage, introduced by Macy
(1936), can be used as such a standard or reference value. It suggests that there is a
critical nutrient concentration for each nutrient and for each kind of plant. Lemaire
& Gastal (1997) defined the critical N concentration as: at a given crop dry matter
a certain critical plant N concentration (PNCc) in the dry matter mass is needed to
obtain the maximum instantaneous growth rate (GR). The PNCc usually refers to
optimal growth, but could as well refer to other properties such as susceptibility to
physiological disorders or to diseases.

The results of plant analysis are affected by environmental factors such as soil
and climate, and plant factors such as the plant’s development stage which have to
be taken into account when interpreting the result of the analysis (Lewis et al.,
1993). Also, critical concentrations may vary depending on the conditions when
they were determined (Bates, 1971) and therefore critical nutrient ranges (CNR)
have been preferred instead of a sharp limit between deficiency and sufficiency
(Dow & Roberts, 1982). The use of CNR does however not allow for a precise
determination of the nutrient status. Therefore, one of the reasons for the low
adoption of plant analysis may be the difficulty in interpreting the results against

11
reliable reference or standard data because of the different factors affecting them.
These factors more or less generally affect the outcome of plant analysis
independently of the method used.

Several environmental factors may affect the results of plant analysis. The site
affects the results of the analysis by various factors such as soil type (Westerveld et
al., 2003b), soil moisture content (Swaider et al., 1988), fertilizer source (Barker et
al., 1971) and climate (Sorensen et al. 2006). Differences and variability in climate
can be accounted for by relating growth to degree days rather than to time or
growth stage (Grevsen, 1998). Plant nitrate concentration has been found to vary
with the time of day. Lower nitrate concentrations are found in early afternoon and
on sunny days (Iversen et al., 1985). Scaife & Stevens (1983) recommended taking
samples within two hours of midday.

Despite local variations in soil types and management, there was good agreement
between critical petiole nitrate concentrations of potato derived from different
studies by Gardner & Jones (1975), MacMurdo et al., (1988), Porter & Sisson
(1991) and Bélanger et al., (2003). The three latter studies were conducted in the
same region of Atlantic North East America, whereas the study by Gardner &
Jones (1975) was conducted in Idaho in the North West. The same cultivars were
used except by MacMurdo et al., (1988). In spite of the variability in
environmental conditions, the results support the possibility of establishing a
general applicable critical concentration in regions with similar climatic conditions,
as suggested by Bélanger et al., (2003).

Nutrient interactions may affect the concentration of a certain element. Higher

concentrations may appear when growth is limited by another element, compared
to when the nutrients are available in adequate proportions. Focus is often on N
being the most decisive element for growth. However, there is always the risk for
erroneous interpretation of the results if the concentrations of other elements are
unknown. Multi-element analysis is therefore preferable although more costly.

The critical petiole NO3-N concentration and total N concentration vary with the
cultivar as was shown for potato by MacMurdo et al., (1988) and Porter & Sisson
(1991) and for onion by Westerveld et al., (2003b). Therefore, cultivar specific
critical N concentrations may be needed.

Nutrient concentrations decline ontogenetically during the growth period, even

with sufficient N supply (Siman 1974; Sorensen, 2000). Therefore, the critical
concentration has to be related to a carefully defined growth stage (Lorenz &
Tyler, 1977). However, the way the growth stages are defined is often imprecise.
Typical examples of growth stages for cabbage referred to in the literature are “2 to
3 months old” (Mills & Jones, 1996) and “at heading” (Maynard & Hochmut,
1997). Westerveld (2003b) concluded that the main difficulty in using critical
concentrations is to match the stage of sampling to the growth stages for which
critical concentrations are given in the literature.

12
 Concentrations vary between plant parts. The plant part to be sampled should be
sensitive to variation in nutrient supply and should be easy to identify for correct
sampling. For small plants whole shoots may be sampled, whereas for bigger plants
sampling just a part of the plant is more convenient. The leaf is the plant part
commonly recommended for sampling (Benton Jones Jr, 1985). According to
Geraldson et al. (1973), the youngest fully mature leaf is preferable for nutrient
analysis for many crops. Tabor et al. (1984) suggested sampling of fully mature
leaves because immature and not fully extended leaves were less sensitive to
changes in nitrate N content and the nitrate content varied considerably between
immature leaves, whereas there was no significant difference in nitrate N content
between the first, second and third mature leaf. For cabbage the wrapper leaves
(Westerveld et al., 2003b), the most recently fully expanded leaf and the youngest
fully opened leaf which later become the wrapper leaves (Huett & Rose, 1989)
have been sampled. Concentrations of nutrients vary between individual leaves at
different positions (Dole & Wilkins, 1991). Therefore, the leaf position sampled
for diagnostic analysis must agree with the leaf position that the critical
concentration refers to. The lowest concentration is found in the upper leaves,
except for the youngest leaves, as shown by Geyer & Marschner (1990) for maize.
In contrast, leaves at the bottom of the canopy will first indicate a sudden
deficiency as nutrients will be translocated from lower leaves to upper leaves in the
case of deficiency (Girardin et al., 1985; Ogunlela et al., 1990). The variability is
however larger for the lower leaves (Scaife & Stevens, 1983). Therefore, the lower
leaves are not commonly used for plant analysis. Thus, there has been a general
preference for choosing upper leaves for plant analysis, but it is difficult to define a
very precise specific leaf position to sample. This may result in a certain variability
in the analysis as concentrations vary between individual leaves.

Besides leaves, petioles are used for plant analysis. Petioles have often been used
for quick tests to estimate nitrate-N in sap (tissue nitrate-N). The quick tests have
been developed for field use to avoid the time lag between sampling and result as
well as the costs of laboratory analysis. Petiole plus midrib nitrate has been shown
to reflect the N status, as for example in potato (Bélanger et al., 2003). The N
status was based on measurements of total N and expressed by the nitrogen
nutrition index (NNI) (Lemaire & Gastal, 1997). The NNI is defined as the ratio
between the actual and the critical concentrations. The relationship between nitrate
and NNI indicated a certain variability (0.29 < R2 < 0.62). Petiole critical nitrate-N
concentrations on a dry matter basis have been derived for several crops and also
for brassicae such as broccoli (Gardner & Roth 1989a), cabbage (Gardner & Roth
1989b) and cauliflower (Gardner & Roth 1990). Linear relationships between field
determinations of nitrate-N in sap and on dry matter in the laboratory have been
found in many studies (Kubota et al., 1996; Kubota et al., 1997, Coulombe et al.,
1999). The quick tests provide rapid answers but with some loss in accuracy.

Westerveld et al. (2003b) estimated both nitrate-N and total-N but found it
difficult to match concentrations from their experiments with recommended critical
concentrations. Cabbage, carrot and onion with varying N status ranging from
deficiency to excess were grown for two years and on two soils and the

13
tissue concentrations were compared with corresponding literature data on critical
concentrations. They found that the nitrate concentrations were very variable so
that total N matched literature data better compared to nitrate, however fertilizer
rates according to these data would have resulted in either under or over-
application of fertilizer. The main difficulty was the discrepancies and lack of
accurate definitions of the stages of sampling. They concluded that a greater
standardisation of sampling procedures would improve the usefulness of tissue
analysis. High variability from tissue analysis was supported by Matthäus & Gysi
(2001). Broccoli petiole sap nitrate-N varied between less than 2500 and up to a
peak of 4000 ppm within a period of four days. Neverthelesss, the coefficient of
variation was lower for the plant sap analysis (9%) compared to the soil-Nmin
analyses (29%). Moreover, the time required for sampling in the field was much
lower for sampling petioles compared to extracting soil cores. Thus, plant analysis
has advantages over soil analysis because there is less variability and sampling is
less time consuming, however difficulties due to variability and defining the
growth stage for sampling remain.

The predictive value of plant analysis has been under debate. For a predictive
function, the nutrient status at a certain growth stage during the growing period
should relate to final yield or to the optimum N application rate. Such relationships
have been shown; for example yield of broccoli was related to nitrate in the midrib
(press sap as well as on a dry matter basis) and to total N in the most recently fully
expanded leaf at different growth stages throughout the growth period (Castellanos
et al., 2001). The optimum N rate for maize was related to tissue N and chlorophyll
meter readings at the time of sidedress application (Scharf, 2001). However, such
relationships are not generally found. A poor relationship between nitrate sap tests
and N uptake of potato together with spatial and temporal variability caused
MacKerron et al. (1995) to question the benefit of the sap nitrate test for adjusting
top-dressings of N-fertilizers. Their criticism was directed against the prognostic
value of the sap nitrate test. Neither was a relationship of sap nitrate to the
optimum amount of N for top dressing found for Brussels sprouts in 46 field trials
conducted by Scaife & Turner (1987). They concluded that the variations in the
amount mineralised after top-dressing and in the N demand obscure such a
relationship. Therefore, Scaife (1988) rejected attempts to derive critical
concentrations of petiole sap nitrate from final yields but instead stressed the
importance of relating the concentrations to the instant growth rate. Hartz (2003)
concludes that plant analysis does not generate data that are useful for estimating
appropriate seasonal N applications, but is valuable to identify N deficiencies.
Thus, in spite of problems with variability, plant analysis may be useful for its
diagnostic function but not for a predictive, prognostic function.

Leaf area - nitrogen - mass relationships; –whole plants

An important advance in taking account of nutrient variability was made with the
observation that despite large differences in growth and nitrogen uptake rates of
tall fescue (Festuca arundinacea Schreb. cv. Ludelle) between years, the
relationship between the plant N concentration on a dry matter basis (PNC) and

14
the weight per unit ground area of plant dry matter (W) was the same in all years
for plants grown under non N-limited conditions (Lemaire and Salette, 1984). This
means that differences in climatic conditions between years affected crop growth
and N uptake, but not the relationship between PNC and W. This overcomes the
problems of variability caused by climate. The PNC has been found to decline in a
typical pattern with an increase in the amount of biomass per unit ground area. The
pattern can be described by the following equation:
PNC = aW b (1)
where a (%) represents the initial N concentration at low biomass densities (< 1 t
ha-1), and b describes the pattern of the decrease of PNC with growth (Lemaire &
Salette, 1984). For a non-limiting N supply, the parameter a was 4.8% and
parameter b was –0.324 for tall fescue up to W around 6 t ha-1 (Lemaire & Salette,
1984). Moreover, this equation has been found to hold for a very wide variety of
crops such as grasses, vegetable crops and cereals with the main differentiation
being between plants with the C3 and those with the C4 pathway of photosynthesis,
which differ in terms of parameter a rather than b (Greenwood et al., 1990).
Therefore, species- or cultivar-specific critical concentrations may not have to be
defined. Another advance is that the growth stage is continuously defined so that
there is a unique PNC for each value of W. The problem in plant analysis of
defining the growth stage or time for sampling is thereby avoided.

Mathematically different, but principally similar equations to describe the

decline in PNCc with plant biomass have been evaluated for a range of agricultural
crops, including vegetables. Greenwood & Draycott (1989) used an exponential
function for several vegetables, but a linear function specifically for brassicae
(Greenwood et al., 1996). The equations were derived from data sets for crops
grown under non N-limiting conditions and do therefore not necessarily describe
the relationship between PNC and W at optimum N supply (i.e. the lowest N
supply and PNC that result in maximum growth) as some luxury consumption may
have occurred (Greenwood et al., 1986). To obtain the relationship between PNC
and W at optimum N supply (PNCc), data from multi-N level experiments are
needed with frequent harvests during the growing season. Similarly, because of
difficulties in statistically determining the PNCc at each harvest with sufficient
accuracy, data from a large number of experiments are also usually needed (e.g.
Justes et al., 1994 and Plénet & Lemaire, 1999). Riley & Guttormsen (1999)
compared the equations describing the decline in PNCc with W using data of PNCc
and W for cabbage obtained from multi-N level experiments in Norway. They
found that the equation of Greenwood & Draycott (1989) underestimated PNCc,
whereas the equation proposed by (Greenwood et al., 1996) overestimated PNCc.

The equations mentioned above are entirely empirical and it is therefore not
possible to test their validity experimentally. In contrast, Ingestad & Lund (1986)
and Ingestad & Ågren (1992) developed theoretically well founded relationships
between the relative growth rate (RGR) and PNC during exponential growth (that
occurs in the plant’s early growth phase) under conditions of steady-state with
respect to the relative nutrient addition rate. Their studies have shown that

15
under these conditions, and with relative addition rate at optimal or suboptimal
rates, PNC is linearly related to the RGR, and therefore, by multiplication by W,
absolute growth rate to the amount of N in the plant. This therefore confirms the
validity that under steady state conditions with a growth-limiting N supply, the
resulting lower growth rate is reflected in a lower PNC of the plant. Even though
the relation between RGR and PNC is well-established under conditions of
constant relative addition rate and exponential growth, the relation between the
declining PNC and increasing W after the period of exponential growth is not as
well-founded in a theoretical framework.

After the exponential growth phase PNC and RGR decline ontogenetically
(Greenwood et al., 1991). During vegetative growth the cause of the decline in
PNC is mainly attributed to a relative increase in ‘structural’ (supportive) tissues
rich in cellulose and lignin with a low N content compared to the higher N content
of ‘metabolic’ tissues accommodating the photosynthetic function with biomass
production (Warren Wilson, 1972). Based on a conceptual model of the two
compartments of plant nitrogen, Caloin & Yu (1984) related the decline in PNC to
RGR. They assumed that the N concentrations of the two compartments PNCm
(metabolic) and PNCs (structural) were constant and that the change in biomass
proportion of the two types of tissues caused the decline in PNC. Moreover, they
assumed that the GR was proportional to the amount of metabolic biomass by the
proportionality constant k as this tissue accommodates the function for biomass
production with photosynthesis. Based on this assumption, Caloin & Yu (1984)
related PNC to RGR as
PNC = ( PNC m − PNC s ) / k × RGR + PNC s (2)

Equation 2 implies proportionality between PNC and RGR, but to examine the
ontogenetic decline in relation to growth and development they have to be related
to W. Caloin & Yu (1982) derived an expression for GR as a function of W:
GR = k Wa
and in consequence
RGR = GR/W = k Wa-1
where k is a constant and a a scaling exponent.
According to Caloin & Yu (1984), GR is proportional to the biomass of the
metabolic compartment (Wm):
GR = k’ Wm
This means a proportional relationship between Wm and Wa and a parallel decline
of PNC and RGR at a rate of a–1. According to these theories, PNC can be
predicted from RGR (Greenwood et al., 1991).

Proportionality between PNC and RGR implies that PNC will decline parallel to
RGR in relation to W according to the power function based on the theories of
Caloin & Yu (1982) and Caloin & Yu (1984). It has been shown that the GR of
many organisms, plants as well as animals, relates to W according to the 3/4 –

16
Power Law (Niklas, 1994), i.e. the scaling exponent a = 0.75. A similar
relationship exists between surface area and biomass of cells and organisms
according to the 2/3-Power Law (Niklas, 1994). Hardwick (1987) derived a 2/3-
Power relationship for plant communities –the “core-skin” hypothesis. The
hypothesis is based on the same idea of two compartments as described by Caloin
& Yu (1984) with a “skin” of outer tissues engaged in energy exchange with the
environment and a “core” of structural and supportive inner tissues. Based on the
geometric relation between volume and periphery of a three-dimensional core
relative to its length, Hardwick (1987) derives a 2/3-power relationship between
“skin” and “core” and assumes that the amount of energetically active “skin” tissue
is allometrically proportional to the accumulated plant N (PN). The 2/3 power
relationship is in agreement with the value of the scaling exponent (0.63) in the
relation between PN and W by Plénet & Lemaire (1999):
PN = 34W0.63
Leaf area (LA) can also be assumed to be proportional to “skin” tissue which is
supported by the similar allometric scaling exponent b (equation 1) found by Plénet
& Lemaire (1999) for the leaf area index (LAI) as for PN:
LAI = 1.234W0.679
PN and LAI therefore scale with similar proportionality to W with the allometric
scaling exponent b making it possible to compare the rates of change between
variables that have different units (Niklas, 2006). A similar scaling exponent b,
therefore implies that LAI and PN will change at similar rates relative to W.

The similar values of the scaling exponents b in their relation to W indicates

proportionality between LA and PN with a constant PN on an area basis (PNCa).
PN is linearly related to LA but not to W (Grindlay et al., 1997). A linear
relationship between PN and LA has been shown for various crops such as wheat
(Sylvester-Bradley, 1990 and Olesen et al., 2002), tomatoes (Tei et al., 2002),
Brussels sprouts and leeks (Booij et al., 1996) and cabbage (Ekbladh et al., 2007).
Thus, the 1:1 relationship of LA and PN, during the exponential growth phase
(Glimskär & Ericsson, 1999), continues also after the exponential growth phase,
but not a 1:1 relationship of LA or PN to W.

Plant nitrogen is determined from, and therefore more conveniently expressed as

the plant nitrogen concentration. The relationship between PN (kg N ha-1) and W
(ton DM ha-1), PN = 34W0.63 (Plénet & Lemaire, 1999) can be converted to PNC
(equation 1) as
PNC = 34/10 x W0.63-1
PNC = 3.4W-0.37
An expression for leaf area ratio (LAR) is obtained by a similar conversion of the
expression of LA as a function of W (bLAR = bLA - 1).

The value of parameter a in equation (1), 3.4% for maize, represents the PNC of
young plants in their exponential growth phase when most of the above ground
plant biomass consists of photosynthetic tissue.

17
 The value of parameter b found for maize (-0.37) is similar to that found for tall
fescue by Lemaire & Salette (1984). The relationship for tall fescue was originally
introduced as an empirical relationship but it can be linked through parameter b to
the value of 2/3 suggested by the “core-skin” hypothesis. The allometric
relationship between LA and PN implies a corresponding allometric relationship
between LAR and PNC as bLAR = bLA – 1 and bPNC = bPN – 1 so there is a similar
proportionality between LAR and PNC as between LA and PN. However as
described above, Caloin & Yu (1982), Caloin & Yu (1984) and Greenwood et al.
(1991) suggested proportionality between PNC and RGR. Proportionality implies
similar rates of decline in both LAR and RGR relative to PNC. Similar rates would
imply a constant net assimilation rate (NAR) as RGR can be factorised into its
growth components LAR and NAR (Hunt, 1978). Moreover, in the case of a
constant PNCa, a constant NAR implies constant nitrogen productivity (NP =
GR/PN) because NAR is PNCa × NP. As mentioned above PNC and RGR are
linearly related with a constant slope (=NP) during the exponential growth phase
(Ingestad & Ågren, 1992). After the exponential growth phase, nitrogen
productivity was however expected to be reduced for example by self shading
(Ingestad & Ågren, 1992). So, LAR, NP and RGR will change with growth, and
the relative importance of their ontogenetic changes in relation to the ontogenetic
changes in PNC needs to be determined in the attempt to relate the ontogenetic
decline of PNC to any of them. It is the leaf nitrogen that is involved in biomass
production, so to distinguish between the productivity of the leaf nitrogen and
nitrogen allocation, nitrogen productivity can be factorised into leaf nitrogen
productivity (LNP = GR/LN; LN=leaf nitrogen) and leaf nitrogen ratio (LNR =
LN/PN) (Hirose, 1988). Then RGR can be factorised into:

RGR = LNP × LNR × PNC (3)

Moreover, NAR can be factorised into leaf nitrogen productivity and leaf nitrogen
concentration on an area basis (LNCa = LN/LA). The ontogenetic changes in LNCa
are of importance both for the relation of PNC to RGR and for plant analysis of
individual leaves as proposed by Lemaire et al. (1997).

Leaf area - nitrogen - mass relationships – individual leaves

Sampling whole plants is inconvenient in the case of crops with large plants such
as cabbage. Moreover, although the critical PNC can be well-defined in relation to
W, the problem of estimating W for samples from commercial fields remains. As
an alternative to sampling on whole plants, leaf area based assessment of leaf N
content can be used instead as proposed by Lemaire et al. (1997). LNCa has been
shown to remain constant during growth and development of the last visible collar
leaf of maize (Lemaire et al., 1997) and for new leaves appearing on top of the
canopy of lucerne (Lemaire et al., 1991). The constancy of LNCa of the top leaves
offers a possibility of overcoming the problem in plant analysis of a varying N
content according to the growth stage.

LNCa of the top leaves remains constant because they are exposed to a constant
light intensity and their LNCa is adjusted to the ambient light intensity (Grindlay,

18
1997). The successively emerging new leaves at the top of the canopy are exposed
to full light, to which the LNCa for these leaves will be adjusted (Lemaire et al.,
1997). The unshaded leaves at the top of the canopy would be relatively easy to
identify and to reach for sampling. For the sampling, an appropriate technique and
tool is needed to punch out leaf discs of a well-defined leaf area. Alternatively, if
the readings of chlorophyll meters can be calibrated to LNCa, instant determination
would be possible (Richardson et al., 2002); however uncertainty remains about
the constancy of the calibration equation over time. Good correlation, although not
entirely consistent between years, between Minolta SPAD-502 chlorophyll meter
readings and both tissue nitrate and total N found by Westerveld et al. (2003a)
shows that the chlorophyll meter can be used for cabbage with good results.

Possibilities for plant analysis based on critical N concentrations

derived from leaf area or growth
Plant analysis is mainly diagnostic as it retrospectively reflects the nutrient status.
It is a powerful tool for evaluation of fertilizer strategies and may serve as a
complement to prognostic methods with a predictive function. Farmers in
vegetable production are encouraged to use the Nmin-method or the pre-side nitrate
test-test for adjusting the N application rates to the soil mineral N supply. Adoption
by farmers has however been low and therefore research and development on
computer-based recommendation systems are in progress. Eurotate-N, based on
WELL-N and N_ABLE (Rahn et al., 1996; Greenwood, 2001), and N-Expert
(Fink & Scharpf, 1993; Fink & Feller, 1997) are the most important for vegetable
production. Evaluation of WELL_N showed promising results as the combination
of acceptable yields and low soil mineral N residues was achieved (Goodlass et al.,
1997). These methods have been developed to predict fertilizer requirements by a
precise matching of supply to demand. Trying to be precise may imply a certain
risk for underfertilising compared to if fertiliser is applied with a certain margin of
excess (Ekbladh, 2003) and therefore evaluation of the fertiliser strategy may be
needed. Thus, the diagnostic function of plant analysis provides a complement and
can be used together with prognostic methods. Organic vegetable production is
another example where evaluation of the nutrient management strategy is of great
importance. The strategy is often a combination of several practices, such as crop
rotation and use of organic fertilizers for each of which the nutrient effect is
difficult to predict. Plant analysis may serve as an indicator from which the
management techniques can successively be improved in a participatory learning
process, in which both farmers and advisers are engaged.

Plant analysis may advance by deriving critical concentrations according to the

principles outlined above, so as to overcome some of the problems and drawbacks
stated above:
• Growth stage: The most important advantage expected is that a continuously
defined N concentration with growth eliminates the difficulty in defining
specific growth stages to which the sampling date have to match. This
applies both to analysis of the whole plant and of individual leaves.

19
• Plant part: According to the theory of the relationship between LNCa and light
intensity, LNCa remains constant for unshaded leaves on top of the canopy.
Sampling of top leaves is in agreement with earlier praxis.
• Differences between crops and species: The theories refer basically to general
principles for the relation between growth and N; crop specific deviations
from the theories need to be further explored.
• Environmental variability: Relating N concentration to growth instead of to time
removed much of the variability between years. Moreover agreement has
been found between critical concentrations determined within regions with
similar climate and favourable growth conditions.
• Nutrient interactions: Nutrient interactions occur independently of method used.
However, in contrast to soil analysis, multi-element analysis offers the
possibility of showing the nutrient composition experienced by the plant.
• Total N of leaves gives less variability compared to petiole nitrate.

Aims and hypothesis

The overall aim of this work was to advance plant analysis as a tool for evaluation
of the crop N status by establishing critical N concentrations based on a) the
relationship between plant nitrogen concentration (PNC) and weight per unit
ground area of plant dry matter (W) for whole plants, and b) a constant leaf
nitrogen on an area basis (LNCa) for leaves exposed to constant light intensity for
analysis of individual leaves. The work builds on the theoretical frameworks
proposed in the literature that describe the relation between N and growth.
Specifically the study was intended to

I: compare the rate of ontogenetic decline in PNC with the rates of decline in
relative growth rate (RGR) and the growth components of RGR (leaf area ratio
(LAR), leaf nitrogen productivity (LNP), leaf nitrogen ratio (LNR) and net
assimilation rate (NAR) with the aim of finding a predictor of the rate of
decline in PNC;
II: estimate the PNCc for white cabbage, including the pre-linear growth phases.
III: relate the rate of decline in the derived PNCc to the rate of decline in LAR at
ample N supply;
IV: examine if a constant leaf area based value of the critical N concentration can
be used for unshaded leaves at top of the canopy.

The main hypotheses of this study were that during growth and development,
LNCa of unshaded leaves at top of the canopy as well as whole plant PNCa remain
constant and that as a consequence of a constant PNCa, LAR and PNC will decline
at similar rates relative to W. An alternative hypothesis was that PNC declines at a
similar rate as RGR.

20
Materials and Methods
The relationships between N and growth components of white cabbage were
examined in both field experiments (Paper I and IV) and in a climate chamber
(Paper II and III). The time-course of change in N concentrations and in growth
components was examined by repeated samplings from seedling to mature plant
during the entire growth cycle, both for individual leaves and for the whole plant.
Data from the field experiments were used to describe the ontogenetic decline in
PNCc and to relate it to LAR under ample N-supply. Experiments were conducted
in climate chamber, to closely examine the relation between growth and N without
interference from fluctuations in N supply or in other environmental conditions.

Ontogenetic changes were compared by means of scaling relationships between

various growth components and variables derived from the measured variables of
leaf area, N and weight. Variables over different scales can be related by scaling
relationships (Wright & Westoby, 2001; Niklas, 2006). The scaling exponent b in
the power function y=axb or the scaling coefficient b in the linear regression log(y)
= log(a) + b µ log(x) show the proportional relationship between changes in the
variables x and y, and where x and y can be of different units (scales) as in
allometric relationships. The studied variables (y) were all related to W, i.e. (x) and
the rate of their ontogenetic changes relative to W were described by the
coefficient b. The variables compared were leaf area (LA) and accumulated plant
nitrogen (PN) and the growth components leaf area ratio (LAR), leaf nitrogen on
an area basis (LNCa), leaf nitrogen productivity (LNP), leaf nitrogen ratio (LNR),
net assimilation rate (NAR), plant nitrogen concentration (PNC) and relative
growth rate (RGR). A similar rate of decline in PNC and another variable indicated
a relationship although necessarily not a causal one. In addition, for direct
comparisons with PNC, the scaling relationships were used to show how the data
fit to the theories.

The relationship between PNC and growth components during the exponential
growth phase was studied on plants grown in a spray-based flowing solution
system (Ingestad & Lund, 1986) by which nutrient additions were automatised.
Nutrients were added to an extent just compensating for nutrients removed from
the culture solution. Nutrients were kept at a very low concentration and were
replenished by small nutrient additions at short intervals, thus minimising “luxury”
consumption. A major advantage of this technique is that the plant regulates its
own nutrient demand and it avoids effects of fluctuating N concentrations in the
plants caused by an N supply rate that does not closely match N demand, as occurs
in the field or in traditional nutrient film or pot-growing techniques. This technique
of growing plants at optimum conditions offers the possibility of carefully
examining relations of PNC to W within the theoretical framework described
above.

The growth device could only accommodate very small plants (exponential
phase). For the growth phases thereafter, the principle of growing plants in large
volumes of culture solution whereby the nutrient concentration of the solution

21
could be maintained constant, as the uptake is negligible relative to the amounts
available in the culture solution, was used (Asher et al., 1965). In both of the
systems, plant nutrients are available at free access. The ontogenetic changes are
thereby controlled by the plant itself. The alternative would be to control nutrient
availability by applied rates. Then, however, there would be the risk of imposing
changes in nutrient status which could interfere with the ontogenetic changes.
The PNCc was estimated from growth response curves for N supply ranging from
limitation to excess. Separate response curves were used for the pre-linear and
linear growth phases. For the pre-linear growth phase response functions
(piecewise regression and inverse polynomial regression) of PNC against W were
used. For the linear growth phase a response curve was used that related growth
rate (GR) to the parameters a and b of equation 1. The critical N concentration was
related to 95% of the maximal GR (Olfs, 2005). Details are given in Paper IV and
in Figure 3 in Paper IV. For determination of critical N concentrations data from
T2001, T2002a and T2002b were used.

In all experiments white cabbage (Brassica oleracea L. var. capitata L. f. alba

D.C. cv. Heckla F1, Svalöf Weibull AB, Hammenhög, Sweden) was used. The
field experiments were carried out in SLU, Ultuna Horticultural Research Station,
Uppsala (N59±49’, E17±39’) (Paper I) and Torslunda Experimental Station
(N56±38´, E16±31´) (Paper IV). The soil at Ultuna was a clay loam and at
Torslunda a loamy sand. Irrigation was applied according to a deficit balance of
precipitation and evapotranspiration. Between and in-row plant spacings were 0.6
and 0.5 m, respectively. Plant protection measures and weed control were carried
out as needed. Full supply of nutrients other than N (P, K, Mg, S and
micronutrients) was applied to all treatments (N-levels). In total, seven experiments
were performed as follows:

Ultuna 2000 (U2000) (Paper I)

Multi-harvest (0, 7, 14, 22, 35, 49, 63, 77, 97 and 114 days after transplanting,
DAT) and multi-N-level (0, 25, 50, 100, 150, 200 and 250 kg N ha-1) field
experiment. N was applied as Ca(NO3)2 with ⅛ applied at transplanting (0 DAT),
¼ 23 DAT, ⅜ 50 DAT and the remaining ¼ of the N applied 78 DAT. The
experiment had a split-plot design with N rate in the main plots and harvest day in
sub-plots with three replicates in complete blocks. Four plants were harvested from
each sub-plot and combined to provide one sample from each treatment and
block/replicate. Determination of leaf area, total N and dry weight.

Torslunda 2001 (T2001) (Paper IV)

Multi-harvest (0, 10, 18, 26, 38, 52, 67, 84, 101 and 130, DAT) and multi-N-level
field experiment (0, 50, 100, 150, 225, 300 and 375 kg N ha-1). Fertilizer N
applications were split at 11, 15, 23, 23 and 28% of the total N rate and applied
immediately after harvests at 0, 27, 53, 67 and 94 DAT. The treatments were
arranged in a split-plot design with N rate in the main plots and harvest day in sub-
plots with three complete replicate blocks. At each harvest day 8 plants were

22
harvested and combined to provide one sample from each treatment and
block/replicate. Determination of total N and dry weight.

Ultuna 2001 (U2001) (not presented elsewhere)

Multi-harvest (7, 14, 21, 35, and 49 DAT) and multi-N-level pot experiment.
Chlorophyll content meter readings (CCM200, ADC BioScientific Ltd,
Hoddesdon, Herts, UK). Determination of leaf area, total N and dry weight. Data
from this experiment were used to study how chlorophyll content meter readings
vary for different N-application rates for different individual leaves at 14, 35 and
49 DAT and to study the effect of varying N supply rates on LNCa at 49 DAT.

Torslunda (T2002a) (Paper IV)

Multi-N-level field experiment for the pre-linear growth phase. The experiment
consisted of seven different rates of N supply with three replicate blocks and was
harvested during the exponential growth phase (30 DAT). Number of plants
sampled, and design of experiment were as described for experiment T2001.
Fertigation was used for nutrient supply to ensure a rapid, direct transport of
nutrients to the roots. Determination of total N and dry weight at 30 DAT.

Torslunda (T2002b) (Paper IV)

Multi-harvest (1, 20, 26, 33, 40, 47, 54, 61, 68, 75, 82, 96, 110 and 134 DAT) field
experiment at only one application rate: ample N supply. The amount applied was
determined from the expected N uptake and from weekly soil mineral N samplings
and analyses. The weekly demand was calculated on basis of maintaining a buffer
of 50 kg N ha–1 in the soil, the expected weekly uptake (based on data from
experiment T2001) and the actual soil mineral N measured two days before the
next application. At each sampling time 14 plants were harvested and combined to
one sample from each of the two replicate blocks. Determination of leaf area, total
N and dry weight.

The Biotron (B2004) (Paper II)

Multi-harvest (5.9, 10.4, 11.9, 13.9, 17.2, 20.9 and 24.0 DAT) experiment
conducted during the exponential growth phase in growth units (Ingestad & Lund,
1986) in a climate chamber (the Biotron, Alnarp) with one level of nutrient supply
at free access. Determination of leaf area of individual leaves, nitrate-N, total N
and dry weight.

The Phytotron (Ph2004) (Paper II and III)

Multi-harvest (1, 21, 29, 36, 41, 54, 62, 76, 90 and 104 DAT) pot experiment
during the post-exponential growth phases in a flowing solution system in a climate
chamber (the Phytotron, Ultuna, Uppsala) with one level of nutrient supply at free
access. Determination of leaf area of individual leaves, chlorophyll, photon flux

23
density (PFD) distribution within the canopy, neutral detergent fibre (NDF) nitrate-
N, total N and dry weight.

Results and discussion

Growth – determination of growth rate and relative growth rate

Growth analysis (Hunt, 1978) was the major tool used to relate N to growth by the
means of growth components. GR and RGR were used in the growth analysis and
were estimated by linear or non linear regressions of W (or lnW) against DAT.
Growth was characterised by four growth phases. Growth was exponential (lnW to
DAT, R2 = 0.999) for the plants grown in the growth device of Ingestad & Lund
(1986) (Figure 1). W increased from 4.6 mg of the seedling (germ) to 2 g per plant
at 24 DAT. The next phase, the approximately exponential one, was characterised
by a good non linear fit of W to DAT (R2 = 0.997) but a gradual decline in the
slope of the linear regression of lnW to DAT resulting in a poor fit of the linear
regression (R2 = 0.93) (Figure 1). During this phase, W increased to around 50 g.
After a short transition phase, growth was linear in all experiments (Figure 1c and
Figure 1 in Paper IV). The very last sampling date at all field sites was not
included in the linear regression as GR declined. W at final harvest was slightly
more than 600 g, similar for plants grown in the climate chamber and the maximal
yield in the field (T2002b) and corresponded to 20.6 t ha-1. The plants in the
climate chamber grew faster (GR = 9.5 g plant-1 day-1; Figure 1) compared to those
in the field (GR = 6.2 g plant-1 day-1; Figure 1 in Paper IV). The cabbage plant of
cultivar Heckla developed around 35 green leaves, the leaves developed thereafter
formed the head.

Whole plant critical N

The critical N concentration on a whole plant basis (PNCc) for white cabbage cv.
Heckla was estimated as (Paper IV):
PNCc = 4.5 (% of DM) W < 1.5 t ha-1
PNCc = 5.1W-0.33 (% of DM) W > 1.5 t ha-1 (eq. 1)
The PNCc for W < 1.5 t ha-1 was estimated at 30 DAT. The value of 4.5% was
calculated using an inverse polynomial, whereas piecewise regression estimated the
value to 4.7 with the confidence interval of 4.60–4.74% (Figure 2 in Paper IV).
The PNCc for W > 1.5 t ha-1 was calculated from GR and parameters a and b of
equation 1 during the linear growth phase and the confidence intervals were for ac
= 4.71–5.62 and for bc = (–0.30) – (–0.36) based on the regressions of parameters a
and b against the average rate of daily accumulated plant N (PNR) (Paper IV).

24
 (a) (b)
8 0.4
Biotron experiment
6
0.3
4
Phytotron experiment
0.2
2

RGR
lnW

0
0.1

-2
0.0
-4

-6 -0.1
0 20 40 60 80 100 120 0 20 40 60 80 100 120
Days after transplanting, DAT Days after transplanting, DAT
(c)
a):
700 lnW exponential growth phase
2
lnW = -6.25+0.30*DAT; R =0.999, P<0.001
Dry weight, W (g plant )

600
-1

lnW approx. exp. and linear growth phases

2 3
500 lnW = -5.15+0.37*DAT-0.0042*DAT +0.170e-4*DAT
2
b): R =0.999
400
-1
300 Exponential growth phase RGR = 0.30 day
Approximately exponential growth phase
200 RGR = 0.37+2*-0.0042*DAT+3*0.170e-4*DAT
2

100 Linear growth phase RGR = 9.5/(-381+9.5*DAT)

c):
0
W exponential growth phase
0 20 40 60 80 100 120 W approx. exp. and linear growth phases
W = exp(lnW)
Days after transplanting, DAT W = -381+9.5*DAT, linear growth phase
2
R =0.9996, P<0.001

Figure 1. Growth and determination of RGR: a) lnW against DAT, b) RGR = d(lnW)/dt
(slope of lnW) and c) growth curves: W = exp(lnW) (1-104 DAT) and W = a + GR × DAT
(54-104 DAT). Exponential phase: Biotron experiment (up triangle); approximately
exponential and linear growth phases: Phytotron experiment (circle). Error bars denote ±SE.
(=Figure 1 in paper II).

The flat slope of parameter a against average rate of daily accumulated plant N
(Figure 3b in Paper IV) implies an uncertainty in parameter ac. The flat slope was
most likely caused by the weak response to fertiliser N at the beginning of the
growth period. Response of young plants to soil mineral N is not always to be
expected (Binford, 1992). The recovery of fertiliser is low during early growth,
especially for row-grown transplanted vegetables (Greenwood et al., 1989). In this
study when applied as a nutrient solution at a very wide range of N rates PNC of
field grown young plants responded to different N supply rates (experiment
T2002a, Paper IV).

25
 (a) (b)
-1
300 0 kg N ha 4 LAI 250
-1 0.68
250 100 kg N ha LAI=0.78×W

PN (kg N ha )

PN (kg N ha )
-1
200

-1
-1
250 kg N ha 3 2
R =0.99
200

LAI
150 2 150

100 1 100
PN
50 0.66
0 PN=43.8×W 50
0 2
R =0.99
-50 0
0 1 2 3 4 5 0 2 4 6 -1 8 10
LAI W (t ha )

Figure 2. a) Plant nitrogen (PN) versus leaf area index (LAI). Data from N application rate
0 (PN = 42.3 × LAI + 0.27, R2 = 0.99), 100 (PN = 54.4 × LAI – 0.36, R2 = 0.99) and 250
kg N ha-1 (PN = 64.4 × LAI – 1.42, R2 = 0.99) are shown for illustrative purpose; b) Leaf
area index (LAI) and plant nitrogen (PN) versus dry weight of plant. Data from N supply
rate 100 kg N ha-1 and DAT 1–97. Error bars indicate ±SE. Data from the U2000
experiment (=Figure 3 in Paper I).

(a) (b)
1.6
5 11
Leaf area index, LAI

1.4 PNCac
Critical PNCac (g m )
-2

10
Leaf area, LA (m )

Parameter aPNC c
2

1.2 4
9 Parameter aLAR
1.0
3 8
0.8 7
0.6 2 6
0,67
0.4 LA=0.32*W ;
2
5
1
0.2 R =0.993 4
0.0 0 3

0 5 10 15 20 20 40 60 80 100 120 140

-1
(c) Dry weight, W (t ha ) Days after transplanting, DAT
Plant nitrogen, PN (kg N ha )
-1

600
PN
500 PNc Linear growth phase (exp. T2002b):
> 40 DAT
400
LAI > 2.0
2
300 LA > 0.6 m
-1
200 W > 2.2 t ha
100 PN = 171.1LA-8.5;
2
0 R =0.964

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6

2
Leaf area, LA (m )

Figure 3. Relations to leaf area (LA) and leaf area index (LAI). a) LA against W; b) Critical
plant nitrogen conc. on an area basis (PNCac), dotted line shows the ratio aPNCc/aLAR; c)
Plant nitrogen PN and PNc against LA. Measured data from experiment T2002b. Error bars
denote ≤SE. (=Figure 5 in Paper IV).

26
Plant N per unit leaf area (PNCa)
The results from all experiments supported the main hypothesis of a constant
PNCa, except in cases where luxury consumption was suspected. PN was linearly
related to LA in all of the three experiments where LA was measured (Figure 2;
Figure 2a in Paper II; Figure 3), which is in agreement with linear relationship
between PN and LA for various crops as described above. PNCa remains constant
as the rate of increase of mass with structural functions is made up for by the lower
concentration of that mass (Grindlay et al., 1997). To examine the time-course of
PNCa in more detail during growth and in order to relate PNCa to changes in
growth components, PNCa was calculated for each sampling date (PNCa = PN/LA).
Under field conditions (U2000), PNCa remained constant with growth, expect for
the highest N supply rate (Figure 4). However, PNCa for plants grown in the
climate chamber increased (Figure 2b in Paper II), most likely because of luxury
uptake of N. The critical PNCa, PNCac, was calculated to be 4.7 g N m-2 (total N)
by dividing parameter aPNCc (5.1) by parameter aLAR (0.011) (Figure 3).

Ontogenetic changes in PNC and in growth components

During the exponential growth phase, not only LA and PN, but also W increased at
a close to 1:1 relationship to LA and PN, as shown for grasses by Glimskär &
Ericsson (1999). The scaling coefficients for LA and PN relative to W were very
close to 1 (Figure 3 in Paper II), which implies constant LAR, PNC and PNCa. The

10

7
PNCa (g N m )
2

-2
3 0 kg N ha
-2
2
100 kg N ha
-2
250 kg N ha
1

0 20 40 60 80 100 120 140

Days after transplanting, DAT

Figure 4. Accumulated plant N per unit leaf area (PNCa). Data from the
experiment U2000. Error bars show ≤SE.

27
constant RGR (Figure 1) implies a 1:1 relationship of GR to W, so that NAR
would also have been constant. The constant relationship between PNC and RGR
means a constant nitrogen productivity, which is the base for theory of N nutrition
and growth proposed by Ingestad & Ågren (1992). Nevertheless, in spite of the 1:1
relationship between LA and W, the slope of decrease in LAR (linear regression of
logLAR against logW, 10–24 DAT) was significantly different from zero
(p=0.0094) (Figure 5a). Thus, LAR declined slightly in spite of the almost 1:1
relationship of LA to W. And as LAR is a growth component of RGR, the constant
RGR during the exponential growth phase (0.30 day-1) was most likely an
approximation in spite of the very good fit of lnW to lnDAT (Figure 1a). Smolders
(1991) also showed by means of accurate non-destructive measurements that RGR
declines slightly during the exponential growth phase of spinach. During the first
sampling interval in the Ph2004 experiment (21–29 DAT), LAR still declined at
similar rate as during the exponential phase in the Biotron experiment (10–21

(a) (b)

Plant nitrogen concentration, PNC (%)

200 7 *
180 12 21
160
10 14 29* 29
Leaf area ratio, LAR (cm g )

21
17 -0.040 6 21
-1

140 36 0.0084 36
41
*
-0.047 -0.20 *
41 0.0042 -0.053 54
120
2

100 -0.17 54* 5 10 12

14 21 * 62
* 17 -0.25
-0.38 62 -0.0025 *
80 *76
-0.72 4 0.017 -0.073 90
104
60 76

0.95 90
3 PNCorg Biotron experiment -0.16
40 *
104 PNCorg Phytotron experiment
Phytotron experiment PNCtot Biotron experiment
Biotron experiment PNCtot Phytotron experiment
20 2
0.01 0.1 1 10 100 1000 0.01 0.1 1 10 100 1000
Leaf nitrogen productivity, LNP (g day g )

(c) (d)
-1

Net assimilation rate, NAR (g day m )

-2

10 30 0.9 21
-1

-0.03017 21
Leaf nitrogen ratio, LNR (fraction)

9 21 10 * 29
-1

0.013 36 41
8 10
12 14 20 0.8
7 -0.029 *
0.047 54
6 0.7 -0.11 62
-0.10 *
5 21 10
9
4 -0.12 *29* * -0.28 8 0.6
7
* -0.62 6 -0.51
3
36 -0.28 5
76
-0.17 41 * 0.5
*
* 54 -0.047 4
90
-0.67
2 NAR Biotron experiment *62 3
NAR Phytotron experiment 0.4 104
-0.39 76 2
LNP Biotron experiment 90 Biotron experiment
LNP Phytrotron experiment 104 Phytotron experiment
1 1 0.3
0.01 0.1 1 10 100 1000 0.01 0.1 1 10 100 1000

-1
Dry weight, W (g plant )

Figure 5. Time-course of change in a) leaf area ratio (LAR), b) organic and total plant
nitrogen concentration (PNCorg, PNCtot), c) leaf nitrogen productivity (LNP) and net
assimilation rate (NAR) and d) leaf nitrogen ratio (LNR). The lines are the log-log linear
regressions slopes and the figure in italics adjacent to the line indicates the slope (scaling
coefficient, bi, eq. 4 in Paper II). Smaller figures adjacent to data points denote days after
transplanting (DAT). An asterisk at a data point indicates that the two adjacent slopes are
statistically different (P < 0.05). Error bars denote ±SE. (Figure 4 in Paper II).

28
DAT) (Figure 5a). According to the hypothesis of a constant PNCa, PNC can also
be expected to decline at a similar rate as LAR during the exponential phase.
However, in the Phytotron experiment PNCa increased because of suspected luxury
consumption and a decline in PNC, as in LAR, could not be confirmed by the data.
PNC was almost constant during the exponential phase, in the Biotron experiment
and up to 29 DAT in the Phytotron experiment (Figure 5b).

During the approximately exponential growth phase, there was no longer a 1:1
relationship of LA and PN to W (Figure 3 in Paper II), neither was there of GR to
W as RGR declined (Figure 1b). PNC and RGR declined at different rates, –0.07
and –0.30 (Figures 5a and 5c). The rate of decline in RGR is the sum of the scaling
coefficients of LAR (–0.20) and NAR (–0.10). Their different rate of decline
means in terms of the theory by Ingestad & Ågren (1992) that nitrogen productivity
declined. Leaf nitrogen productivity was responsible for the major part of the
difference in rates of decline between PNC and RGR as leaf nitrogen ratio was
nearly constant during the approximately exponential growth phase and in the
beginning of the linear growth phase (Figure 5 in Paper I; Figure 5b, 5c and 5d).
However, as shown for field data, the 1:1 relationship of PN to LA remained, as
PNCa was constant and LA and PN increased at similar rates relative to W (Figure
2b). Thus, the constant relationship between RGR and PNC during the exponential
phase changes to a constant relationship between PNC and only one of the growth
components of RGR, LAR. The difference between the rates of decline in PNC and
RGR were due to changes in NAR and nitrogen productivity (NP). NAR and
nitrogen productivity will change at similar rates as long as PNCa is constant (NAR
= PNCa × NP). The deviations between NAR and the sum of leaf nitrogen
productivity and leaf nitrogen ratio shown for the rates of decline for the Phytotron
data were related to the fact that PNCa increased, contrary to what was
hypothesised (Figure 4c and 4d in Paper II). For the same reason the rates of
decline in LAR and PNC differed. The changing relationships between GR, LA,
PN and W between the exponential and the approximately exponential growth
phases were reflected in the change to a steeper rate of decline for LAR and PNCorg
(Figure 5a and 5b). The scaling coefficient for LAR was -0.04 during the
exponential phase and -0.20 during the approximately exponential phase. Thus,
one level of rate of decline was associated with the exponential growth phase and
another rate with the approximately exponential growth phase, rather than a
gradually decreasing rate of decline. However, the critical N concentration for
field-grown young plants was approximated by a constant value of 4.5% during the
whole pre-linear growth period because the rate of decline for young plants could
not be estimated under field conditions.

During the linear growth phase, PNCa was constant in the field (Figure 4) so that
LAR and PNC declined at similar rates, bLAR = -0.32 and bPNC = -0.31. These
values were estimated from regressions of LA and PN against W for the entire
growth period at U2000, 1–97 DAT (Figure 2b). They agreed with the value for
critical PNC, bPNCc = -0.33 (Paper IV) and for LAR at ample supply, bLAR =
-0.33. The latter value was estimated from the regression of LA against W up to
62 DAT as bLA -1 = 0.67 - 1, up to 62 DAT. After 62 DAT this relationship did
not hold as leaf extension ceased. The rate of decline in PNCc during the entire

29
linear growth phase could thus be predicted from bLA up to 62 DAT. Also, in the
Phytotron experiment PNC and LAR declined at similar rates (bPNCorg = -0.16 and
bLA -1 = 0.85 - 1 = -0.15) up to the time when the extension of LA ceased at 62
DAT (Figure 5b and Figure 3 in Paper II). Both experiments therefore supported
the hypothesis of similar rates of decline for LAR and PNC. Direct regression on
data of LAR and PNC against W gave more variable values of bLAR and bPNC and
larger confidence intervals. Direct regression analysis was not possible for climate
chamber data as only data from one sampling interval during the short period of the
linear growth phase before heading were available. During heading LAR declined
more sharply so that LAR and PNC declined at different rates. PN increased
because of growth of the head whereas LA extension ceased. LAR seemed to
gradually become more and more responsible for the decline in RGR and finally
declined at a rate of -1 whereas the decline in NAR ceased. RGR declined at a rate
of -1 because GR was constant: RGR = GR/W = GR * W-1. Because of the
increasing PNCa, NAR and nitrogen productivity changed at different rates. NAR
was constant because GR and LA were constant (Figure 5c), whereas leaf nitrogen
productivity decreased because of increasing leaf thickness (Figure 5c) and leaf
nitrogen ratio decreased mainly because of head growth (Figure 5d). Thus, none of
the relationships shown during the vegetative phase before heading held during
heading. Therefore, LAR and PNC can only be expected to be related during
vegetative growth as long as leaf extension and mass develop proportionally and N
is not allocated to other plant organs. In spite of this, PNC continued to decline at
the same rate during the entire linear growth phase. In summary, the results
therefore show that the rate of decline in PNCc could be predicted from the rate of
increase in LA with ample N supply.

The rate of increase in LA, bLA = 0.68 across all N rates in U2000 (1–97 DAT)
(Figure 2) and bLA = 0.67 with ample N supply in T2002b (1–62 DAT) (Figure 3a )
during the linear growth phase agree very well with the 2/3-Power Rule (Niklas,
1994) or “core-skin” hypothesis (Hardwick, 1987) which postulates that plants and
plant communities that optimise the utilisation of the available radiation for growth
will increase their PN in proportion to W²•³ in a dense canopy. This applies to LA
as well. LA and PN will increase allometrically at the same rate (the same scaling
exponent) relative to W (Figure 3 in Paper I), because of the constant PNCa. The
“core-skin” hypothesis was forwarded to supplement the self-thinning rule, which
states that the increase in mean mass per plant is related to a decrease in the
number of plants in a dense plant community (Westoby, 1984). The energy
available for growth is limited by the amount of light intercepted per ground area
(Donald, 1961). When a plant dies, its share of energy flux can be used by
neighbouring plants. Although the self-thinning rule only applies to dense canopies
and not to row-grown crops such as cabbage, the growth of row-grown crops is
limited as well by the incident light flux per unit ground area, which estimates the
maximal daily growth rate that can be reached during the linear growth phase. The
basic idea of the core-skin hypothesis, that periphery (LA and PN) relates to the
length squared of a three dimensional core and the volume (proportional to W) to
the length cubed, apparently also applies to cabbage although the leaves to a great
extent were responsible for the increase in mass (Paper II), rather than the stem.
When growth is limited by light, LA and PN in cabbage allometrically relate to W

30
by the proportionality factor 2/3, as periphery relates to volume, as proposed by
Hardwick (1987).

The rates of decline in PNC were different in the climate chamber compared
with those in the field. In the climate chamber, the average rate of decline in
PNCorg during the linear growth phase was –0.16, which was similar to the rate of
decline in LAR during the approximately exponential growth phase. This is in
contrast to PNC of plants grown in a closed canopy (i.e. linear growth phase) in the
field. For the latter, LAR and PNC declined in agreement with the core-skin
hypothesis at a rate of –1/3. The rate of decline of LAR and PNC in the climate
chamber was halfway between when LAR and PNC were close to constant (zero
decline rate) and when they were –1/3 for the closed canopy, i. e. approximately –
1/6. Therefore, results from the climate chamber cannot directly be applied to field
conditions. The reason for the difference between the intermediate rate of decline
and the rate of decline of –1/3 for cabbage plants grown in the climate chamber
and cabbage grown in the field was mainly the difference in leaf size and thickness,
with the former being larger and thinner. The mass fractions were similar for
cabbage in the three experiments for which plant parts were weighed separately
(U2000, T2002a, Ph2004, data not shown) so that the difference in rate of decline
can only be attributed to the difference in specific leaf area (SLA). Broccoli grown
at different plant densities by Francesangeli et al. (2006) behaved in a similar
manner – mass fractions were unaffected whereas LAR increased at higher plant
densities whereby the degree of shading increased. The larger LA of the cabbage in
the climate chamber resulted in a lower rate of decline in LAR compared to the
decline under field conditions and was most likely caused by the lower light
intensity in the climate chamber and because the plants were grown as isolated
plants. Rates intermediate between 2/3 and unity have also been shown,in other
studies, such as for example 0.84 and 0.88 for PN ∂ W for lucerne grown in a
greenhouse at low plant densities and at high densities under field conditions,
respectively (Lemaire et al., 2005). Values of 0.76–0.79 were found for the scaling
exponent a of the relation between metabolic tissues and total biomass
Wm = k ' ( W ) α proposed by Caloin & Yu (1982). These values were based on
experiments with Dactylis glomerata in a climate chamber with light intensities of
35–85 W m-2. The values for both of the experiments are closer to the 3/4-Power
Rule by Niklas (1994) or to 5/6 instead of to the 2/3-Power Rule. It is evident that
the allometric scaling exponent for LA and PN increases at a lower rate (2/3)
within the constraints of a canopy compared to plants grown at low plant densities
or at low light intensities. This applies to plants with very different morphology
such as grasses and cabbage.

Effect of self shading on the ontogenetic decline in PNC

The increasing difference in rate of decline between PNC and RGR was, as shown
above, attributed to decline in the nitrogen productivity and specifically in the leaf
nitrogen productivity. After the closure of the canopy (the linear growth phase) GR
was constant, which resulted in a decreased NAR as LA still increased at the very
beginning of the linear growth phase. Leaf nitrogen productivity and NAR declined
rapidly, with scaling coefficients of -0.62 and -0.67 respectively, compared to the

31
rate of decline of -0.16 in PNCorg (Figure 5b and 5c). Similar rates of decline were
found for leaf nitrogen productivity and NAR under field conditions during the
first part of the linear growth phase (Figure 5 in Paper I). The decline in leaf
nitrogen productivity and NAR was preceded by a decline in the average amount of
photon flux density (PFD) incident on the LA of the cabbage (Figure 5 in Paper II).
NAR is known to be closely related to light intensity (Blackman & Wilson, 1951;
McDonald et al., 1992). The decline in leaf nitrogen productivity and NAR was
therefore most likely caused by self shading. The hypothesis by Caloin & Yu
(1984) related the ontogenetic decline in PNC to the ontogenetic decline in RGR
which was based on an assumption that GR is proportional to the amount of
metabolic tissue in the plant (Wm). The proportionality constant k was assumed to
be constant for a given set of environmental conditions. But the conditions within
the canopy change with increasing self shading. The proportionality constant k is
not constant when self shading occurs and PNC and RGR are then not related,
which confirms the reservation by Lemaire & Gastal (1997) of the constancy of
constant k in a dense canopy. It is most likely that the metabolic tissue increases in
proportion to the increasing leaf area. GR responds more directly to changes in
light intensity than Wm. A relation of Wm to LA rather than to GR is therefore more
likely and it can be shown that PNC relates to LAR rather than to RGR by
changing the assumption of a constant GR/Wm to that of a constant Wm/LA
(appendix B). The rapid response of NAR to self-shading, but no corresponding
response of PNC is in agreement with PNC for the whole plant of potato, for which
PNC, measured 70 days after shade treatment, was not affected by partial shading
(Vos & Putten, 2001). In the study by Vos & Putten (2001), LNC of the shaded
leaves was only affected at 90% shading but not at 50%, so that very large light
reductions are needed before LNC is affected. A minor effect on PNC only at the
lowest of the light intensity treatments during the exponential growth phase of crop
was also found by Ingestad & McDonald (1989), whereas LNCa but not LNC of
lettuce was affected by shading (De Pinheiro Henrique & Marcelis, 2000). The
leaves became thinner (higher SLA) due to shading. So, self shading affects N on
an area basis but not N on a basis of mass. Thus, the ontogenetic rate of decline in
leaf nitrogen productivity was affected by self shading, whereas PNC was not
directly affected.

Leaf N on an area basis (LNCa)

The average LNCa of the field-grown cabbage declined with growth (Figure 5 in
Paper 1). For cabbage in the climate chamber, the average LNCa did not decline as
expected and LNCa of the unshaded leaves exposed to constant light was not
constant as expected but increased with growth, which was most likely due to
luxury consumption. Nevertheless, a vertical gradient in the LNCa of individual
leaves was developed downwards through the canopy parallel to a vertical gradient
in photon flux density (Figure 6a and 6b). The light is attenuated with increasing
canopy depth because of self shading (Monsi and Saeki, 2005) and LNCa is
adjusted to the incident photon flux density at the leaf surface to maximise
photosynthesis (Hirose & Werger, 1987). Maximal LNCa and photon flux density

32
 (a) (b)

Photon fluux density, PFD (µmole m s )

-1
Leaf nitrogen per leaf area, LNCa (g N m )

7.0 70
-2

-2
6.5
20,24 60
6.0 16
5.5 50
28 20
5.0 12 40
24
4.5 32
30 16
4.0
12
3.5 8 20

3.0 28
4 10
2.5
4 4 8, 32
2.0 8 0
12
20 30 40 50 60 70 80 90 100 110 120 20 30 40 50 60 70 80 90 100 110 120
16
Leaf nitrogen concentration, LNC (percent)

(c) (d)
20
350
24

Specific leaf area, SLA (cm g )

7

-1
28
300
32
6 250

200 4
8
5
12 150 8
16 100 12
4 20 16
24 20-32
28 50
32
3 0
20 30 40 50 60 70 80 90 100 110 120 20 30 40 50 60 70 80 90 100 110 120

Days after transplanting, DAT

Figure 6. a) Leaf nitrogen concentration on an area basis (LNCa), b) Photon flux density,
(PFD); c) Leaf nitrogen concentration on a weight basis (LNC) and d) specific leaf area
(SLA) all for leaves of positions 4, 8, 12, 16, 20, 24, 28 and 32 counted from the bottom of
the canopy. Error bars denote ≤SE. (=Figure 2 in paper III).

coincided at the same leaf positions (Figure 6a and 6b). The maximal LNCa and
photon flux density were reached for leaves at successively higher leaf positions
and finally at leaf position 20 (Figure 6a and 6b). These leaves were characterised
by being larger and more horizontally orientated than the more vertical orientated
leaves above them. From these leaves LNCa and photon flux density decreased in
parallel gradients downwards as well as upwards in the canopy. The decrease in
photon flux density at higher leaf positions was a result of their vertical orientation.
The decrease in LNCa was a consequence of changes in SLA and LNC (LNCa =
LNC/SLA). LNC and SLA were highest at the lowest leaf position (Figure 6c and
6d), LNC was high but the leaves were thin (high SLA) which leads to a low
LNCa. Even if LNC decreased upwards in the canopy the increasing thickness
(decreasing SLA) caused LNCa to increase. At higher leaf positions, above the leaf
positions with maximal LNCa, SLA was fairly constant with leaf position, whereas
LNC continued to decrease upwards, which led to decreasing LNCa upwards in the
canopy from the leaves with the maximal LNCa. Such an interaction between LNC
and SLA leading to a constant LNCa was shown by Charles-Edwards et al., (1987)
for forage sorghum. The pattern of LNCa between leaf positions is not clearly
shown for the early growth stages in Figure 6a. However, chlorophyll content
readings from the U2001 pot experiment at 14, 35 and 49 DAT showed a similar
pattern of the relative chlorophyll content between individual leaves with a

33
 (a) (b)
60
25
50

20

CCI
40

CCI
15 30

20
10
10

5
0
1 2 3 4 5 6 7 0 2 4 6 8 10 12 14 16

Leaf position Leaf position

(c)
80 N level
70
1
3
60
5
50 7
CCI

40 9
30

20

10

2 4 6 8 10 12 14 16 18 20 22

Leaf position

Figure 7. Chlorophyll content index (CCI) for various N application rates as a function of
leaf position at a) 14 days after transplantation (DAT) b) 35 DAT and c) 49 DAT. Data
from pot experiment U2001. Error bars show ≤SE.

maximum in the middle of the canopy (Figure 7). The fully developed horizontally
orientated unshaded leaves had maximal LNCa from young plant stage to a fully
developed canopy.

For each sampling date, LNCa values of those leaves with the highest LNCa,
were plotted against W (Figure 3a in Paper III). The LNCa of these leaves, that
were fully exposed to constant light, increased. A similar trend was shown for
chlorophyll (Figure 3b Paper III). The rate of increase in maximal LNCa was the
same as the rate of increase in PNCa, the scaling coefficient was 0.11 for both of
them. The ratio between LNCa for the leaves fully exposed to light and PNCa was
constant, around 0.8 for whole leaves and slightly lower for punched leaf discs
(Figure 4 in Paper III). For young plants, as long as all leaves are unshaded, the
LNCa can be expected to be similar. The ratio of LNCa of all leaves to PNCa (leaf
nitrogen ratio) of young plants will therefore be similar to the ratio of LNCa of
leaves fully exposed to light to PNCa during the continued growth. The results in
Paper II (Figure 4d) show that the leaf nitrogen ratio of young plants was indeed
close to 0.8, similar to the ratio of LNCa of leaves fully exposed to light to PNCa.
The leaf nitrogen ratio was similar in the field and in the climate chamber (Figure
8), which is in agreement with results of a study on lettuce grown at different light
intensity and at varying N supply by De Pinheiro Henriques & Marcelis (2000).
There is therefore strong evidence that the critical LNCa of leaves fully exposed to
light for field conditions can be predicted from the critical PNCa, PNCac, as 0.8 µ
PNCac = 0.8 µ 4.7 = 3.7 g N m-2. The hypothesis of a constant LNCa of unshaded
leaves was therefore supported.

34
 1.2

1.0
LNR (Fraction)

0.8

0.6
T2002
U2000
0.4 Ph2004 tot
Ph2004 org
0.2

0.1 1 10 100 1000

-1
Dry weight, W (g plant )

Figure 8. Leaf nitrogen ratio (LNR) of field experiment U2000 and T2002 and of the
experiment in the climate chamber Ph2004. Error bars show ≤SE.

The results from the field experiment at Ultuna in 2000 provide additional,
indirect evidence to support the conclusion that a critical LNCa can be predicted
from PNCac. Unfortunately LNC of individual leaves was not analysed from data
from the field experiments. Therefore, critical LNC of the youngest fully expanded
leaves (YFEL) of cabbage cv. Rampo was taken from Table 2 in Huett & Rose
(1989). These values of critical LNC were used to estimate critical LNCa of leaves
fully exposed to light under field conditions (Appendix B). It was assumed that the
maximal LNCa is reached at the same leaf positions both under field conditions and
in the climate chamber so that values of SLA from the T2002b experiment were
selected for the same leaf positions at which LNCa was maximal in the Ph2004
experiment. The estimated critical LNCa was 3.5 g N m-2 (Table 1) and, except for
deviations in weeks 8 and 12, remained constant over 10 weeks. The estimated
value of 3.5 g N m-2 was based on GR of the lowest N application rate (29 mmol
L-1) that gave the highest yield. The estimated LNCa was slightly lower than the
value derived from PNCac (3.7 g N m-2). The two values of critical LNCa agree
rather well taking into consideration that different cultivars were grown in a quite
different climate.

35
Table 1. LNCa for leaves fully exposed to light estimated from LNC of Huett & Rose (1989)
and SLA of T2002b at leaf positions, at which LNCa and PFD had their maximal values in
the Ph2004-experiment
Week W1 SLA2 LNC3 LNCa4
2 8.1 123.5 4.38 3.55
4 22.5 121.8 4.35 3.57
6 57.7 117.5 4.15 3.53
8 124.5 109.3 3.10 2.84
10 206.1 99.2 3.50 3.53
12 265.7 91.9 3.10 3.37
1
W (g plant-1), estimated from GR in Table 1 of Huett & Rose (1989), see appendix B
2
SLA (cm2 g-1), estimated from SLA data of T2002b as SLA=124.5 – 0.123 ä W, see
appendix B
3
LNC (%), from Table 2 in Huett & Rose (1989)
4
LNCa = LNC/SLA (g N m-2)

Applications
The results shown in this thesis indicate that assessing the plant N status on the
basis of analysis of leaves fully exposed to light as suggested by Lemaire et al.
(1997) is promising. Matching the actual sampling stage to a critical concentration
which precisely corresponds to that growth stage has hitherto been one of the
greatest difficulties in plant analysis. This obstacle is removed by leaf area based
plant analysis of individual leaves, because the critical concentration remains
constant during plant growth and development. Another difficulty has been to
choose which part of the plant to sample. For leaf area based analysis, the leaves
fully exposed to full light should be sampled. These leaves should be relatively
easy to identify and reach, however attention is needed at the sampling to ensure
the leaves selected are not shaded at any time during the day. Also the inherent
variability between adjacent leaf positions has to be further examined. LNCa has to
be sensitive to variations in N status. A plant adjusts its LA to the N supply in
order to maintain LNCa at a functional level for photosynthesis (Grindlay, 1997).
Therefore, it can be expected that LNCa will not respond to N supply. The
response of average LNCa of cabbage to N supply (Figure 9) and of LNCa of
unshaded maize top canopy leaves to N supply (Lemaire et al., 1997) show
however that LNCa is sensitive to changes in N supply. An appropriate tool has to
be developed, such as for example punching tongs for a convenient and rapid
sampling of leaf discs with a well-defined leaf area. Rapid sampling is required in
commercial practice, and sampling of plants is less time consuming compared to
extracting soil cores (Matthäus & Gysi, 2001). Another advantage of plant analysis
is that many elements can be diagnosed. Leaf area based plant analysis for
diagnostic evaluation of fertiliser strategies has several advantages and seems
promising together with the prognostic function of computer-based
recommendation systems.

Assessing plant N status on a whole plant basis by relating to W, according to the

critical curve derived, has the advantage that each value of PNCc is uniquely
related to a value of W. As with measuring on unshaded leaves, the problem of

36
 3.0 5.0

LNC
4.5
LNCa
2.5
4.0

LNC (%)
LNCa (g m )
-2

3.5
2.0

3.0

1.5 2.5

2.0
0 2 4 6 8 10
N supply rate

Figure 9. Dependence of leaf nitrogen on an area basis LNCa (average of canopy) and leaf
nitrogen concentration on a weight basis (LNC) on N status. Data at 49 DAT from pot
experiment U2001.

defining specific growth stages is avoided. The similar decline in PNC and LAR
means that LA measurements from only one level of ample N supply can be used
to predict the rate of decline in PNCc. Only a few repeated samplings are needed to
estimate PNCac from which the parameter ac for PNCc can be derived. If W cannot
be estimated, for example by sampling in commercial fields, it can be estimated
from the number of leaves developed, as was shown in Figure 6 in Paper IV. A
universal scale for defining growth stages, the BBCH-code, was developed jointly
by BASF, Bayer, Ciba-Geigy and Hoechst (Lancashire et al., 1991). The BBCH-
code suggests using the number of leaves developed and diameter of the head to
define the growth stage of cabbage (Meier, 1997). Although sampling of whole
plants may be of limited use for routine sampling, at least for larger plants, the
method is essential in the assessment of plant N status in research and development
work. PNCc is necessary for calculating the optimal N demand of a crop and is
basic information needed to develop software for computer-based
recommendations. A correct critical concentration is crucial; the error of the
calculated accumulated N uptake will increase each day if based on daily
calculations from a wrong critical concentration. Increasing PNC with +0.5% adds
100 kg N ha-1 more in N-uptake to the 390 kg N ha-1 calculated for the PNCc.

The principles behind a constant PNCa leading to the same rate (–1/3) of decline
of LAR and PNC, which is agreement with the skin-core hypothesis, and the
principles behind constant LNCa of unshaded leaves seem to be general for various
crops. However, the –1/3 rate does not generally apply to all crops as is shown by

37
38

deviation
the varying
Table 2. Values of parameters ac and bc of equation 1 for critical PNC for total N of various C3 and C4- crops
Species a b References

values of
C3
White cabbage (Brassica oleracea L. var. capitata L. alba DC.) 5.1 -0.33 Paper 4
Lettuce (Lactuca sativa L. Var. capitata L.) 4.6 -0.36 Tei et al. (2003)
Linseed (Linum usitatissimum L.) 4.7 -0.53 Flénet et al. (2006)
Pea (Pisum sativum L.) 5.1 -0.32 Ney et al. (1997)

parameter bPNCc for

Potato, (Solanum tuberosum L)
Potato, cv. Russet Burbank 4.6 -0.42 Belanger et al., (2001)
Potato, cv. Shepody 5.0 -0.42 Belanger et al., (2001)
Potato, cv. Bintje & cv. Kaptah Vandel 5.2 -0.56 Duchenne et al., (1997)
Rapeseed (Brassica napus L) 4.5 -0.25 Colnenne et al. (1998)
Tall fescue (Festuca arundinacea Schreb.) 4.8 -0.32 Lemaire & Salette (1984)
Tomato (Lycopersicon esculentum Mill.) 4.5 -0.33 Tei et al. (2002)
Wheat (Triticum aestivum L.) 5.3 -0.44 Justes et al., (1994)
C4

different crops (Table 2). The

Maize (Zea mays L.) 3.4 -0.37 Plenét & Lemaire (1999)
Sorghum (Sorghum bicolour L.) 3.9 -0.39 Plenét & Cruz (1997)
seems to be caused by special traits of the crop. The linseed crop was originally
bred for fibre which explains the steep decline with a bPNCc of –0.53 (Flénet et al.
2006). The lower rate of decline in bPNCc (bPNCc of –0.25) of winter oilseed rape
was explained by a higher N absorption capacity due to increased root growth in
autumn and N reallocation from senescent leaves to the youngest leaves during
winter (Colnenne et al., 1998). Further research is needed into the causes of
differences between crops and cultivars and to identify possible common traits for
crops and cultivars with similar critical concentrations.

The actual N status of cabbage at any time during the growth period can be
evaluated by means of the critical concentrations presented. The NNI gives a
convenient measure of the N status and is defined as the actual concentration
divided by the critical concentration (Lemaire & Gastal, 1997). In the case of
analysis of the horizontally orientated unshaded leaves, the same value of critical N
concentration can be used during the entire growth period for calculation of the
NNI.

The NNI can be used to evaluate the actual plant nutrient management strategy.
Such an evaluation is an important component of a nutrient management strategy
which tries to as closely as possible match N supply with demand during the entire
growing period. Such nutrient management strategies have been made possible
through the use of computer-aided decision support tools such as N-Expert and
N_Able (Fink & Feller, 1997; Greenwood, 2001). Evaluation will also be called
for when more attention is paid to the N supply from manures and residues from
the preceding crop (Rahn, 1992; Torstensson & Ekbladh, 2002) as this supply is
difficult to predict and therefore best evaluated in terms of its effect on crop N
uptake. This is of course of special interest in organic vegetable production. In
such production systems crop yields are often lower than in their conventional
counterparts and this is often believed to -at least partially- be due to crop nutrient
deficiencies. In such systems the farmer often has very little information as to the
extent to which the nutrient demand of the crop is met, unless the nutrient status of
the crop can be evaluated. Also in research is the nutrient supplying capacity of
different organic manures (e.g. legumes, animal manures, etc.) or in different crop
rotations often evaluated in terms of their effects on total yield, rather than in terms
of their ability to meet crop nutrient demand (e.g. Ögren et al., 1998). At both the
research and at the practical level an assessment of the crop nutrient status in terms
of nutrient excess or deficiency in relation to potential crop growth would therefore
be extremely valuable.

Conclusions
The critical N concentration of cabbage on a whole plant basis was estimated as
PNCc = 4.5 (% of DM) W < 1.5 t ha-1
PNCc = 5.1W-0.33 (% of DM) W > 1.5 t ha-1

39
 The rate of decline in PNC was characterised by the exponent –0.33. This rate
was similar to the rate of decline in LAR for plants grown with ample N supply.
The rate of decline in PNC can thus be predicted from that of LAR, but prediction
from the regression of LA to W is preferable because pre-linear growth phase data
points can also be included in the regression. The rate in LAR changed stepwise
following changes from one growth phase to the next (shown for plants in the
climate chamber). LAR and PNC declined at similar rates when PNCa remained
constant with growth.

RGR and PNC were constant during the exponential growth phase (with a
resulting constant nitrogen productivity according to the nutrient productivity
theory). After this phase, RGR and PNC declined at different rates with the
difference being equal to the rate of decline in nitrogen productivity and, as PNCa
was constant, equal to the rate of decline in NAR. Thus one growth component of
RGR, LAR, declined at the same rate as PNC whereas the other, NAR, made up
the difference in the rate of decline between PNC and RGR. The difference in rate
of decline in PNC and RGR was most likely mainly due to self shading.

Leaf area based plant analysis on individual leaves appears promising as the
critical concentration will be constant during the entire growth period for leaves
fully exposed to light. A value of 3.7 g N m-2 for white cabbage was derived from
the critical PNCa but the ratio between LNCa for leaves fully exposed to light and
PNCa has yet to be validated under field conditions.

References
Asher, C. J., Ozanne, P. G. & Loneragan, J. F. 1965. A method for controlling the ionic
environment of plant roots. Soil Science 103, 155-161.
Barker, A.V., Peck, N.H. & MacDonald, G.E. 1971. NO3-N accumulation in vegetables. I.
Spinach grown in upland soil. Agronomy Journal 63, 126-129.
Bates, T.E. 1971. Factors affecting critical nutrient concentration in plants and their
evaluation: A review. Soil Science 112, 116-130.
Belanger, G., Walsh, J.R., Richards, J.E., Milburn, P.H. & Ziadi, N. 2003. Critical petiole
nitrate concentration of two processing potato cultivars in eastern Canada. American
Journal of Potato Research 80, 251-262.
Belec, C., Villeneuve, S., Coulombe, J. & Tremblay, N. 2001. Influence of nitrogen
fertilization on yield, hollow stem incidence and sap nitrate concentration in broccoli.
Canadian Journal of Plant Science 81, 765-772.
Benton Jones Jr, J. 1985. Soil testing and plant analysis: guides to the fertilization of
horticultural crops. Horticultural Reviews 7, 1-68.
Binford, G. D., Blackmer, A. M. & Cerrato, M. E. 1992. Nitrogen concentration of young
corn plants as an indicator of nitrogen availability. Agronomy Journal 84, 219-223.
Blackman, G.E. & Wilson, G.L. 1951. Physiological and ecological studies in the analysis
of plant environment. VI. The constancy for different species of a logarithmic
relationship between net assimilation rate and light intensity and its ecological
significance. Annals of Botany 15, 63-94.

40
Booij, R., Kreuzer, A.D.H., Smit, A.L. & Vanderwerf, A. 1996. Effect of nitrogen
availability on dry matter production, nitrogen uptake and light interception of Brussels
sprouts and leeks. Netherlands Journal of Agricultural Science 44, 3-19.
Booij, R., Groenwold, K., Rovers, J., Clevering, O., Pijnenburg, H., Hekkert, M. &
Langeveld, H. 2003. Nutrient management on vegetable farms; What will be the Future?
Acta Horticulturae 627, 275-282.
Caloin, M. & Yu, O. 1982. An extension of the logistic model of plant growth. Annals of
Botany 49, 599-607.
Caloin, M. & Yu, O. 1984. Analysis of the time course of change in nitrogen content in
Dactylis glomerata L. using a model of plant growth. Annals of Botany 54, 69-76.
Castellanos, J. Z., Villalobos, S., Delgado, J. A., Munoz-Ramos, J., Sosa, A., Vargas, P.,
Lazcano, I., Alvarez-Sanchez, E. & Enriquez, S. A. 2001. Use of best management
practices to increase nitrogen use efficiency and protect environmental quality in a
broccoli-corn rotation of central Mexico. Communications in Soil Science and Plant
Analysis 32, 1265-1292.
Charles-Edwards, D.A., Stutzel, H., Ferraris, R. & Beech, D.F. 1987. An analysis of spatial
variation in the nitrogen content of leaves from different horizons within a canopy.
Annals of Botany 60, 421-426.
Colnenne, C., Meynard, J. M., Reau, R., Justes, E. & Merrien, A. 1998. Determination of a
critical nitrogen dilution curve for winter oilseed rape. Annals of Botany 81, 311-317.
Coulombe, J., Villeneuve, S., Lamy, P., Yelle, S., Belec, C. & Tremblay, N.N. 1999.
Evaluation of soil and petiole sap nitrate quick tests for broccoli in Quebec. Acta
Horticulturae 506, 147-152.
De Pinheiro Henriques, A.R. & Marcelis, L.F.M. 2000. Regulation of growth at steady-state
nitrogen nutrition in lettuce (Lactuca sativa L.): Interactive effects of nitrogen and
irradiance. Annals of Botany 86, 1073-1080.
Dole, J.M. & Wilkins, H.F. 1991. Relationships between nodal position and plant age on
the nutrient composition of vegetative poinsettia leaves. Journal of American Society of
Horticultural Science 116, 248-252.
Donald, C.M. 1961. Competition for light in crops and pastures. In: Mechanisms in
biological competition. (Ed. F. L. Milthorpe). Symposium of the society for experimental
biology Vol. 15, Cambridge University Press. Cambridge, pp. 282-313.
Dow, A.I. & Roberts, S. 1982. Proposed: critical nutrient ranges for crop diagnoses.
Agronomy Journal 74, 402-404
Duchenne, T., Machet, J.M. & Martin, M. 1997. Potatoes. In: Diagnosis of the nitrogen
status in crops. (Ed. G. Lemaire). Springer-Verlag. Berlin Heidelberg, pp. 119-130.
Ekbladh, G. 2003. Optimal kvävegödsling till vitkål med N-min och N-Expert 2003. SLU,
Swedish University of Agricultural Sciences, Torslunda Försöksstation, Årsrapport 2003.
Färjestaden, 7-14.
Ekbladh, G., Witter, E. & Ericsson, T. 2007. Ontogenetic decline in the nitrogen
concentration of field grown white cabbage -Relation to growth components. Scientia
Horticulturae 112, 149-155.
Eurostat. 2007. Eurostat yearbook
http://epp.eurostat.ec.europa.eu/extraction/retrieve/en/theme5/apro/apro_cpp_fruveg?Out
putDir=EJOutputDir_341&user=unknown&clientsessionid=965B287863DABCF321514
CDB55B14D08.extraction-worker-
1&OutputFile=apro_cpp_fruveg.htm&OutputMode=U&NumberOfCells=1&Language=e
n&OutputMime=text%2Fhtml&) Accessed 2007-05-02
Feller, C., Fink, M., Maync, A., Paschold, P.J., Scharpf, H.-C., Schlaghecken, J.,
Strohmeyer, K., Weier, U. & Ziegler, J. 2001. Düngung im Freilandgemüsebau.
Gartenbauliche Berichte, Schriftenreihe des Institutes für Gemüse- und Zierpflanzenbau,
Großbeeren/Erfurt e.V. Heft 4. 196 pp.
Fink, M. & Scharpf, H.C. 1993. N-Expert - a decision support system for vegetable
fertilization in the field. Acta Horticulturae 339, 67-74.
Fink, M. & Feller, C. 1997. N-Expert II verfügbar. Deutscher Gartenbau 51, XVII-XIX.

41
Flénet, F., Guerif, M., Boiffin, J., Dorvillez, D. & Champolivier, L. 2006. The critical N
dilution curve for linseed (Linum usitatissimum L.) is different from other C3 species.
European Journal of Agronomy 24, 367-373.
Francescangeli, N., Sangiacomo, M. A. & Marti, H. 2006. Effects of plant density in
broccoli on yield and radiation use efficiency. Scientia Horticulturae 110, 135-143.
Gardner, B.R. & Jones, J.P. 1975. Petiole analysis and the nitrogen fertilization of Russet
Burbank potatoes. American Potato Journal 52, 195-200.
Gardner, B. R. & Roth, R. L. 1989a. Midrib nitrate concentration as a means for
determining nitrogen needs of broccoli. Journal of Plant Nutrition 12, 111-125.
Gardner, B. R. & Roth, R. L. 1989b. Midrib nitrate concentration as a means for
determining nitrogen needs of cabbage. Journal of Plant Nutrition 12, 1073-1088.
Gardner, B. R. & Roth, R. L. 1990. Midrib nitrate concentration as a means for determining
nitrogen needs of cauliflower. Journal of Plant Nutrition 13, 1435-1451.
Geraldson, C.M., Klacen, G.R. & Lorenz, O.A. 1973. Plant analysis as an aid in fertilizing
vegetable crops. In: Soil testing and plant analysis (Eds. L.M. Walsh & J.D. Beaton).
Soil Science Society of America, Inc. Madison, Wisconsin USA, pp. 365-379.
Geyer, B. & Marschner, H. 1990. Application of the rapid nitrate test in maize. Verband
Deutscher Landwirtschaftlicher Untersuchungs- und Forschungsanstalten,
Kongressband 1989 Bayreuth, Reihe Kongressberichte. VDLUFA-Schriftenreihe 30,
141-142, ISBN 3-922712-39-8.
Girardin, P., Tollenaar, M. & Deltour, A. 1985. Effect of temporary N starvation in maize
on leaf senescence. Canadian Journal of Plant Science 65, 819-829.
Glimskär, A. & Ericsson, T. 1999. Relative nitrogen limitation at steady-state nutrition as a
determinant of plasticity in five grassland plant species. Annals of Botany 84, 413-420.
Goodlass, G., Rahn, C., Shepherd, M.A., Chalmers, A.G. & Seeney, F.M. 1997. The
nitrogen requirement of vegetables: Comparisons of yield response models and
recommendation systems. Journal of Horticultural Science 72, 239-254.
Goulding, K. 2000. Nitrate leaching from arable and horticultural land. Soil Use and
Management 16, 145-151.
Greenwood, D. J., Neeteson J. J., & Draycott, A. 1986. Quantitative relationships for the
dependence of growth rate of arable crops on their nitrogen content, dry weight and aerial
environment. Plant and Soil 91, 281-301.
Greenwood, D.J. & Draycott, A. 1989. Experimental validation of an N-response model for
widely different crops. Fertilizer Research 18, 153-174.
Greenwood, D. J., Kubo, K., Burns, I. G. & Draycott, A. 1989. Apparent recovery of
fertilizer-N by vegetable crops. Soil Science and Plant Nutrition 35, 367-368.
Greenwood, D. J., Lemaire, G., Gosse, G., Cruz, P., Draycott, A. & Neeteson, J. J. 1990.
Decline in percentage N of C3 and C4 crops with increasing plant mass. Annals of
Botany 66, 425-436.
Greenwood, D. J., Gastal, F., Lemaire, G., Draycott, A., Millard, P. & Neeteson, J. J. 1991.
Growth rate and %N of field grown crops: Theory and experiments. Annals of Botany 67,
181-190.
Greenwood, D.J., Rahn, C., Draycott, A., Vaidyanathan, L.V. & Paterson, C. 1996.
Modelling and measurement of the effects of fertilizer-N and crop residue incorporation
on N-dynamics in vegetable cropping. Soil Use and Management 12, 13-24.
Greenwood, D.J. 2001. Modeling N-Response of field vegetable crops grown under diverse
conditions with N_Able: A review. Journal of Plant Nutrition 24, 1799-1815.
Grevsen, K. 1998. Effects of temperature on head growth of broccoli (Brassica oleracea L.
var. Italica): parameter estimates for a predictive model. Journal of Horticultural Science
& Biotechnology 73, 235-244.
Grindlay, D.J.C. 1997. Towards an explanation of crop nitrogen demand based on the
optimization of leaf nitrogen per unit leaf area. Journal of Agricultural Science,
Cambridge 1278, 377-396.
Grindlay, D. J. C., Sylvester Bradley, R., & Scott, R. K. 1997. The relationship between
canopy green area and nitrogen in the shoot. In: Diagnostic procedures for crop N
management. (Eds. G. Lemaire & I.G. Burns), Poitiers (France) November 22-23, 1995.

42
 INRA Editions, Vol. 82, Colloques Institut National Recherche Agronomique, Paris, pp.
53-60.
Hardwick, R.C. 1987. The nitrogen content of plants and the self-thinning rule of plant
ecology: a test of the core-skin hypothesis. Annals of Botany 60, 439-446.
Hartz, T.K., Bendixen, W.E. & Wierdsma, L. 2000. The value of presidedress soil nitrate
testing as a nitrogen management tool in irrigated vegetable production. HortScience 35,
651-656.
Hartz, T.K. 2003. The assessment of soil and crop nutrient status for the development of
efficient fertilizer recommendations. Acta Horticulturae 627, 231-240.
Hartz, T. 2004. Efficient fertilization in field vegetable production. Key note speech, NPK-
improved, Workshop, Malmö, 30 November-1 December 2004.
http://www.improved.se/workshop/pdf/ppt_workshop_thartz_vers2.pdf (accessed 15-
March-2007)
Heckman, J. R., Morris, T., Sims, J. T., Sieczka, J. B., Krogmann, U., Nitzsche, P. &
Ashley, R. 2002. Pre-sidedress soil nitrate test is effective for fall cabbage. Hortscience
37, 113-117.
Hirose T. 1988. Modelling the relative growth rate as a function of plant nitrogen
concentration. Physiologia Plantarum 72, 185-189.
Hirose, T. & Werger, M.J.A. 1987. Maximizing daily canopy photosynthesis with respect to
the leaf nitrogen allocation pattern in the canopy. Oecologia 72, 520-526.
Huett, D.O. & Rose, G. 1989. Diagnostic nitrogen concentrations for cabbages grown in
sand culture. Australian Journal of Experimental Agriculture 29, 883-892.
Hunt, R. 1978. Plant Growth Analysis. The Institute of Biology’s Studies in Biology,
Vol. 96. Arnold. London. 67 pp.
Huus-Bruun, T. 2003. Potential leaching from vegetable farms in Denmark. Acta
Horticulturae 339, 219-227.
Ingestad, T. & Lund, A.B. 1986. Theory and techniques for steady state mineral nutrition
and growth of plants. Scandinavian Journal of Forestry Research 1, 439-453.
Ingestad, T. & McDonald, A.J.S. 1989. Interaction between nitrogen and photon flux
density in birch seedlings at steady state nutrition. Physiologia plantarum 77, 1-11.
Ingestad, T. & Ågren, G.I. 1992. Theories and methods on plant nutrition and growth.
Physiologia Plantarum 84, 177-184.
Iversen, K.V., Fox, R.H. & Piekielek, W.P. 1985. Diurnal, shade and hybrid effects on
nitrate content of young corn stalks. Agronomy Journal 16, 837-852.
Justes, E., Mary, B., Meynard, J.M., Machet, J.M. & Thelierhuche, L. 1994. Determination
of a critical nitrogen dilution curve for winter-wheat crops. Annals of Botany 74, 397-
407.
Kalra, Y.P. 1998. Handbook of reference methods for plant analysis. CRC Press. Boca
Raton. 300 pp.
Kubota, A., Thompson, T. L., Doerge, T. A. & Godin, R. E. 1996. A petiole sap nitrate test
for cauliflower. Hortscience 31, 934-937.
Kubota, A., Thompson, T. L., Doerge, T. A. & Godin, R. E. 1997. A petiole sap nitrate test
for broccoli. Journal of Plant Nutrition 20, 669-682.
Lancashire, P.D., Bleiholder, H., Van den Boom, T., Langeluddeke, P., Stauss, R., Weber,
E. & Witzenberger A. 1991. A uniform decimal code for growth stages of crops and
weeds. Annals of Applied Biology 119, 561-601.
Lemaire, G. & Salette, J. 1984. Relation entre dynamique de croissance et dynamique de
prélèvement d'azote pour un peuplement de graminées fourragères. I. Etude d l'effet du
milleu. Agronomie 4, 423-430.
Lemaire, G., Onillon, B., Gosse, G., Chartier, M. & Allirand, J. M. 1991. Nitrogen
distribution within a lucerne canopy during regrowth: Relation with light distribution.
Annals of Botany 68, 483-488.
Lemaire, G. & Gastal, F. 1997. N uptake and distribution in plant canopies. In: Diagnosis of
the nitrogen status in crops. (Ed. G. Lemaire). Springer-Verlag. Berlin Heidelberg, pp. 3-
43.

43
Lemaire, G., Plénet, D., Grindlay, D., & Gastal, F. 1997. Leaf N content as an indicator of
crop N nutrition status. In: Diagnosis of the nitrogen status in crops. (Ed. G. Lemaire).
Springer-Verlag. Berlin Heidelberg, pp. 189-199.
Lemaire, G., Avice, J-C., Kim, T-H. & Ourry, A. 2005. Development changes in shoot N
dynamics of lucerne (Medicago sativa L.) in relation to leaf growth dynamics as a
function of plant density and hierarchical position within the canopy. Journal of
Experimental Botany 56, 935-943.
Lewis, D.C., Grant, I.L. & Maier, N.A. 1993. Factors affecting the interpretation and
adoption of plant analysis services. Australian Journal of Experimental Agriculture 33,
1053-1066.
Lorentz, H-P., Schlaghecken, G., Engl, G., Maync, A., & Ziegler, J. 1989.
Ordnungsgemäße Stickstoffversorgung in Freilandgemüsebau nach dem "
Kulturbegleitenden Nmin Sollwerte (KNS)-System. Ministerium für Landwirtschaft,
Weinbau und Forsten, Rheinland-Pfalz, Mainz. 85 pp. ISSN 0931-9026.
Lorenz, O. A. & Tyler, K. B. 1977. Plant tissue analysis of vegetable crops. Division of
Agricultural Sciences, University of California. Bulletin No 1879, 21-24.
MacKerron, D.K.L., Young, M.W. & Davies, H.V. 1995. A critical assessment of the value
of petiole sap analysis in optimizing the nitrogen nutrition of the potato crop. Plant and
Soil 172, 247-260.
MacMurdo, W., Prange, R.K. & Veinot, R. 1988. Nitrogen fertilization and petiole tissue
testing in production of whole seed tubers of the potato cultivars Sebago and Atlantic.
Canadian Journal of Plant Science 68, 901-905.
Macy, P. 1936. The quantitative mineral nutrient requirements of plants. Plant Physiology
11, 749-764.
Magdoff, F.R. 1991. Understanding the Magdoff pre-sidedress nitrate test for corn. Journal
of Production Agriculture 4, 297-305.
Matthäus, D. & Gysi, C. 2001. Plant-sap analysis in vegetables - a tool to decide on
nitrogen top dressing. Acta Horticulturae 563, 93-102.
Maynard, D. & Hochmut, G.J 1997. Knott's Handbook for Vegetable Growers. 4th edition.
John Wiley & Sons. New York. 600 pp.
McDonald, A.J.S., Lohammar, T. & Ingestad, T. 1992. Net assimilation rate and shoot area
development in birch (Betula pendula Roth.) at different steady-state values of nutrition
and photon flux density. Trees: Structure and Function 6, 1-6.
Meier, U. 1997. Growth stages of mono- and dicotyledonous plants: BBCH-Monograph.
Blackwell Wissenschafts-Verlag. Berlin Boston. 622 pp.
Mills, H.A. & Jones, J.B. 1996. Plant analysis handbook II. A practical sampling,
preparation, analysis and interpretation guide. Micro-Macro Publishing Inc. Athens.
422 pp.
Monsi, M. & Saeki, T. 2005. On the factor light in plant communities and its importance for
matter production. Annals of Botany 95, 549-567.
Monteny, G.J. 2001. The EU Nitrates Directive: a European approach to combat water
pollution from agriculture. TheScientificWorld 1, 927-935.
Neeteson, J.J. & Carton, O.T. 2001. The environmental impact of nitrogen in field
vegetable production. Acta Horticulturae 563, 21-28.
Neeteson, J.J., Langeveld, J.W.A. & Smit, A.L. 2003. Nutrient balances in field vegetable
productions systems. Acta Horticulturae 627, 13-23.
Ney, B., Dore, T. & Sagan, M. 1997. The N requirement of major agricultural crops: grain
legumes. In: Diagnosis of the nitrogen status in crops. (Ed. G. Lemaire). Springer-
Verlag. Berlin Heidelberg, pp. 107-118.
Niklas, K.J. 1994. Plant Allometry. The scaling of form and process. The University of
Chicago Press. Chicago London. 395 pp.
Niklas, K.J. 2006. Plant Allometry, Leaf nitrogen and phosphorus stoichiometry, and
interspecific trends in annual growth rates, Annals of Botany 97, 155-163.
Ögren E., Båth B. and Rämert B. 1998. First and second year nitrogen effects of autumn
and spring- incorporated green-manure crops in field vegetable production. Swedish
Journal of Agricultural Research 28, 137-146.

44
Ogunlela, V.B., Kullmann, A. & Geisler, G. 1990. Nitrogen distribution and dry matter
accumulation in oilseed rape (Brassica Napus L.) as influenced by N nutrition. Journal of
Agronomy and Crop Science 164, 321-333.
Olesen, J.E., Berntsen, J., Hansen, E.M., Petersen, B.M. & Petersen, J. 2002. Crop nitrogen
demand and canopy area expansion in winter wheat during vegetative growth. European
Journal of Agronomy 16, 279-294.
Olfs, H.W., Blankenau, K., Brentrup, F., Jasper, J., Link, A. & Lammel, J. 2005. Soil- and
plant-based nitrogen-fertilizer recommendations in arable farming. Journal of Plant
Nutrition and Soil Science 168, 414-431.
Plénet, D. & Cruz, P. 1997. The N requirement of major agricultural crops: maize and
sorghum. In: Diagnosis of the nitrogen status in crops. (Ed. G. Lemaire). Springer-
Verlag. Berlin Heidelberg, pp. 3-43.
Plénet, D. & Lemaire, G. 1999. Relationships between dynamics of nitrogen uptake and dry
matter accumulation in maize crops. Determination of critical N concentration. Plant and
Soil 216, 65-82.
Porter, G.A. & Sisson, J.A. 1991. Petiole nitrate content of Maine-grown Russet Burbank
and Shepody potatoes in response to varying nitrogen rate. American Potato Journal 68,
493-505.
Rahn, C. R. 1992. Nitrogen residues from brassica crops. Aspects of Applied Biology 30,
263-270.
Rahn, C. R., Greenwood, D. J. & Draycott, A. 1996. Prediction of nitrogen fertiliser
requirement with HRI WELL_N computer model. In: Proceedings of 8th Nitrogen
Workshop, Ghent, 5-8 September 1994. (Eds. O. Van Cleemput, G. Hofman, & A.
Vermoesen). Progress in Nitrogen Cycling Studies, Developments in Plant & Soil
Sciences, Vol. 68, Kluwer Academic Publishers. Dordrecht, Netherlands, pp. 255-258.
Reuter, D. J. & Robinson, J. 1997. Plant Analysis: an Interpretation Manual. B. 2nd
edition. CSIRO Publishing, Collingwood, Australia, 572 pp.
Richardson, A.D., Duigan, S.P. & Berlyn, G.P. 2002. An evaluation of noninvasive
methods to estimate foliar chlorophyll content. New Phytologist 153, 185-194.
Riley, H. & Guttormsen, G. 1999. Alternative equations for critical N-concentration in
cabbage. Acta Horticulturae 506, 123-128.
Rimpau, J. 1984. Mit einem „Düngefenster“ die Stickstoffnachlieferung abschätzen.
Deutsche Landwirtschafts-Gesellschaft. DLG-Mitteilungen 2, 72-73.
Roling, N.G. & Wagemakers, A.E. 2000. Facilitating Sustainable Agriculture -
Participatory Learning and Adaptive Management in Times of Environmental
Uncertainty. Cambridge University Press. Cambridge. 318 pp.
Salomez, J., Hofman, G. & Demyttenaere, P. 2005. A future for vegetable cultivation within
the Flemish manure decree? Communications in Soil Science and Plant Analysis 36, 661-
672.
Santamaria, P. 2006. Nitrate in vegetables: toxicity, content, intake and EC Regulation.
Journal of the Science of Food and Agriculture 86, 10-17.
Scaife, A. & Stevens, K. L. 1983. Monitoring sap nitrate in vegetable crops: comparison of
test strips with electrode methods, and effects of time of day and leaf position.
Communications in Soil Science and Plant Analysis 14, 761-771.
Scaife, A. & Turner, M. K. 1987. Field measurements of sap and soil nitrate to predict
nitrogen top-dressing requirements of Brussels sprouts. Journal of plant nutrition 10,
1705-1712.
Scaife, A. 1988. Derivation of critical nutrient concentrations for growth-rate from data
from field experiments. Plant and Soil 109, 159-169.
SCB and SJV. 2006a. Sveriges officiella Statistik, Statistiska meddelanden
Trägårdsproduktion 2005. 16. SCB and SJV. Serie JO, Jordbruk, skogsbruk och fiske.
SCB and SJV. 2006b. Yearbook of agricultural statistics 2006 including food statistics. 49.
Örebro, SCB and SJV. Official Statistics of Sweden, Statistics Sweden.
SCB and SJV. 2006c. Sveriges officiella Statistik, Statistiska meddelanden
Trädgårdsproduktion 2005. 25. SCB and SJV. Serie JO, Jordbruk, skogsbruk och fiske.
Scharf, P.C. 2001. Soil and plant tests to predict optimum nitrogen rates for corn. Journal
of Plant Nutrition 24, 805-826.

45
Scharpf., H-C. 1991. Stickstoffdüngung im Gemüsebau. Auswertungs- und
Informationsdienst für Ernährung, Landwirtschaft und Forsten (AID) e.V., Bonn. 36 pp.
Schröder, J.J., Neeteson, J.J., Oenema, O. & Struik, P.C. 2000. Does the crop or the soil
indicate how to save nitrogen in maize production? Reviewing the state of the art. Field
Crops Research 66, 151-164.
Siman, G. 1974. Nitrogen status in growing cereals, with special attention to the use of
plant analysis as a guide to supplemental fertilization. Doctoral thesis. The Royal
Agricultural College of Sweden, Uppsala. 93 pp.
Smolders, E., Merckx, R., Schoovaerts, F. & Vlassak, K. 1991. Continuous shoot growth
monitoring in hydroponics. Physiologia-Plantarum 83, 83-92.
Sorensen, J.N. 2000. Ontogenetic changes in macro nutrient composition of leaf-vegetable
crops in relation to plant nitrogen status: A review. Journal of Vegetable Crop
Production 6, 75-96.
Sorensen, J.N., Staugaitis, G. & Cizauskas, A. 2006. Optimum nitrogen supply for leeks at
two climatic conditions in the Baltic sea region. Acta Horticulturae 700, 149-152.
Swaider, J.M., Sullian, J.G., Grunau, J.A. & Freiji, F. 1988. Nitrate monitoring for pumpkin
production on dryland and irrigated soils. Journal of American Society of Horticultural
Sciene 113, 684-689.
Sylvester-Bradley, R., Stokes, D.T., & Scott, R. K. 1990. A physiological analysis of the
diminishing responses of winter wheat to applied nitrogen. 2. Evidence. Cereal quality II.
Aspects of Applied Biology Vol. 25, Association of Applied Biologists. Wellesbourne,
England. pp. 289-300.
Tabor, J. A., Pennington, D. A. & Warrick, A.W. 1984. Sampling variability of petiole
nitrate in irrigated cotton. Communications in Soil Science and Plant Analysis 15, 573-
585.
Tei, F., Benincasa, P. & Guiducci, M. 2003. Critical nitrogen concentration in lettuce. Acta
Horticulturae 627, 187-194.
Tei, F., Benincasa, P. & Guiducci, M. 2002. Critical nitrogen concentration in processing
tomato. European Journal of Agronomy 18, 45-55.
Torstensson, G. & Ekbladh, G. 2002. Kväveutlakning i frilandsodling av sallat och vitkål på
sandig mojord med olika kvävegödslingsmodeller. Resultat från södra Halland, perioden
1995-97. Swedish Universityof Agricultural Sciences, Division of Water Quality
Management, Ekohydrologi 62, pp 29.
van Dijk, W. & Smit, A.L. 2006. How to meet the EC-Nitrate directive in Dutch vegetable
growing? Acta Horticulturae 700, 191-198.
Vos, J. & Van Der Putten, P.E.L. 2001. Effects of partial shading of the potato plant on
photosynthesis of treated leaves, leaf area expansion and allocation of nitrogen and dry
matter in component plant parts. European Journal of Agronomy 14, 209-220.
Warren Wilson, J. 1972. Control of crop processes. In: Crop processes in controlled
environments. (Eds. A.R. Rees, K.E. Cockshull, D.W. Hand & R.G. Hurd). Academic
Press. London New York, pp. 7-30.
Wehrmann, J. & Scharpf, H.C. 1979. Der Mineralstickstoffgehalt des Bodens als Maßstab
für den Stickstoffdüngerbedarf (Nmin-Methode). Plant and Soil 52, 109-126.
Westerveld, S.M., McKeown, A.W., McDonald, M.R. & Scott-Dupree, C.D. 2003a.
Chlorophyll and nitrate meters as nitrogen monitoring tools for selected vegetables in
southern Ontario. Acta Horticulturae 627, 259-266.
Westerveld, S.M., McKeown, A.W., Scott-Dupree, C.D. & McDonald, M.R. 2003b. How
well do critical nitrogen concentrations work for cabbage, carrot, and onion crops?
HortScience 38, 1122-1128.
Westoby, M. 1984. The self thinning rule. Advances in Ecological Research 14, 167-225.
Wright, I.J. & Westoby, M. 2001. Understanding seedling growth relationships through
specific leaf area and leaf nitrogen concentration: Generalisations across growth forms
and growth irradiance. Oecologia 127, 21-29.
Ziegler, J., Strohmeier, K., and Brand, T. 1996. Nitrogen supply of vegetables based on the
"KNS-system". Acta Horticulturae 428, 223-233.

46
Appendices

Appendix A, Relationship of LAR and PNC

According to Caloin & Yu (1984), Greenwood et al. (1991) and Lemaire & Gastal
(1997), PNC is linearly related to RGR, equation 2:
 PNC m − PNC s 
PNC =   RGR + PNC s
 k 
The constant k is GR divided by Wm and was assumed to be constant. Wm is the
“active” biomass and is mainly located in the leaf. It can be assumed that this type
of tissue constitutes a certain proportion of the leaf. When leaf area increases it can
therefore be expected that the amount of metabolic dry matter also increases in
proportion to the increase of leaf area. NAR decreases due to increasing LA at
constant GR during the linear growth phase. If Wm is related to LA rather than to
GR, it can be assumed that k decreases at the same rate if metabolic tissues
constitute a certain proportion of the leaf. The metabolic processes become less
active and less effective although the metabolic tissues remain intact. An analogy
between GR to LA and GR to Wm can be assumed as long the proportion of
metabolic tissues within the leaf are not reduced by decreasing intercepted
radiation:
GR GR
= k2 ×
Wm LA

The equation is rearranged and W is inserted on both sides:

Wm 1 LA GR
= * *
W k2 GR W

Simplified:
Wm 1
= × LAR
W k2

Equation 9 in Lemaire & Gastal (1997):

PNC m Wm + PNC s Ws
PNC =
W
The pool of structural biomass Ws= W-Wm:
Wm PNC s × W W
PNC = PNC m + − PNC s × m
W W W
Wm/W is substituted with 1/k2 × LAR and the equation is rearranged:
1 (3)
PNC = ( PNC m − PNC s ) × LAR + PNC s
k2

47
If the assumption of an analogy between GR/Wm and GR/LA is valid, PNC will be
linearly related to LAR during the linear growth phase.

Appendix B, Determination of LNCa from LNC of Huett & Rose

(1989)

To estimate LNCa for leaves fully exposed to light under field conditions data of
LNC of the youngest fully expanded leaves (YFEL) from Huett & Rose (1989)
were used and data of SLA from the T2002b and Ph2004 experiments. It was
assumed that LNCa was at its maximum at the same leaf positions in the field as in
the climate chamber at a given plant size (W). SLA was selected at leaf positions
where LNCa was at its maximum and at a similar W as in the climate chamber.
These data of SLA were plotted against W, the linear regression for 20 – 82 DAT
was SLA = 124.5-0.123*W, R2 = 0.983. Data for W that corresponded to the LNC
data was not given by Huett & Rose (1989). Therefore, W was estimated from the
biweekly data of GR for 29 mmol L-1 given in Table 1. Data for the N application
rate 29 mmol L-1 were used. Decreasing GR during the last weeks suggested use of
the logistic growth function:

dW(t)  W(t) 
= a W(t) 1− 
dt  W f 

The biweekly data of GR were fitted to the derivative with respect to time of the
logistic growth function:
dW (t) a W f exp(− a ( t − b))
=
dt [1 + exp(−a(t − b))]2
The parameters were (confidence intervals within parenthesis):
a = 0.54 (0.48 - 0.59), b = 8.77 (8.54 -9.01), Wf = 313 (289-336)
The accumulated growth (W) was calculated from the analytical solution to the
logistic growth function with the values of the parameters inserted:
313
W(t) =
1 + exp (−0.54 ( t − 8.77 ))

This expression was used to calculate W corresponding to the LNC of Table 2 in

Huett & Rose (1989). The linear relationship (124.5-0.123*W) was used to
calculate SLA. LNCa was finally calculated from the calculated SLA and LNC of
Table 2 in Huett & Rose (1989) (LNCa = LNC/SLA) (Table 2).

48
Acknowledgements
Swedish Council for Environment, Agricultural Sciences and Spatial Planning
(Forskningsrådet för miljö, areella näringar och samhällsbyggande, FORMAS) is
acknowledged for the financial support. The Swedish Farmers’ Foundation for
Agricultural Research (Stiftelsen Lantbruksforskning, SLF) and The Foundation
for Plant Nutrition Research (Stiftelsen Svensk Växtnäringsforskning, SV) are
acknowledged for supplementary support which made it possible to include
analysis of individual leaves in the study.

I am sincerely grateful to all of you who in one way or another have contributed
to this work.

My thanks go first to my main supervisor, Ernst Witter, for introducing me to the

topic of plant analysis, the chapter by Lemaire & Gastal was a challenge from the
first day. Thanks for all discussions over tables and diagrams on the interpretations
of the results. Without your support all the way through, especially with your
patience with my manuscripts, I would not have reached as far as this. A special
thanks as you gave your time although you are now no longer working at SLU.

Many thanks to my supervisor Tom Ericsson who came into the “match at half
time” and guided me past the pitfalls of growing in culture solution in a climate
chamber, which was new to me. Without your experience from the “Ingestad-
method” the match would have been lost! Many thanks also for the discussions on
the results and for your patience all the way through with reading and giving
comments on the manuscripts.

Many thanks for help with statistical analysis and manuscripts to:
Lennart Norell for advice concerning statistical analyses.
Ingvar Nilsson, Olof Andrén and Johan Arvidsson from the Editorial Board
(Redaktionsrådet) for valuable comments on the thesis and manuscripts.
Shelagh Green for checking and correcting the English in the thesis.

Thanks to all of you who worked with the experiments:

The staff at Ultuna Experimental station with the first experiment the year before I
started.
Ingrid Björklund, Ola Jönsson, Janne Pettersson and Eva Bjelkendal at Torslunda
Experimental Station and also Torsten Kellander for all your effort that have made
Torslunda an inspiring place for horticulture.
Emma, Gabriella, John, Lars-Erik, Lisa and Patrik for patient work with roots and
leaves during your summer holiday 2001.
Special thanks go to Göran Nilsson for technical support (also Friday evening!) –
without your kind help and sharing of experience, I would never have managed the
growth units in the Biotron. Thanks also to Olof Hellgren and Jan Rydén at the
Biotron.

49
 Bertil Schelander, Leif Andersson and Per Lindén at the Phytotron and special
thanks to Jan Parsby for innovative help with suction cups and drip irrigation.
Pär Hillström and Allan Lundquist for assistance with drying and grinding samples.
Inger Juremalm and Rose-Marie Eriksson for analyses of numerous samples and
for advice and answers when I needed your help.
Beatrice Tjernberg and Bodil Lindström for valuable help with all the samples
from the Biotron and the Phytotron – I really appreciated your talent for
organization.

Gabriele Engqvist, Svalöf Weibull AB, for providing me with seeds of your
cultivar Heckla although they were out of stock. Thanks to your support, the same
cultivar could be used in all experiments.
Monica Kähr, for valuable discussions and advice.

Thanks to all my PhD colleagues, most of you have already left “Markvetenskap” –
Many thanks to Anke, for continued contact from the other side of the globe and
special thanks for comments to Paper II & III.

I am grateful for the time at “Markvetenskap” –my thanks go to all colleagues.

Last but not least my greatest thanks to my beloved Christhild for all support,
without your support the last time of finishing the thesis would not have been
possible.

50