Review

Landrace Germplasm for

Improving Yield and Abiotic
Stress Adaptation
Sangam L. Dwivedi,1 Salvatore Ceccarelli,2 Matthew W. Blair,3
Hari D. Upadhyaya,1 Ashok K. Are,1 and Rodomiro Ortiz4,*
Plant landraces represent heterogeneous, local adaptations of domesticated
Trends
species, and thereby provide genetic resources that meet current and new
Global climate change emphasizes the
challenges for farming in stressful environments. These local ecotypes can need to use better-adapted cultivars of
show variable phenology and low-to-moderate edible yield, but are often highly the main crops and landraces as
potential donors of useful genes.
nutritious. The main contributions of landraces to plant breeding have been
traits for more efficient nutrient uptake and utilization, as well as useful genes for The contribution of modern agriculture
adaptation to stressful environments such as water stress, salinity, and high to total human-made greenhouse gas
emissions is approximately 30%, and a
temperatures. We propose that a systematic landrace evaluation may define shift to agro-ecological modes of pro-
patterns of diversity, which will facilitate identifying alleles for enhancing yield duction is increasingly seen as urgent,
with landraces playing an important
and abiotic stress adaptation, thus raising the productivity and stability of staple
role in breeding programs.
crops in vulnerable environments.
During the past few decades interest in
landrace conservation has been grow-
Modern Agriculture: A Threat to Landrace Diversity ing, with much research focusing on
The current industrial agriculture system may be the single most important threat to biodiversity the maintenance of on-farm crop
genetic diversity.
(see Glossary) [1,2]. A serious consequence of biodiversity loss is the displacement of locally
adapted landraces with adaptation traits to future climates [3–6] by mono-cropping with There is increasing consumer concern
genetically uniform hybrids and improved cultivars [7]. Modern agriculture has contributed to worldwide about food safety and nutri-
decreasing agricultural biodiversity; most of humankind lives now on only 12 crops, with wheat tion. Landraces or old crop cultivars
may provide solutions as sources of
(Triticum aestivum L.), rice (Oryza sativa L.), maize (Zea mays L.), and potato (Solanum tuber- healthy and nutritious food.
osum L.) accounting for 60% of diets [8]. For example, 74% of rice cultivars in Indonesia are
derived from the same stock, while 50% of wheat, 75% of potato, and 50% of soybeans in USA
trace back to 9, 4, and 6 cultivars, respectively [9]. Likewise, there was a significant loss of
genetic variability when analyzing data of collecting missions in Albania in 1941 and 1993, and in
southern Italy in 1950 and the late 1980s. The genetic erosion was estimated at 72.4% and 1
International Crops Research Institute
72.8%, respectively [10]. Furthermore, the number of rice cultivars declined in Indian farms from for the Semi-Arid Tropics (ICRISAT),
Patancheru 502324, India
about 400 000 before colonialism to 30 000 in the mid-19th century, with unknown thousands 2
Rete Semi Rurali, Via di Casignano,
more being lost after the Green Revolution. Greece also lost 95% of its wheat landraces after 25 50018 Scandicci (FI), Italy
3
being encouraged to replace them with modern cultivars [11]. Department of Agriculture and
Natural Sciences, Lawson Hall,
Tennessee State University, 3500
The evolution of plant breeding may explain genetic erosion and the gradual shift towards a John A. Merritt Boulevard, Nashville,
model of agriculture based on uniformity. For millennia, plant breeding was carried out by TN, USA
4
Swedish University of Agricultural
farmers who selected for specific adaptations leading to the formation of landraces. By contrast, Sciences, Department of Plant
modern plant breeding has emphasized wide adaptation, which has resulted in modern Breeding, Sundsvagen, 14 Box 101,
agriculture depending on a small number of cultivars for major crops. Hence, the main sources 23053 Alnarp, Sweden

of food are more genetically vulnerable than ever before. The importance of diversity loss is
becoming more important today as we face the need to adapt crops to climate change [12,13]. *Correspondence:
However, there are no unanimous views on whether and where diversity has been lost [14]. [email protected] (R. Ortiz).

Trends in Plant Science, January 2016, Vol. 21, No. 1 http://dx.doi.org/10.1016/j.tplants.2015.10.012 31

© 2015 Elsevier Ltd. All rights reserved.
Landraces have long been recognized as source of traits for local adaptation, stress tolerance, Glossary
yield stability, and seed nutrition. This review is focused on identifying and deploying gene(s) for Abiotic stress: climatic or edaphic
yield and abiotic stress adaptation from landraces to modern cultivars in cereal and legume conditions that affect cellular
grain crops. The use of haplotype diversity for identifying superior alleles and genomic homeostasis adversely impacting on
growth and fitness. The major abiotic
signatures of environmental adaptation is also discussed, in addition to stressing the greater stresses include water surplus
use of landrace diversity (Figure 1) in developing climate-smart crops. (leading to flood) or deficit (resulting
in drought), extreme variation in
Underuse of Exotic Germplasm Including Landraces in Plant Breeding temperature (cold and heat), salinity,
nutrient deficiency, ion deficiency,
Programs and ion toxicity.
Improvement in crop plants is occurring in breeding cycle after breeding cycle, and plant breeders Biodiversity: entire diversity of
are optimistic about future improvements. While optimism appears justified, there is an important genes, species, cultivars, and
ecosystems in agriculture, including
consequence associated with this genetic gain that has significant implications for the future. As
taxonomic, ecological, morphological,
favorable alleles are selected and fixed, genetic variability is reduced, thus presumably reducing the and molecular diversity.
potential for future progress. The obvious remedy is to augment improved plant populations with a Fertile Crescent: region of the
continuing infusion of genetic diversity. It appears, however, that programs for adding genetic Middle East extending from the
Persian Gulf, through modern-day
diversity have had relatively low priority, or have had only limited success, compared to programs
southern Iraq, Syria, Lebanon,
that develop cultivars using already improved elite germplasm [15]. Jordan, Israel, and northern Egypt.
Considered to be the birthplace of
Plant breeders are always under pressure to fulfill short-term breeding goals, and to achieve this agriculture, urbanization, writing,
trade, science, history, and organized
they use elite germplasm with which they are familiar and that is reasonably adapted to the target
religion, the area was first populated
population of environments, versus exotic germplasm which may require lengthy pre-adaptation 13 000–11 000 BCE when animal
or pre-breeding [16]. The reluctance of plant breeders to use exotic germplasm is largely due to and plant domestication began in this
fear of loss of co-adapted gene complexes, linkage drag, and the protracted time needed to region, signing the start of agriculture.
Genetic erosion: loss of variation
develop cultivars. In addition, the potential value of such germplasm for the stress environments which may occur at the level of crop,
was neither fully appreciated nor appropriately documented until recently, when advances in cultivar, and allele. Reduction in allelic
plant phenotyping and genomics facilitated access to the discovery and deployment of allelic evenness and richness is the greatest
diversity into improved genetic backgrounds, which resulted in large-scale adoption of abiotic concern in the agriculture of today.
Genome-wide association:
stress-tolerant cultivars, and phosphorous deficiency- and submergence-tolerant rice [17], in assessing marker–trait association
some parts of the world. Today, agricultural production worldwide is affected by climate change. (MTA) by studying a large number of
Landraces, given their past evolutionary history and adaptation to stress environments, often germplasms. This approach takes
advantage of historic linkage
out-yield modern cultivars under low-input production systems. A paradigm shift in breeding
disequilibrium to link phenotypes with
program is needed to ensure greater use of landraces as a resource for improving edible yield, genotypes to find MTAs underlying
nutrition content, and abiotic stress adaptation of locally adapted cultivars for sustaining global complex traits.
food and nutritional security. A pre-breeding program should be carried out by public-sector Genomic signature: a characteristic
frequency of oligonucleotide
breeders and, to achieve this, the private sector should support pre-breeding programs, and the
sequences in a genome or sequence.
product with required characteristics should be shared globally. This two-way interaction is Green Revolution: a suite of
expected to facilitate greater use of unadapted germplasm, including landraces, in the devel- agricultural technologies including
opment of climate-smart crops. irrigation, synthetic nitrogen fertilizer,
pesticides, and improved cultivars
that revolutionized maize, rice, and
Value and Impact of Landraces in Crop Improvement wheat production in many parts of
Landraces Enhance Adaptation and Trait Variability the world.
Cultivars with increased tolerance to abiotic stress are necessary to enhance food and nutritional Haplotype and haplotype map: a
set of alleles in cluster of tightly linked
security. Landraces provide a source for discovering novel alleles for enhancing crop adaptation
genes on a chromosome that are
to abiotic stress. Maize landraces from Mexico show immense diversity, growing from arid to inherited together. The haplotype
humid environments and from temperate to tropical environments, while observed climatic map is a catalogue of common
adaptability ranged from 0 to 2900 m altitude, 12 to 29.1 8C growing season mean temperature, genetic variants describing what
these variants are, where they occur
426 to 4245 mm annual rainfall, and 12.46 to 12.98 h growing season day-length [18]. Mexican in the DNA, and how they are
maize landraces are known for high-altitude adaptation and tolerance to abiotic stresses [19], distributed among individuals within a
while the naturally occurring mutant Opaque-2 – which gives a chalky appearance to the kernel population or between populations.
and improves protein quality – was found in a Peruvian maize landrace. This trait was thereafter Landrace: a dynamic population(s)
of a cultivated species that has a
used to breed quality-protein maize lines and cultivars, which today grow in Africa, Asia, and historic origin and distinct identity that
Latin America, because it improves the diets of those depending on maize as a staple. Recent

32 Trends in Plant Science, January 2016, Vol. 21, No. 1

Key Figure was locally adapted in association
with traditional farming systems.
Cereal and Legume Landraces: Showing Various Plant and Grain Types, Linkage disequilibrium: nonrandom
association of alleles at different loci.
where They Are Still Grown Near-isogenic lines (NILs):
introgression lines containing a single
fragment or a small number of
genomic introgression fragments
(A) (B) from a donor parent into an
otherwise homogeneous genetic
background.
Pleiotropic effect: genes that
control more than one trait.
Quantitative trait loci (QTLs): DNA
segments carrying genes controlling
quantitative traits.
Recombination: the production of
offspring with a combination of traits
that differ from either parent.
Seed dormancy: seeds unable to
germinate in a specific period of time
under optimal environmental
conditions that are suitable for the
(C) (D) germination of the non-dormant
seed.
Vernalization: a process of
prolonged exposure of plants to low
temperatures that is a prerequisite for
the plants to flower.

(E) (F)

Figure 1. (A) Diversity in cob and seeds among maize landraces. (B) Variation for seed characteristics in barley landraces.
(C) Maldandi 35-1, a popular landrace of sorghum from India. (D) ICTP 8203, an open pollen pearl millet cultivar derived from
the ‘Iniadi’ landrace. (E) Mixture of bush-type or climbing-type red mottled common bean landraces grown between or on
dry corn stalks in Colombia. (F) Tepary bean landraces from Mexico differ in leaf size and shape, or seed color ranging from
black to yellow and white.

Trends in Plant Science, January 2016, Vol. 21, No. 1 33

work on introgression of alleles from diverse landraces into adapted maize cultivars led to
populations producing more grains than recurrent parents, thereby demonstrating that
improved populations are valuable sources for developing cultivars and parental lines of hybrids
combining drought adaptation and high grain yield [19–21].

Post-rainy sorghum [Sorghum bicolor (L.) Moench] in India is a predominant crop grown on
5.7 million ha in the Deccan Plateau of India under receding soil moisture, with highly variable
production environments and low productivity [22]. ‘M35-1’ (Figure 1C) – a selection from a local
landrace ‘Maldandi’ nearly 75 years ago – is the most popular cultivar largely because of its
earliness, high grain and stover yields, bold and lustrous grains, and drought adaptation. Several
Maldandi variants collected from farmers’ fields showed immense phenotypic variation and
molecular diversity, and such phenotypically diverse and molecularly distinct variants may
be used to enrich the Maldandi race genepool [23]. Improving grain yield but retaining grain quality
and adaptive capacity of M35-1 was the major challenge. The use of Maldandi variants in crosses
led to the development of several restorer lines (ICSR#) and cultivars (ICSV#) which, when tested
separately across six locations during the 2014 postrainy season in India, six restorer lines and
two cultivars, respectively, outyielded M35-1 (grain yield: restorer trial, 1.54 t ha 1; variety trial,
2.25 t ha 1) by 83–128% and 33–37%, with some having the same grain quality as M35-1.

‘Iniadi’ pearl millets [Pennisetum glaucum (L.) R.Br] from West Africa are widely adapted,
drought-tolerant, relatively photoperiod-insensitive, and are known for early maturity, and have
a conical head shape with large globular bluish-grey colored grains [24]. Some of the open
pollinated cultivars (OPV) bred in the 1980s are still grown on large acreages in India, for
example, ‘ICTP 8203’ [25] (Figure 1D) grown on about 0.65 million ha despite stiff competition
from hybrid cultivars [26]. The composite cultivar ‘CZP 9802’ – bred from the landrace-based
early Rajasthan population – recorded 24% and 56% higher grain and stover yields than ICTP
8203 (675 kg ha 1 and 1903 kg ha 1, respectively) in drought-prone environments [27].

Mediterranean Landraces Are the Source of Variability for Abiotic Stress Tolerance in Barley
and Wheat
The domestication of wheat and barley took place before 7000 BC in the Fertile Crescent,
which includes parts of Jordan, Lebanon, Palestine, Syria, southeastern Turkey, Iraq, and
western Iran. Evidence suggests that the most important of the early cereals was barley,
and that the first barleys were two-rowed [28]. The wild progenitor of cultivated barley, Hordeum
vulgare ssp. spontaneum, is still widely distributed along the Fertile Crescent where, particularly
in the driest areas, it can be identified within fields of cultivated barley from a distance because of
its height. It is likely that H. spontaneum contributes to the evolutionary processes of barley
landraces through continuous gene introgression [29].

The main contributions of landraces to plant breeding are useful traits for more-efficient nutrient
uptake and utilization, and genes associated with adaptation to water stress, salinity, and high
temperatures [30]. Several comparisons between barley landraces and modern cultivars have
shown that landraces consistently outyield modern cultivars under stress by as much as 20%
maximum grain yield. Furthermore, in contrast to modern cultivars, landraces rarely fail in the most
extremely stressed environments [31]. The yield advantages of landraces over modern cultivars
that were observed under stress conditions in barley cannot be necessarily extrapolated to other
crops with a different range of adaptation: for example, an environment defined as stressful for
wheat or maize may be moderately favorable for barley or sorghum. Hence, comparisons between
crops must consider carefully what is defined as a stress environment in each case.

Traits associated with grain yield under stress of landraces were early growth vigor, earliness,
growth habit, plant height under drought, long peduncle, and a short grain-filling duration [32].

34 Trends in Plant Science, January 2016, Vol. 21, No. 1

Landrace analysis has identified and mapped quantitative trait loci (QTLs) that control these
traits [33]. The strength of the correlation between each of these traits and grain yield under
stress varied from year to year, suggesting that all these traits are important, but their relative
importance depends on the timing, duration, and intensity of drought [34]. Hence, at least for
the abiotic stresses predominant in the Fertile Crescent, what seems to be important is not the
individual trait but the trait complex.

When barley landraces were used in a breeding program based on direct selection in the target
environment, an interesting pattern emerged: lines derived from crosses with landraces not only
outyielded lines derived from crosses without landraces under drought conditions, but the grain
yield of lines derived from crosses with specific landrace lines, such as ‘Tadmor’, a pure line
cultivar selected from a Syrian landrace [35], was twice (1237 kg ha 1) that of lines lacking
Tadmor in their pedigree (604 kg ha 1). This result could suggest that there are blocks of genes
in chromosomal regions with a low frequency of recombination that can confer specific
adaptation to stress environments [2].

It is very likely that landraces possess useful genes for adaptation to future climates. Farmer-
based programs exploiting landrace gene pools may reconcile an increase in available and
accessible food production with an increase in agro-biodiversity [12]. The evidence in wheat also
supports the view that landraces can provide sources for increased biomass and thousand-
kernel weight, both important traits for adaptation to tolerate drought and heat stress [36]. More
recently landraces have also been acknowledged as potential sources of nutritious and healthy
food [37–40].

Common Bean Landraces and Relatives Enhance Abiotic Stress Tolerance

Common bean (Phaseolus vulgaris L.) is a New World domesticate that is arguably the most
diverse of the pulses based on its multiple domestications in the Andes and Mesoamerica [41]. In
conjunction, Andean, Mesoamerican, and inter-genepool landraces have been selected for
growth in disparate environments ranging from lowlands to high altitude mountains, and ranging
from equatorial regions to temperate climates up to 508N [42–46]. As a result, landraces of
common bean are extremely diverse in their adaptation to different soils and weather regimes
both as dry grain and vegetable types. Because many common bean landraces were selected
on marginal lands, they contain traits of importance for adaptation to soil problems such as
nutrient deficiencies and ion toxicities. This is true of the ‘Chaucha Chuga’ landrace G 19833
from Perú that was the subject of QTL analysis for its aluminum (Al) toxicity tolerance as well as its
tolerance of phosphorus (P) deficiency [47–49]. These traits seem to be correlated given the
pleiotropic effects of organic acid exudation, a trait that is widespread and inducible in many
landraces of common beans grown in acid soils that are high in Al or poor in P. G 19833 was
selected as the reference genome for a whole-genome sequencing effort in common bean [50]
because of its status as a multiple abiotic stress resistant landrace.

Particular common bean genotypes share the ability to tolerate the acid soil complex while
others adapt to high temperatures or drought. Al tolerance is found among Andean genotypes
and some lowland landraces from Brazil and Mexico [51]. Al tolerance ranges from high to
extremely low and is often worst in recently bred advanced lines selected on neutral pH soils at
centralized breeding stations. Populations generated from crosses of abiotic stress tolerant
landraces  improved but abiotic stress susceptible cultivars have even been the subject of
various QTL analyses, for example, offspring with G 122 and Indeterminate Jamaica Red that are
landraces with reproductive heat-tolerance [52]. Drought adaptation, based on deep-rooted-
ness or other traits, has been found in specific Mexican landraces of the Durango–Jalisco
germplasm group, but generally breeder-selected advanced lines are used for breeding and
QTL analysis [53–56].

Trends in Plant Science, January 2016, Vol. 21, No. 1 35

In addition to common bean, two of its relatives have been sources of abiotic stress adaptation
for the major cultigen. In this case, these secondary and tertiary genepool relatives are also
cultivated, but have not been subject to breeding programs and are therefore represented
entirely by landraces or wild accessions. These relatives include scarlet runner beans
(P. coccineus L.) from Central and northern South America, that have tolerance to high Al soils
and P acquisition efficiency, and cultivated tepary beans (P. acutifolius Gray) (Figure 1F) from
Mexico and the southwestern USA that show heat- and drought-tolerance. While scarlet runner
bean are easily crossed with common bean, their progeny are rarely useful in breeding programs
because of dominance of the aggressive climbing growth habit. Tepary beans have been,
however, successfully used for breeding advanced lines of common bean with some level of
drought adaptation in congruity backcross lines, although genomic introgression levels are low
[57]. Empirical results indicate that tepary bean landraces may contribute to heat tolerance, and
their characteristic water use efficiency under high temperatures may give them a clear advan-
tage. Salt tolerance was noticed in tepary bean landraces but not in common bean.

Identifying Genomic Signatures of Environmental Adaptation

Improving environmental adaptation in crops is essential for food and nutrition security under
global climate change. Single nucleotide polymorphism (SNP) variants associated with envi-
ronment origin in crop landraces are likely to reflect adaptations that could then be used to
predict phenotypic variation for adaptive traits. A study using 404 627 SNPs on 1943 georefer-
enced sorghum landraces found that environment explained a substantial portion of SNP
variation, independently of geographical distance, and genic SNPs were enriched for environ-
mental (drought and Al toxicity) associations [58]. Thus, genomic signatures of environmental
adaptation may be useful for crop improvement, enhancing germplasm identification and
marker-assisted selection.

Haplotype Diversity and Identifying Superior Alleles in Landraces

Maize was domesticated in southern Mexico between 6000 and 9000 years ago, but its landraces
are found where farmers have grown this crop for many centuries [59], thus selecting germplasm
with better adaptability; for example, Al tolerance in ‘Cateto’ landraces from South America and
their use in local maize hybrid breeding [60]. Maize inbred lines capture less than 80% of landrace
alleles, suggesting that landraces have significant value as sources of genetic diversity for further
maize breeding [61]. Linkage disequilibrium, which has increased as a result of domestication,
declines to below 1 kb in maize landraces [62], which provides more opportunity for selection.
A DNA survey of maize landraces may reveal functional alleles existing before modern industrial
agriculture, thereby facilitating the understanding of the frequency distributions of haplotypes
among MesoAmerican landraces [19]. Such an undertaking will also determine the extent of allelic
variation in available maize genetic resources. The genome sequence of the highland popcorn
‘Palomero Toluqueño’ landrace provided a means to identify genes related to abiotic stress
response or heavy-metal detoxification [63]. The re-sequencing of maize landrace core subsets
[64] will give further insights about the distribution of SNP allelic variants in the genome, thus
facilitating both gene discovery and establishing their distinct haplotypes. For example, the
resequencing of 517 rice landraces led to the identification of about 3.6 million SNPs that were
further used in developing a high-density haplotype map and for genome-wide association
research to unravel the genetic basis of 14 agronomic traits [65].

Landraces Contribute Yield and Abiotic Stress Adaptation Alleles to Modern

Cultivars
Abiotic Stress Adaptation Alleles
About 29% of the 135 million ha of rainfed lowland rice area in Asia experiences one or two
weeks of prolonged flood, leading to substantial production loss. The discovery of submer-
gence-tolerant landrace ‘FR13A’ led to the identification of SUBMERGENCE 1 (SUB1) located

36 Trends in Plant Science, January 2016, Vol. 21, No. 1

on chromosome 9, which is associated with submergence tolerance [66]. SUB1 was further
mapped to a 0.16 cM region [67] and fine-mapped to 0.075 cM on chromosome 9 [68]. The
positional cloning of SUB1 locus reveals three genes, SUB1A, SUB1B, and SUB1C, the former
being found in tolerant lines and the latter two only in intolerant lines. SUB1A has two allelic
forms, SUB1A-1 being associated with tolerant lines and SUB1A-2 with intolerant lines [68].
SUB1A-1 allele has been introgressed using marker-aided backcrossing into locally adapted
and widely grown rice cultivars in Asia, which were released as ‘Swarna-Sub1’ in India,
Indonesia, and Bangladesh; ‘BR11-Sub1’ in Bangladesh; and ‘IR64-Sub1’ in the Philippines
and Indonesia [69]. These cultivars had significant impact on rice growing in Asia; for example,
Swarna-Sub1 submerged for 7 to 14 days across 128 villages of Odisha offers a 45% increase in
yields over the popular cultivar ‘Swarna’, with no yield penalty under non-flooding [70].

Upland rice is one of the major rice production systems worldwide, and drought stress during
the crop season often adversely impact on rice production. The landraces ‘Aus 257’, ‘Aus Bak
Tulsi’, ‘Azucena’, ‘Basmati 370’, ‘Dular’, ‘Kalia’, ‘Kali Aus’, ‘Lal Aus’, and ‘N22’ were the source
for grain yield under drought stress, which led to the development and release of 17 high-yielding
drought-tolerant rice cultivars in Asia and Africa. Moreover, landraces contributed 14 large-effect
QTLs associated with drought adaptation, of which six QTLs were effective in multiple genetic
backgrounds and production environments. Pyramiding of large-effect QTLs improved drought
adaptation of widely grown rice cultivars such as Swarna (India, Nepal and Bangladesh), ‘IR64’
(many countries in South and Southeast Asia), ‘Sabitri’ (Nepal), ‘TDK1’ (Laos), and ‘BR11’
(Bangladesh). Furthermore, introgressed offspring bearing SUB1 and drought-tolerant QTL
alleles on Swarna, TDK1, and IR64 backgrounds are being agronomically evaluated in some
countries in Asia [71].

Phosphorus (P) deficiency is widespread in tropical soils [72]. The rice landrace ‘Kasalath’ is
highly tolerant to P deficiency [73]. Phosphate uptake 1 (Pup1) on chromosome 12 increases
P uptake [74], which confers significant grain yield advantage in P-deficient soils. It is found in
landraces or cultivars adapted to drought-prone environments [75]. Pup1 is effective in different
genetic backgrounds and environments, and introgressed lines containing Pup1 allele signifi-
cantly increased grain yield on P-deficient soils [76]. A study on the functional mechanism
revealed the presence of a Pup1-specific protein kinase gene, named phosphorus starvation
tolerance 1 (PSTOL1), which is absent from the rice reference genome and P-starvation-
intolerant cultivars. The overexpression of POSTL1 in such cultivars significantly enhances grain
yield in P-deficient soils. POSTL1 promotes early root growth, thereby enabling plants to acquire
more P and other nutrients [77].

Salt stress adversely impact on rice productivity in rainfed and irrigated agro-ecosystems. Rice
landraces ‘Nona Bokra’ and ‘Pokkali’ are excellent source of salt tolerance. Nona Bokra
contributed major QTLs for shoot K+ concentration on chromosome 1 (qSKC-1) and for shoot
Na+ concentration on chromosome 7 (qSNC-7); qSKC-1 encodes a sodium transporter that
control K+ homeostasis under salt stress [78]. Pokali contributed another major QTL, Saltol1,
associated with Na/K ratio and salinity tolerance [79]. Further research revealed that Saltol1 is a
complex locus, with multiple Pokkali alleles regulating shoot Na+/K+ homeostasis [80]. The
Saltol1 region was recently mapped with a SNP set on chromosome 1. The SNPs associated
with Na+/K+ ratio were in complete linkage disequilibrium [81]. Another rice landrace ‘Hasawi’
contributed many QTL alleles, some with major effects associated with increased grain yield
and salinity tolerance (at reproductive stage) across different genetic backgrounds [82].
Pyramiding of Sub1 and Saltol1 is in progress to develop flood- and salinity-tolerant rice
cultivars [17]. Similarly, the barley landrace ‘TX9425’ contributed two QTLs for drought
tolerance (accounting for 42% and 14% of variation) and one QTL for salinity tolerance
(explaining 29% of variation) [83].

Trends in Plant Science, January 2016, Vol. 21, No. 1 37

Boron (B) toxicity has potential to limit crop production, and tolerance to toxicity is associated
with the ability to maintain low B concentrations in the shoot [84]. Barley landrace ‘Sahara 3763’
is highly tolerant to B toxicity [85]. A gene associated with tolerance to boron toxicity, Bot1, a
BOR1 ortholog, was delimited to a 0.15 cM interval between markers xBM178 and xBM162
which encodes Bot1 protein [86]. Bot1 was introgressed into commercial barley cultivars using
linked markers. The introgressed lines show lower yield than the recipient cultivars, which have
been further tailored to develop lines carrying recombination events directly adjacent to Bot1 for
use in barley breeding [86]. A B-tolerant landrace ‘G61450’ contributed the B toxicity gene Bo4,
which was mapped on chromosome 4AL in bread wheat [87]. Preharvest sprouting (PHS) often
reduces grain yield and end-use quality of wheat flour [88]. Landraces from China showed strong
seed dormancy and resistance to PHS. ‘Wanxianbaimaizi’ contributed two major QTLs that
could be used as source alleles to incorporate strong seed dormancy and resistance to PHS in
new cultivars [89].

The discovery of abiotic stress tolerant alleles in landraces of barley, maize, rice, and wheat
clearly shows the importance of conserving and exploring landrace germplasm as a means to
identify agronomically beneficial alleles for enhancing adaptation and productivity in stress-prone
environments.

Grain Yield-Enhancing Alleles in Rice

The productivity of many food crops has stagnated or is not keeping pace with the growing
global food demand. Rice is the most important staple. To date, genetic resources have
contributed to the identification and isolation of at least nine genes associated with grain
yield-enhancing traits in rice [90]. The discovery and deployment of new grain yield-enhancing
genes together with known yield-enhancing alleles are expected to provide additional sources
for raising cereal productivity. SPIKELET NUMBER (SPIKE), identical to NARROW LEAF1
(NAL1), was identified in the tropical japonica rice landrace ‘Daringan’ [90]. SPIKE enhances
grain productivity in indica cultivars through pleiotropic effects on plant architecture. Near-
isogenic lines (NILs) derived from ‘IR64’ and including SPIKE showed increased spikelet
number, leaf size, root system, and number of vascular bundles, and achieved 13–36% grain
yield increase without any negative effect on grain appearance. Uniform expression of SPIKE in
panicles, leaves, roots, and culms support its pleiotropic effect on plant architecture. Further-
more, SPIKE increased grain yield by 18% in the recently released indica cultivar ‘IRRI146’, and
increased spikelet number in the genetic background of other popular indica cultivars. Grain
weight is a major determining factor for rice yield. THOUSAND-GRAIN WEIGHT 6 (TGW6) was
isolated from landrace ‘Kasalath’ [91]. TGW6 encodes a novel protein with indole-3-acetic acid
(IAA)-glucose hydrolase activity. A loss of function Kasalath allele enhances grain weight through
pleiotropic effects on source organs and leads to significant grain yield increase.

Lessons from Barley Landraces for Breeding in Stress-Prone Environments

Landraces have been grown continuously since domestication, often without inputs. After the
adoption of modern cultivars, landraces of several crops moved into unfavorable and stress-
prone environments [92]. Hence, their evaluation could teach a few lessons about adaptation to
low-input and stressful environments [93]. Such lesson(s) may be useful where landraces are still
predominant [94]. This hypothesis was tested with barley grown in the dry areas of the Fertile
Crescent.

The evaluation of pure lines extracted from Syrian and Jordanian barley landraces [95] revealed a
high degree of seed dormancy and variability in the requirement for vernalization. Genetic
variation was further found for seed color (Figure 1B), growth habit, days to heading, culm length,
leaf width, awn length, early growth vigor, lodging, and powdery mildew resistance [34]. Thus,
these landraces showed adaptive traits such as vernalization and seed dormancy. They harbor

38 Trends in Plant Science, January 2016, Vol. 21, No. 1

considerable functional diversity. Additional findings were related to within-landrace genetic Outstanding Questions
variation noted in stress-prone sites, and most landraces outyielded modern cultivars in stressful Why do global public or private breed-
environments, as confirmed recently in Spain and Iran [96,97], although modern cultivars ing programs not use landraces
extensively?
outyielded landraces, though not always significantly, under non-stress conditions. There
was also a large and significant variability for almost all traits both between and within collection To what extent will modern plant
sites. breeding be able to adapt crops to
climate change by pursuing genetic
uniformity? Or do we need a return
Heterogeneity did not decrease in most adapted landraces after thousands of years of natural to diversity?
and human selection in a stressful environment. Therefore, the variability available within the
population appears itself to be an important adaptive trait in stressful environments and under What are the thresholds between food
safety and food security?
low inputs. The concept of using cultivar mixtures as a means to increase stability and maintain
evolutionary potential is increasing because it seems to be a sound strategy to cope with climate Will an agro-ecological mode of pro-
change [98]. Given that the major strength of landraces is their adaptation to the area where they duction be able to feed the 9 billion
evolved, a strategy to increase their yield potential is to consider them as recipients, rather than people projected for 2050?

as donors, of genes which improve their performance without altering their adaptation.
How can we provide healthy food at a
price affordable to all people?
Landraces fit into the farming systems of their area of adaptation, and are often essential
components in the diet: using their genes in formal breeding programs addressing either
nutritional or adaptation issues, and maintaining them through on-farm management, is an
obligation toward the many farmers who have maintained landraces over millennia [3].

Concluding Remarks and Future Perspectives

About 70% additional food production from the same arable lands currently used for agriculture
will be necessary to feed 9 billion people by 2050. This land-use intensification will occur mostly
in the developing worldi. Drought, salinity, low soil fertility, and heat stress rank among the most
important abiotic constraints, which together have a significant negative impact on agricultural
productivity [99]. Plant breeding for adaptation to abiotic stress-prone environments remains a
challenging task. Genetic enhancement of crops needs to identify genetic variation that would
permit plants to tolerate the expected levels of abiotic stress beyond those experienced by the
present-day agriculture; adopt new tools to dissect the genetic, physiological and molecular
basis of stress tolerance; identify genetic tags associated with enhanced tolerance; and inte-
grate these genetic tags in applied breeding to develop ‘smart’ crop cultivars that yield more with
low inputs.

Landraces, given their more than a millennium of evolutionary history and adaptation to stressful
environments, are ideal genetic resources to explore novel genetic variation that overcomes
challenges to crop production. Landraces are heterogeneous with variable phenology, low to
moderate but stable edible yield, and are often nutritionally superior. The genebanks of the
CGIAR (Consortium of International Agricultural Research Centers) and the World Vegetable
Center hold over 791 565 cereal and legume germplasm accessions, of which about 35% are
landraces (accessed on March 6, 2015ii). Traditional agricultural production systems in the past
have played a vital role in the evolution and conservation of on-farm diversity, allowing farmers to
circumvent crop failure by reducing vulnerability to environmental stresses. A systematic
evaluation of landraces for assessing the pattern of diversity is urgently needed to identify
alleles for enhancing yield and adaptation to abiotic stress for raising the productivity of the staple
food crops in stressful environments.

Acknowledgments
S.L.D. is grateful to Ishrath Durafsha of the Knowledge Sharing and Innovation Program of ICRISAT for arranging reprints on
landrace germplasm as a valuable resource for enhancing abiotic stress adaptation and grain yield in cereal and legume
crops. We are grateful to M. Govindraj, pearl millet breeder at ICRISAT, for providing a field picture of a cultivar widely grown
in India that is derived from ‘Iniadi’ landrace pearl millet germplasm.

Trends in Plant Science, January 2016, Vol. 21, No. 1 39

Resources
i
www.fao.org/fileadmin/templates/wsfs/docs/expert_paper/How_to_Feed_the_World_in_2050.pdf
ii
www.genesys-pgr.org

References
1. Khoury, C.K. et al. (2014) Increasing homogeneity in global food 24. Andrews, D.J. and Anand Kumar, K. (1996) Use of the West
supplies and the implications for food security. Proc. Natl. Acad. African pearl millet landrace Iniadi in cultivar development. Plant
Sci. U.S.A. 111, 4001–4006 Genet. Resour. Newslett. 105, 15–22
2. Gepts, P. (2006) Plant genetic resources conservation and utiliza- 25. Rai, K.N. et al. (1990) Registration of ICTP 8203 pearl millet. Crop
tion: the accomplishment and future of a societal insurance policy. Sci. 30, 959
Crop Sci. 46, 2278–2292 26. Mula, R.P. et al. (2010) Case study of adoption of a pearl millet
3. Ceccarelli, S. and Grando, S. (2000) Barley landraces from the variety in a non-target region. SAT eJ. 8, 1–5 (ejournal.ICRISAT.
Fertile Crescent: a lesson for plant breeders. In Genes in the Field: org/Volume 8/sorghum_millets/case_study_of.pdf)
On-Farm Conservation of Crop Diversity (Brush, S.B., ed.), pp. 27. Yadav, O.P. and Bidinger, F.R. (2007) Utilization, diversification
51–76, International Development Research Center and improvement of landraces for enhancing pearl millet produc-
4. Sarker, A. and Erskine, W. (2006) Recent progress in the ancient tivity in arid environments. Ann. Arid Zone 46, 49–57
lentil. J. Agric. Sci. 144, 19–29 28. Harlan, J.R. and Zohary, D. (1966) Distribution of wild wheat's and
5. Rodriguez, M. et al. (2008) Genotype by environment interactions barley. Science 153, 1074–1080
in barley (Hordeum vulgare L): different responses of landraces, 29. Russell, I.K. et al. (2011) Analysis of >1000 single nucleotide
recombinant inbred lines and varieties to Mediterranean environ- polymorphisms in geographically matched samples of landrace
ment. Euphytica 163, 231–247 and wild barley indicates secondary contact and chromosome
6. Abay, F. and Bjørnstad, A. (2009) Specific adaptation of barley level differences in diversity around domestication genes. New
varieties in different locations in Ethiopia. Euphytica 167, 181–195 Phytol. 191, 564–578
7. Frison, E.A. et al. (2011) Agricultural biodiversity is essential for a 30. Newton, A.C. et al. (2010) Cereal landraces for sustainable agri-
sustainable improvement in food and nutrition security. Sustain- culture. A review. Agron. Sust. Dev. 30, 237–269
ability 3, 238–253 31. Ceccarelli, S. (1994) Specific adaptation and breeding for marginal
8. Esquinas-Alcázar, J. (2010) Protecting crop genetic diversity for conditions. Euphytica 77, 205–219
food security: political, ethical and technical challenges. Nature 6, 32. Acevedo, E. and Ceccarelli, S. (1989) Role of physiologist breeder
946–953 in a breeding program for drought resistance conditions. In
9. World Conservation Monitoring Centre (1992) . In Global Biodiver- Drought Resistance in Cereals (Baker, F.W.G., ed.), pp. 117–
sity: Status of the Earth's Living Resources (Groombridge, B., ed.), 139, CAB International
Chapman & Hall 33. Comadran, J. et al. (2011) Mixed model association scans of multi-
10. Hammer, K. et al. (1996) Estimating genetic erosion in landraces – environmental trial data reveal major loci controlling yield and yield
two case studies. Genet. Resour. Crop. Evol. 43, 329–336 related traits in Hordeum vulgare in Mediterranean environments.
11. Lopez, P.B. (1994) A new plant disease: uniformity. CERES 26, Theor. Appl. Genet. 122, 1363–1373
41–47 34. Ceccarelli, S. et al. (1991) Breeding for yield stability in unpredict-
12. Ceccarelli, S. (2012) Landraces: importance and use in breeding able environments: single traits, interaction between traits, and
and environmentally friendly agronomic systems. In Agrobiodiver- architecture of genotypes. Euphytica 56, 169–185
sity Conservation: Securing the Diversity of Crop Wild Relatives 35. ICARDA (2007) ICARDA Annual Report 2006, International Center
and Landraces (Maxted, N. et al., eds), pp. 103–117, CAB Inter- for Agricultural Research in the Dry Areas
national 36. Lopes, M.S. et al. (2015) Exploiting genetic diversity from land-
13. Frison, A.E. et al. (2011) Agricultural biodiversity is essential for a races in wheat breeding for adaptation to climate change. J. Exp.
sustainable improvement in food and nutrition security. Sustain- Bot. Published online March 28, 2015. http://dx.doi.org/10.1093/
ability 3, 238–253 jxb/erv122
14. Van de Wouw, M. et al. (2010) Genetic diversity trends in twentieth 37. Shewayrga, H. and Sopade, P.A. (2011) Ethnobotany, diverse
century crop cultivars: a meta-analysis. Theor. Appl. Genet. 120, food uses, claimed health benefits and implications on conserva-
1241–1252 tion of barley landraces in North Eastern Ethiopia highlands. J.
15. Rasmusson, D.C. and Phillips, R.L. (1997) Plant breeding prog- Ethnobiol. Ethnomed. 7, 19
ress and genetic diversity from de novo variation and elevated 38. Uarrota, V.G. et al. (2011) Maize landraces (Zea mays L.): a new
epistasis. Crop Sci. 37, 303–310 prospective source for secondary metabolite production. Int. J.
16. Sharma et al. (2013) Pre-breeding for diversification of primary Agric. Res. 6, 218–226
gene pool and genetic enhancement of grain legumes. Front. Plant 39. Privita Edwina, R.G.E. et al. (2014) Biochemical and antioxidant
Sci. 4, 309 activities of pigmented landraces of Oryza sativa – Koraput District,
17. Mackill, D.J. et al. (2012) Development and rapid adoption of sub- Odisha, India. Int. Food Res. J. 21, 1941–1949
mergence-tolerant (Sub1) rice varieties. Adv. Agron. 155, 299–352 40. Bhat, F.M. and Riar, C.S. (2015) Health benefits of traditional rice
18. Ruiz Corral et al. (2008) Climatic adaptation and ecological varieties of temperate regions. Med. Aromat. Plants 4, 198
descriptors of 42 Mexican maize races. Crop Sci. 48, 1502–1512 41. Mamidi, S. et al. (2011) Investigation of the domestication of
19. Prasanna, B.M. (2012) Diversity in global maize germplasm: char- common bean (Phaseolus vulgaris) using multilocus sequence
acterization and utilization. J. Biosci. 37, 843–855 data. Funct. Plant Biol. 38, 953–967

20. Meseka, S. et al. (2013) Introgression of alleles from maize land- 42. Sharma, P.N. et al. (2013) Genetic diversity of Indian common
races to improve drought tolerance in an unadapted germplasm. beans elucidated with two germplasm collections and by mor-
J. Crop. Improv. 27, 96–112 phological and microsatellite markers. Plant Genet. Resour. 11,
121–130
21. Meseka, S. et al. (2015) Exploitation of beneficial alleles from maize
(Zea mays L.) landraces to enhance performance of an elite variety 43. Avila, T. et al. (2012) Genetic diversity of bean (Phaseolus) land-
in water stress environments. Euphytica 201, 149–160 races and wild relatives from the primary center of origin of the
Southern Andes. Plant Genet. Resour. 10, 83–92
22. Kholová, J. et al. (2013) Drought stress characterization of post-
rainy (rabi) sorghum in India. Field Crops Res. 141, 38–46 44. Díaz, L.M. et al. (2011) Genetic diversity of Colombian landraces
of common bean as detected through the use of silver-stained
23. Rakshit, S. et al. (2012) Morphological and molecular diversity
and fluorescently labeled microsatellites. Plant Genet. Resour. 9,
reveal wide variability among sorghum Maldandi landraces from
86–96
India. J. Plant Biochem. Biotechnol. 21, 145–156

40 Trends in Plant Science, January 2016, Vol. 21, No. 1

45. Blair, M.W. et al. (2010) Extensive diversity and inter-genepool 70. Dar, M.H. et al. (2013) Flood tolerant rice reduces yield variability
exchange of phaseolin alleles found in world-wide snap bean and raises expected yield, differentially benefitting socially disad-
germplasm analyzed with AFLP and microsatellite markers. Theor. vantaged groups. Sci. Rep. 3, 3315
Appl. Genet. 120, 1381–1391 71. Kumar, A. et al. (2014) Breeding high-yielding drought-tolerant
46. Blair, M.W. et al. (2010) Inter-genepool introgression, genetic rice: genetic variations and conventional and molecular
diversity and nutritional quality of common bean (Phaseolus vul- approaches. J. Exp. Bot. 65, 6265–6278
garis L.) landraces from Central Africa. Theor. Appl. Genet. 121, 72. Sanchez, P.A. and Salinas, J.G. (1981) Low input technology for
237–248 managing oxisols and utisols in tropical America. Adv. Agron. 34,
47. López-Marín, H.D. et al. (2009) Quantitative trait loci for root 279–406
morphology traits under aluminum stress in common bean (Pha- 73. Wissuwa, M. and Ae, N. (2001) Genotypic variation for tolerance to
seolus vulgaris L.) Theor. Appl. Genet. 119, 449–458 phosphorus deficiency in rice and the potential for exploitation in
48. Ramaekers, L. et al. (2010) Improving phosphorus uptake effi- rice improvement. Plant Breed. 120, 43–48
ciency of crop plants. Field Crops Res. 117, 169–176 74. Wissuwa, M. et al. (2002) Substitution mapping of the Pup1: a
49. Cichy, K.A. et al. (2009) QTL analysis of root architecture traits and major QTL increasing phosphorus uptake of rice from a phospho-
low phosphorus tolerance in an Andean bean population. Crop rus deficient soil. Theor. Appl. Genet. 105, 890–897
Sci. 49, 59–68 75. Chin, J.H. et al. (2010) Development and application of gene-
50. Schmutz, J. et al. (2014) A reference genome for common bean based markers for the major rice QTL Phosphorus uptake1. Theor.
and genome-wide analysis of dual domestications. Nat. Genet. Appl. Genet. 120, 1073–1086
46, 707–713 76. Chin, J.H. et al. (2011) Developing rice with high yield under
51. Blair, M.W. et al. (2013) Race structure in the Mexican collection of phosphorus deficiency: Pup1 sequence to application. Plant
common bean landraces. Crop Sci. 53, 1517–1528 Physiol. 156, 1202–1216
52. Chavarro, M.C. and Blair, M.W. (2010) QTL analysis of heat and 77. Gamuyao, R. et al. (2012) The protein kinase Pstol1 from tradi-
drought tolerance in an inter-genepool common bean population. tional rice confers tolerance of phosphorus deficiency. Nature 488,
Trop. Plant Biol. 3, 204–218 535–541
53. Mukeshimana, G. et al. (2014) Identification of quantitative trait loci 78. Ren, Z.H. et al. (2005) A rice quantitative trait locus for salt
associated with drought tolerance in common bean using SNP tolerance encodes a sodium transporter. Nat. Genet. 37,
markers. Crop Sci. 54, 923–938 1141–1146
54. Blair, M.W. et al. (2012) Development of a Mesoamerican intra- 79. Bonilla, P. et al. (2002) RFLP and SSLP mapping of salinity toler-
genepool genetic map for QTL detection in a drought tolerant  ance genes in chromosome 1 of rice (Oryza sativa L.) using
susceptible common bean (Phaseolus vulgaris L.) cross. Mol. recombinant inbred lines. Philipp. Agric. Sci. 85, 68–76
Breed. 29, 71–88 80. Thompson, M.J. et al. (2010) Characterizing the Saltol quantitative
55. Asfaw, A. and Blair, M.W. (2012) Quantitative trait loci for rooting trait locus for salinity tolerance in rice. Rice 3, 148–160
pattern traits of common beans grown under drought stress 81. Kumar, V. et al. (2015) Genome-wide association mapping of
versus non-stress conditions. Mol. Breed. 30, 681–695 salinity tolerance in rice (Oryza sativa). DNA Res. Published online
56. Asfaw, A. et al. (2012) Multi-environment quantitative trait locus January 27, 2015. http://dx.doi.org/10.1093/dnares/dsu046
analyses for photosynthate acquisition, accumulation and remo- 82. Bimpong, I.K. et al. (2014) New quantitative trait loci for enhancing
bilization traits in a common bean. Genes Genomes Genet. 2, adaptation to salinity in rice from Hasawi, a Saudi landrace into
579–595 three African cultivars at the reproductive stage. Euphytica 200,
57. Muñoz, L.C. et al. (2004) Introgression in common bean  tepary 45–60
bean interspecific congruity-backcross lines as measured by 83. Fan, Y. et al. (2015) Using QTL mapping to investigate the
AFLP markers. Crop Sci. 44, 637–645 relationships between abiotic stress tolerance (drought and
58. Lasky, J.R. et al. (2015) Genotype-environment associations in salinity) and agronomic and physiological traits. BMC Genomics
sorghum landraces predict adaptive traits. Science Adv. 1, 16, 43
e1400218 84. Yau, S.K. and Ryan, J. (2008) Boron toxicity tolerance in crops: a
59. Yamasaki, M. et al. (2005) A large-scale screen for artificial selec- viable alternative to soil amelioration. Crop Sci. 48, 854–865
tion in maize identifies candidate agronomic loci for domestication 85. Nable, R.O. (1988) Resistance to boron toxicity amongst several
and crop improvement. Plant Cell 17, 2859–2872 barley and wheat cultivars: a preliminary examination of the resis-
60. Liu, K. et al. (2003) Genetic structure and diversity among maize tance mechanism. Plant Soil 112, 45–52
inbred lines as inferred from DNA microsatellites. Genetics 165, 86. Sutton, T. et al. (2007) Boron-toxicity tolerance in barley
2117–2128 arising from efflux transporter amplification. Science 318,
61. Maron, L.J. et al. (2013) Aluminum tolerance in maize is associated 1446–1449
with higher MATE1 gene copy number. Proc. Natl. Acad. Sci. U.S. 87. Paull, J.G. et al. (1992) Physiological and genetic control of the
A. 110, 5241–5246 tolerance of wheat to high concentrations of boron and implica-
62. Yan, J. et al. (2011) Association mapping for enhancing maize (Zea tions for plant breeding. Plant Soil 146, 45–52
mays L.) genetic improvement. Crop Sci. 51, 433–449 88. Xiao, S.H. et al. (2002) Germplasm improvement for pre-harvest
63. Vielle-Calzada, J-P. et al. (2009) The Palomero genome suggests sprouting resistance in Chines white-grained wheat. An overview
metal effects on domestication. Science 326, 1078 of the current strategy. Euphytica 126, 35–38
64. Ortiz, R. et al. (2010) Conserving and enhancing maize genetic 89. Haiping, Z. et al. (2011) Investigation of main loci contributing to
resources as global public goods – a perspective from CIMMYT. strong seed dormancy of Chinese wheat landrace. J. Agric. Bio-
Crop Sci. 50, 13–28 technol. 19, 270–277
65. Huang, X. et al. (2010) Genome-wide association studies of 14 90. Fujita, D. et al. (2013) NAL1 allele from a rice landrace greatly
agronomic traits in rice landraces. Nat. Genet. 42, 961–965 increases yield in modern indica cultivars. Proc. Natl. Acad. Sci. U.
66. Xu, K. and Mackill, D.J. (1996) A major locus for submergence S.A. 110, 20431–20436
tolerance mapped on rice chromosome 9. Mol. Breed. 2, 219–224 91. Ishimaru, K. et al. (2013) Loss of function of the IAA-glucose
67. Xu, K. et al. (2000) A high-resolution linkage map in the vicinity of hydrolase gene TGWS enhances rice grain weight and increases
the rice submergence tolerance locus Sub1. Mol. Gen. Genet. yield. Nat. Genet. 45, 707–711
263, 681 92. Li, J. et al. (2012) Farmers’ adoption of maize (Zea mays L.)
68. Xu, K. et al. (2006) Sub1A is an ethylene responsive-factor-like hybrids and the persistence of landraces in Southwest China:
gene that confers submergence tolerance to rice. Nature 442, implications for policy and breeding. Genet. Resour. Crop Evol.
705–708 59, 1147–1160

69. Bailey-Serres, J. et al. (2010) Submergence tolerant rice: SUB1's 93. Ceccarelli, S. (1984) Utilization of landraces and H. spontaneum in
journey from landrace to modern cultivar. Rice 3, 138–147 barley breeding for dry areas. Rachis 3, 8–11

Trends in Plant Science, January 2016, Vol. 21, No. 1 41

94. Weltzien, E. (1982) Observation on the growth habit of Syrian and 97. Mohammadi, R. et al. (2014) Adaptation patterns and yield stability
Jordanian landraces of barley. Rachis 1, 6–7 of durum wheat landraces to highland cold rainfed areas of Iran.
95. Weltzien, E. and Fischbeck, G. (1990) Performance and variability Crop Sci. 54, 944–954
of local barley landraces in Near-Eastern environments. Plant 98. Ceccarelli, S. (2014) Drought. In Plant Genetic Resources and
Breed. 104, 58–67 Climate Change (Jackson, M. et al., eds), pp. 221–235, CABI
96. Yahiaoui, S. et al. (2014) Spanish barley landraces outperform International
modern cultivars at low-productivity sites. Plant Breed. 133, 99. Dwivedi, S.L. et al. (2013) Food, nutrition and agrobiodiversity
218–226 under global climate change. Adv. Agron. 120, 1–128

42 Trends in Plant Science, January 2016, Vol. 21, No. 1