ArticlePDF Available

Cereal landraces for sustainable agriculture. A review

Authors:

Abstract and Figures

Modern agriculture and conventional breeding and the liberal use of high inputs has resulted in the loss of genetic diversity and the stagnation of yields in cereals in less favourable areas. Increasingly landraces are being replaced by modern cultivars which are less resilient to pests, diseases and abiotic stresses and thereby losing a valuable source of germplasm for meeting the future needs of sustainable agriculture in the context of climate change. Where landraces persist there is concern that their potential is not fully realised. Much effort has gone into collecting, organising, studying and analysing landraces recently and we review the current status and potential for their improved deployment and exploitation, and incorporation of their positive qualities into new cultivars or populations for more sustainable agricultural production. In particular their potential as sources of novel disease and abiotic stress resistance genes or combination of genes if deployed appropriately, of phytonutrients accompanied with optimal micronutrient concentrations which can help alleviate aging-related and chronic diseases, and of nutrient use efficiency traits. We discuss the place of landraces in the origin of modern cereal crops and breeding of elite cereal cultivars, the importance of on-farm and ex situ diversity conservation; how modern genotyping approaches can help both conservation and exploitation; the importance of different phenotyping approaches; and whether legal issues associated with landrace marketing and utilisation need addressing. In this review of the current status and prospects for landraces of cereals in the context of sustainable agriculture, the major points are the following: (1) Landraces have very rich and complex ancestry representing variation in response to many diverse stresses and are vast resources for the development of future crops deriving many sustainable traits from their heritage. (2) There are many germplasm collections of landraces of the major cereals worldwide exhibiting much variation in valuable morphological, agronomic and biochemical traits. The germplasm has been characterised to variable degrees and in many different ways including molecular markers which can assist selection. (3) Much of this germplasm is being maintained both in long-term storage and on farm where it continues to evolve, both of which have their merits and problems. There is much concern about loss of variation, identification, description and accessibility of accessions despite international strategies for addressing these issues. (4) Developments in genotyping technologies are making the variation available in landraces ever more accessible. However, high quality, extensive and detailed, relevant and appropriate phenotyping needs to be associated with the genotyping to enable it to be exploited successfully. We also need to understand the complexity of the genetics of these desirable traits in order to develop new germplasm. (5) Nutrient use efficiency is a very important criterion for sustainability. Landrace material offers a potential source for crop improvement although these traits are highly interactive with their environment, particularly developmental stage, soil conditions and other organisms affecting roots and their environment. (6) Landraces are also a potential source of traits for improved nutrition of cereal crops, particularly antioxidants, phenolics in general, carotenoids and tocol in particular. They also have the potential to improve mineral content, particularly iron and zinc, if these traits can be successfully transferred to improved varieties. (7) Landraces have been shown to be valuable sources of resistance to pathogens and there is more to be gained from such sources. There is also potential, largely unrealised, for disease tolerance and resistance or tolerance of pest and various abiotic stresses too including to toxic environments. (8) Single gene traits are generally easily transferred from landrace germplasm to modern cultivars, but most of the desirable traits characteristic of landraces are complex and difficult to express in different genetic backgrounds.Maintaining these characteristics in heterogeneous landraces is also problematic. Breeding, selection and deployment methods appropriate to these objectives should be used rather than those used for high input intensive agriculture plant breeding. (9) Participatory plant breeding and variety selection has proven more successful than the approach used in high input breeding programmes for landrace improvement in stress-prone environments where sustainable approaches are a high priority. Despite being more complex to carry out, it not only delivers improved germplasm, but also aids uptake and communication between farmers, researchers and advisors for the benefit of all. (10) Previous seed trade legislation was designed primarily to protect trade and return royalty income to modern plant breeders with expensive programmes to fund. As the desirability of using landraces becomes more apparent to achieve greater sustainability, legislation changes are being made to facilitate this trade too. However, more changes are needed to promote the exploitation of diversity in landraces and encourage their use.
Content may be subject to copyright.
“a8259” — 2009/9/14 — 14:36 — page 1 — #1
i
i
i
i
i
i
i
i
Agron. Sustain. Dev. (2009)
c
INRA, EDP Sciences, 2009
DOI: 10.1051/agro/2009032
Review article
Available online at:
www.agronomy-journal.org
for Sustainable Development
Cereal landraces for sustainable agriculture. A review
A.C. Newton1*,T.A
kar2,J.P.Baresel3,P.J.Bebeli4,E.Bettencourt5,K.V.Bladenopoulos6,
J.H. Czembor7,D.A.Fasoula8,A.Kat s i ot i s 9,K.Koutis10,M.Koutsika-Sotiriou10,G.Kovacs11,
H. Larsson12,M.A.A.Pinheiro de Carvalho13,D.Rubiales14,J.Russell1,T.M.M.Dos Santos15,
M.C. Vaz Pat t o 16
1SCRI, Invergowrie, Dundee DD2 5DA, Scotland, UK
2Central Research Institute for Field Crops, PoBox: 226 06042 Ulus- Ankara, Turkey
3Technical University of Munich, Chair of Organic Farming, Alte Akademie 12, 85350 Freising, Germany
4Department of Plant Breeding and Biometry, Agricultural University of Athens, Iera Odos 75, Athens 11855, Greece
5Genetic Resources, Ecophysiology and Plant Bredding Unit, Instituto nacional dos Recursos Biológicos, I.P. (INRB, I.P.), Quinta do Marquês,
2784-505 Oeiras, Portugal
6NAGREF –Cereal Institute 57001, Thermi, Thessaloniki, Greece
7Plant Breeding and Acclimatization Institute – IHAR Radzikow, 05-870 Blonie, Poland
8Agricultural Research Institute, P.O.Box 22016, 1516 Nicosia, Cyprus
9Department of Plant Breeding and Biometry, Agricultural University of Athens, Iera Odos 75, Athens 11855, Greece
10 Laboratory of Genetics and Plant Breeding, Faculty of Agriculture, Aristotle University of Thessaloniki, 54124 Thessaloniki, Greece
11 Department of Genetic Resources and Organic Plant Breeding, Agricultural Research Institute of the Hungarian Academy of Sciences,
Brunszvik u. 2., Martonvasar, 2462, Hungary
12 Swedish University of Agricultural Sciences, Box 104, SE-23053 Alnarp, Sweden
13 ISOPlexis Banco de Germoplasma, BGR, CEM, Universidade da Madeira, 9000-390 Funchal, Portugal
14 Institute for Sustainable Agricultura, CSIC, Alameda del Obispo s/n, Apdo. 4084, 14080 Cordoba, Spain
15 ISOPlexis Banco de Germoplasma, BGR, CEM, Universidade da Madeira, 9000-390 Funchal, Portugal
16 Instituto de Tecnologia Química e Biológica, Apto. 127, 2781-901 Oeiras, Portugal
(Accepted 3 August 2009)
Abstract – Modern agriculture and conventional breeding and the liberal use of high inputs has resulted in the loss of genetic diversity and
the stagnation of yields in cereals in less favourable areas. Increasingly landraces are being replaced by modern cultivars which are less
resilient to pests, diseases and abiotic stresses and thereby losing a valuable source of germplasm for meeting the future needs of sustainable
agriculture in the context of climate change. Where landraces persist there is concern that their potential is not fully realised. Much eort has
gone into collecting, organising, studying and analysing landraces recently and we review the current status and potential for their improved
deployment and exploitation, and incorporation of their positive qualities into new cultivars or populations for more sustainable agricultural
production. In particular their potential as sources of novel disease and abiotic stress resistance genes or combination of genes if deployed
appropriately, of phytonutrients accompanied with optimal micronutrient concentrations which can help alleviate aging-related and chronic
diseases, and of nutrient use eciency traits. We discuss the place of landraces in the origin of modern cereal crops and breeding of elite cereal
cultivars, the importance of on-farm and ex-situ diversity conservation; how modern genotyping approaches can help both conservation and
exploitation; the importance of dierent phenotyping approaches; and whether legal issues associated with landrace marketing and utilisation
need addressing. In this review of the current status and prospects for landraces of cereals in the context of sustainable agriculture, the major
points are the following: (1) Landraces have very rich and complex ancestry representing variation in response to many diverse stresses and
are vast resources for the development of future crops deriving many sustainable traits from their heritage. (2) There are many germplasm
collections of landraces of the major cereals worldwide exhibiting much variation in valuable morphological, agronomic and biochemical
traits. The germplasm has been characterised to variable degrees and in many dierent ways including molecular markers which can assist
selection. (3) Much of this germplasm is being maintained both in long-term storage and on farm where it continues to evolve, both of which
have their merits and problems. There is much concern about loss of variation, identification, description and accessibility of accessions
despite international strategies for addressing these issues. (4) Developments in genotyping technologies are making the variation available in
landraces ever more accessible. However, high quality, extensive and detailed, relevant and appropriate phenotyping needs to be associated with
the genotyping to enable it to be exploited successfully. We also need to understand the complexity of the genetics of these desirable traits in
* Corresponding author: adrian.newton@scri. ac.uk
“a8259” — 2009/9/14 — 14:36 — page 2 — #2
i
i
i
i
i
i
i
i
2 A.C. Newton et al.
order to develop new germplasm. (5) Nutrient use eciency is a very important criterion for sustainability. Landrace material oers a potential
source for crop improvement although these traits are highly interactive with their environment, particularly developmental stage, soil conditions
and other organisms aecting roots and their environment. (6) Landraces are also a potential source of traits for improved nutrition of cereal
crops, particularly antioxidants, phenolics in general, carotenoids and tocol in particular. They also have the potential to improve mineral
content, particularly iron and zinc, if these traits can be successfully transferred to improved varieties. (7) Landraces have been shown to
be valuable sources of resistance to pathogens and there is more to be gained from such sources. There is also potential, largely unrealised,
for disease tolerance and resistance or tolerance of pest and various abiotic stresses too including to toxic environments. (8) Single gene
traits are generally easily transferred from landrace germplasm to modern cultivars, but most of the desirable traits characteristic of landraces
are complex and dicult to express in dierent genetic backgrounds. Maintaining these characteristics in heterogeneous landraces is also
problematic. Breeding, selection and deployment methods appropriate to these objectives should be used rather than those used for high input
intensive agriculture plant breeding. (9) Participatory plant breeding and variety selection has proven more successful than the approach used
in high input breeding programmes for landrace improvement in stress-prone environments where sustainable approaches are a high priority.
Despite being more complex to carry out, it not only delivers improved germplasm, but also aids uptake and communication between farmers,
researchers and advisors for the benefit of all. (10) Previous seed trade legislation was designed primarily to protect trade and return royalty
income to modern plant breeders with expensive programmes to fund. As the desirability of using landraces becomes more apparent to achieve
greater sustainability, legislation changes are being made to to facilitate this trade too. However, more changes are needed to promote the
exploitation of diversity in landraces and encourage their use.
diversity /disease /yield /quality /nutrition /breeding /genotyping /competition /cultivar degeneration /whole-plant field
phenotyping /non-stop selection /adaptive variation
Contents
1 Introduction ........................................................ 3
2 History of cereal landraces . ......................................... 4
3 Diversityand germplasmcollections................................. 5
3.1 Bread wheat landraces diversity . ............................... 5
3.2 Durum wheat landraces diversity ............................... 6
3.3 Barley landraces diversity . . . .................................. 6
3.4 Oat landraces diversity ........................................ 6
3.5 Maize landraces diversity ...................................... 7
4 Genebanks and conservation of cereal landraces . . . ................... 7
4.1 On-farmconservation ......................................... 8
5 Genotyping andphenotyping ........................................ 9
5.1 Genotyping technology . ....................................... 10
5.2 Genotype-phenotypeassociation ............................... 10
5.3 Accurate whole-plant field phenotyping for exploiting variation
within landraces . .............................................. 11
6 Nutrient uptake and utilisation....................................... 11
7 Nutrition and quality . ............................................... 13
8 Bioticandabiotic stress resistanceandtolerance ..................... 14
8.1 Wheatdiseases................................................ 14
8.1.1 Septorialeafblotch.................................... 14
8.1.2 Powderymildew ...................................... 14
8.1.3 FusariumHeadBlight ................................. 14
8.1.4 Bunts andsmuts....................................... 14
8.1.5 Rustdiseases.......................................... 14
8.1.6 Aphids................................................ 15
8.2 Barleydiseases................................................ 15
8.2.1 Powderymildew ...................................... 15
8.2.2 Rustdiseases.......................................... 15
8.2.3 Scald ................................................. 15
8.2.4 Net blotch............................................. 16
8.2.5 Barleystripe .......................................... 16
8.2.6 Commonrootrotand spotblotch....................... 16
8.2.7 The smuts............................................. 16
8.2.8 Fusariumcomplexes................................... 16
8.2.9 Viruses................................................ 16
8.2.10 Aphids................................................ 16
8.3 Disease tolerance.............................................. 16
“a8259” — 2009/9/14 — 14:36 — page 3 — #3
i
i
i
i
i
i
i
i
Cereal landraces for sustainable agriculture. A review 3
8.4 Abioticstresses .......................... 16
8.4.1 Drought tolerance .................... 16
8.4.2 Frosttolerance ...................... 17
8.4.3 Salinitytolerance .................... 17
8.4.4 Acid and alkaline soils and tolerance to heavy metal
toxicity .......................... 17
9 Breeding : conversion of landraces into modern cultivars ....... 18
9.1 Yield-basedselection ....................... 18
9.2 Adaptability ............................ 18
9.3 Conversionintodensity-neutralmoderncultivars ........ 18
9.4 Seeddegradationcultivardegeneration............. 19
10Participatorybreeding.......................... 20
11Legalissues ............................... 21
12Conclusions............................... 22
1. INTRODUCTION
Elite cereal cultivars are derived from a relatively nar-
row germplasm pool and are predominantly well adapted to
high input agriculture. However, climate change will bring
ever greater challenges in response to both biotic and abi-
otic stresses. Together with pressures to move towards more
sustainable agriculture, there is clearly a need to access and
exploit a broader germplasm resource. Cereal landraces are
just such a resource which could be very valuable and yet ap-
parently underutilised in contemporary agriculture. Landraces
have closer anity with modern cultivars than wild species
and can more easily be used as a foundation material in breed-
ing programmes. During the course of the EU COST Action
860 “Sustainable variety development for low-input and or-
ganic agriculture” (2004–2008), a working group on cereal
landraces was formed comprising a variety of scientific ex-
pertise, linked by common agreement on the potential value of
landraces as a resource for contemporary agriculture. In this
paper we therefore bring together a review of recent literature
on multiple aspects of cereal landraces, oering insights and
future direction for their more ecient incorporation and utili-
sation in agriculture for low-input as well for more favourable
environments.
Various definitions of a landrace have evolved since the end
of the 19th century. Owing to their complex nature, Zeven
(1998) concluded that an all-embracing definition cannot be
given. More recently, Camacho Villa et al. (2005) faced with
the challenges this creates for inventory and conservation
purposes, proposed the following working definition: “a lan-
drace is a dynamic population(s) of a cultivated plant that has
historical origin, distinct identity and lacks formal crop im-
provement, as well as often being genetically diverse, locally
adapted and associated with traditional farming systems”.
They additionally recognised that although the above charac-
teristics are commonly present,they are not always present for
any individual landrace.
A distinction is made between landraces and modern, or so
called elite, cultivars, the latter being the result of formal crop
breeding programmes. It is useful to consider the implications
of this from the genetic and breeding points of view. Modern
cereal cultivars are almost always bred to be mono-genotypic
as inbred or pure lines for self-pollinating species, or one-way
hybrids in maize, and thusare genetically homogeneous. They
are bred to exploit high-input environments with increased
yield levels and with an emphasis on broad or wide adaptation.
In the European Union (EU) and other InternationalUnion for
the Protection of New Varieties of Plants (or UPOV (French:
Uion internationale pour la protection des obtentions vtégé-
tales)) countries, modern cultivars are accepted for commer-
cial cultivation after they have passed through the Distinctive-
ness, Uniformity, and Stability (DUS) and Value of Cultivation
and Use (VCU) systems of evaluation trials. In contrast, ce-
real landraces are genetically heterogeneous. They comprise a
large number of distinct homozygous lines in the case of self-
pollinating crops or, in the case of cross-pollinators like maize,
are populations with more heterozygous components. An im-
portant dierence between genetic heterogeneity and genetic
homogeneity is that the former involves genetic competition
among plants, whereas the latter lack such competition.
Landraces were the principal focus of agricultural pro-
duction until the end of the nineteenth century with the ar-
rival of formal plant breeding (Harlan, 1975). Then, in more
favourable environments, gradual replacement during the early
decades of the 20th century by selected component inbred
lines and modern cultivars led to their virtual disappearance.
However, their cultivation persisted in less favourable environ-
ments and despite earlier predictions about their imminent dis-
appearance (Zeven, 1998), landraces still support subsistence
farming worldwide. This persistence is not due to increased
productivity levels (Almekinders et al., 1994), but because of
their increased stability, accomplished through generations of
natural and deliberate selection for valuable genes for resis-
tance to biotic and abiotic stresses and inter-genotypic com-
petition and compensation. The components of a genetically
diverse populationsuch as a landrace, have a disease buering
eect in reducing the pathogen spread (Frankel et al., 1995).
Several important resistance genes were first identified in ce-
real landraces such as the durable mlo mildew resistance gene
in Ethiopian barley landraces (Pianelli et al., 2004) and later
introduced into elite germplasm. In general however, elite ce-
real germplasm has very few durable resistance genes and lit-
tle resistance to emerging or increasingly important diseases
such as Ramularia collo-cygni, Fusarium disease complexes
and rusts in general, a good example being the new race of
stem rust (Puccinia graminis) Ug99. Therefore exploitation of
“a8259” — 2009/9/14 — 14:36 — page 4 — #4
i
i
i
i
i
i
i
i
4 A.C. Newton et al.
landrace diversity to identify relevant resistance genes is an
important goal. In addition, many cereal landraces are known
to possess broad natural variation in valuable neutraceuti-
cals, which has been inadvertently impoverished during the
breeding of modern cultivars (Monasterio and Graham, 2000;
Murphy et al., 2008).
The important issue of increased yield stability attributed
to the inherent heterogeneity of landraces (e.g., Simmonds,
1979) is nicely described by Zeven (1998) as: “Yield stability
of landraces under traditional low input agricultural systems
is due to the fact that whatever the varying biotic and abiotic
stress for each plant, one or more genotypes within the lan-
drace population will yield satisfactorily”. This emphasises the
urgency for solutions to combat production problems encoun-
tered through lack of spatial and temporal uniformity, particu-
larly in stress-prone or marginal environments. A longer-term
breeding goal, described below, would include the incorpora-
tion of those desirable genes found in dierent landrace geno-
types into one or few superior plants with elevated stability
levels. However, a single genotype must express greater trait
plasticity to compensate for its inability to exploit heterogene-
ity in spatial and temporal dimensions as flexibly as multiple
genotypes (Newton et al., 2009).
Awareness that future increases in productivity may de-
pend on improved yields in the high-stress environments,
has focussed breeding on specific or narrow adaptation and
on the need for conservation of genetic diversity (Cleveland
et al., 1999). Conventional plant breeding has been success-
ful in favourable environments and those which can be made
favourable through interventions, but it is less successful in
low-input environments, characterized by increased G×Ein-
teractions, or in organic farming systems. In these cases, par-
ticipatory breeding approaches with the involvement of local
farmers can have a significant and positive influence, as will
be discussed. A renewed focus on cereal landraces for breed-
ing purposes could ameliorate some negative consequences of
modern agriculture and conventional breeding, such as the lib-
eral use of high inputs, the loss of genetic diversity (Tilman,
1996), and the stagnation of yields in less favourable areas
(Annichiarico and Pecetti, 1998).
The role of genebanks and on-farm conservation practices
in preserving landrace genetic diversity is increasingly impor-
tant, particularly since the pivotal involvement of local farmers
is continually declining due to the changes brought about by
modern agricultural and socio-economic practices. Studies on
landrace genetic diversity are a prerequisite for ecient con-
servation and management, registration purposes and eective
use of landraces in breeding programmes. The considerable
advances in molecular genotyping and databasing technolo-
gies in recent years are beginning to make the variation and
resources of landraces more accessible for exploitation. High-
throughput genotyping enables genebank accessions with un-
certain provenance to be elucidated and thereby enable valida-
tion of associated phenotypic data, making them much more
useful. In the quest to bridge the phenotype-genotype gap
(Miflin, 2000; Parry and Shewry, 2003) and exploit landrace
variation, recent advances in genotyping and whole-plant field
phenotyping methodology are discussed.
2. HISTORY OF CEREAL LANDRACES
The Fertile Crescent and Turkey (in Asia Minor) is known
as one of the important centres of diversity of many field
crops, particularly wheat and barley, and therefore has con-
siderable genetic diversity. Plant domestication from this re-
gion over thousands of years has also resulted in the develop-
ment of enormous diversity. Progressive adaptation to a wide
range of environments, responding to various selection pres-
sures including biotic, abiotic and human intervention, has re-
sulted in characteristic intra-specific diversity and dierenti-
ation (Teshome et al., 2001) represented by many landraces
with specific histories and eco-geographical origins. To better
understand the importance of these resources it is necessary to
highlight their history.
Archeobotanical data shows that the first domesticated
wheat species were einkorn (Triticum monococcum ssp. mono-
coccum)andemmer(Triticum turgidum ssp. dicoccon). These
evolved from their wild relatives (T. boeticum and T. dic-
occoides respectively) about 10 000 years ago (Heun et al.,
1997). Both species were the staple food of the human pop-
ulation until the end of the Bronze Age when naked Triticum
species became dominant in agricultural lands. In a recent ex-
pedition of the Hellenic Genebank, at the edges of the centre
of diversity in the northern eastern part of Greece in Thrace
natural populations of Triticum monococcum subsp. boeticum
and Aegilops speltoides progenitors of the genomes A and B
were found in co-existence and in situ conservation has been
planned (Kotali et al., 2008).
Durum wheat (T. turgidum durum) has been of great his-
torical significance, because it provided a range of sub-species
that were cultivated widely across the globe for thousands of
years (Feuillet et al., 2007). Durum wheat spread out from the
Fertile Crescent and through southern Europe, reaching north
Africa around 7000 BC (Feldman, 2001). It came into cultiva-
tion originally in the Damascus basin in southern Syria about
9800 BC (Zohary and Hopf, 2000). A second route of migra-
tion occurred through north Africa during the Middle Ages
(Moragues et al., 2006a). The beginning of modern agricul-
ture transformed durum landraces in obsolete ‘cultivars’ which
were gradually replaced by today’s elite cultivars where only
a fraction of the crop diversity is exploited (Feuillet et al.,
2007). Nowadays, wheat landraces are kept and maintained in
germplasm bank collections around the world, but are grown
in practice only in the more marginal agricultural environ-
ments of their origins.
Bread wheat (T. aestivum L. subsp. aestivum L.)
first emerged in cultivated wheat fields approximately 5–
6000 years ago (Zohary and Hopf, 2000), resulting from a hy-
bridization between T. turgidum LandT. tauschii L. The D
genome of bread wheat originated from Aegilops tauschii and
carried alleles adapted to the more continental climate and thus
enabled bread wheat to be cultivated over more extensive geo-
graphic environmentsthan emmer wheat (Feuillet et al., 2007).
The D genome also encodes proteins that restore the softness
of the grain endosperm (Chantret et al., 2005) thereby improv-
ing bread-making properties. Use for leavened bread produc-
tion has contributed to its migration to Europe and subsequent
“a8259” — 2009/9/14 — 14:36 — page 5 — #5
i
i
i
i
i
i
i
i
Cereal landraces for sustainable agriculture. A review 5
widespread cultivation around the world with the development
of several dierent landraces with a diversity of local adapta-
tions driven by dierent climates and agricultural practices.
Hexaploid bread wheat accounts for approximately 90% of
world wheat production today (Feuillet et al., 2007).
Barley (Hordeum vulgare L.) also became domesticated in
the Fertile Crescent, about 10 000 years ago, migrated through
Europe, and local agricultural practices and natural selection
have led to locally adapted landraces (Jaradat et al., 2004).
Valuable information about domestication of cultivated bar-
ley, ecogeographical diversity and relevant issues are found in
the recent book on barley diversity by Bothmer et al. (2003).
In archaeological records, oats, along with rye, were found
as a weed contaminant of wheat and barley samples. As a
cultivated crop oats was present on a significant scale to-
wards the end of prehistory, mainly in northern European re-
gions (Moore-Colyer, 1995). Domestication of oats is com-
plex, with independent domestication occurring at each ploidy
level (Harlan, 1977). Britain was the main area of cultiva-
tion of the hexaploid oats. Until the early 1900’s cultivated
germplasm was landraces adapted to local growing conditions.
Nowadays, a large number of modern cultivars trace back to a
restricted number of landraces, such as ‘Kherson’ from which
14 cultivars developed by re-selection and another 80 had
‘Kherson’ in their pedigrees (Wesenberg et al., 1992).
Maize (Zea mays L. ssp. mays) was domesticated from its
wild progenitor teosinte (Zea mays ssp. parviglumis) through a
single domestication event, in southern Mexico, between 6000
and 9000 years ago (Matsuoka et al., 2002). After this initial
event, introgressions from other teosinte types may have con-
tributed to the maize gene pool and thereby help explain the
remarkable phenotypic and genetic diversity (Matsuoka et al.,
2002). The same authors suggested also that from early di-
versification in the Mexican highlands, two paths or lineages
of dispersal occurred. One path traces through western and
northern Mexico into the south-western USA and then into
the eastern USA and Canada. A second path leads out of the
highlands to the western and southern lowlands of Mexico
into Guatemala, the Caribbean Islands, the lowlands of South
America and finally the Andean mountains. There is evidence
that these two maize germplasm pools were introduced to Eu-
rope at dierent times and locations. The first introduction
came from the Caribbean germplasm, and remained confined
to southern Spain probably due to poor adaptation to the Eu-
ropean conditions as late maturing populations (Brandolini,
1969). From the 17th century onwards, the north American
flint populations, relatively insensitive to day length and with
low temperature requirements for flowering, were introduced
in northern Europe. Since the introduction of maize to Eu-
rope five centuries ago, cultivated populations have evolved
under the dierent selective pressures imposed by the environ-
ment and farmers. Adaptation to many environmental niches
of European countries for many years explains the large vari-
ability and number of landraces which can be observed today
(Gauthier et al., 2002).
Clearly landraces have very rich and complex ancestry rep-
resenting variation in response to many diverse stresses. These
are vast resources for the development of future crops de-
riving many sustainable traits from their heritage. How these
resources can be made accessible and exploited will be ad-
dressed below.
3. DIVERSITY AND GERMPLASM COLLECTIONS
3.1. Bread wheat landraces diversity
Bread wheat landraces are characterized by their diversity
and heterogeneity. However, this genetic diversity needs to be
described and measured if it is to be used eectively in breed-
ing and management of plant genetic resources. Traditionally
used markers for the description of landrace genetic diver-
sity are morphological and agronomic traits known as descrip-
tors and established by the International Plant Genetic Re-
sources Institute (IPGRI) (IPGRI, 1985). Among them plant
height, flowering and anthesis time, spikelet and leaf emer-
gence (Motzo and Guinta, 2007), grain size (Ferrio et al.,
2007), grain yield and weight, spikes per unit area (Moragues
et al., 2006a, 2006b) and harvest yield index (De Vita et al.,
2007) are considered to be the most important ones, and have
been successfully used in the phenotypingof bread wheat lan-
draces. Besides IPGRI descriptors other morpho-physiological
traits have also been used (Autrique et al., 1996; Nachit et al.,
1988; Dencic et al., 2000). Even though morphological de-
scriptors are highly heritable characters and expressed in all
environments (Frankel et al., 1995), they are limited in num-
ber, and agronomic traits are aected by environmental condi-
tions.
In recent years several physiological, biochemical, molecu-
lar and technological traits have been commonly used to char-
acterise bread wheat landraces and to asses their importance
as a resource and for food security. Among the biochemi-
cal and physiological markers, carbon isotope (13C) discrim-
ination (Ferrio et al., 2007), biomass accumulation (De Vita
et al., 2007; Moragues et al., 2006b), storage protein pat-
terns (Gregová et al., 1999, 2004, 2006), High Molecular
Weight (HMW) glutenin subunits polymorphism (Caballero
et al., 2001) and mineral content (Oury et al., 2006) have
been evaluated. Technological parameters, i.e. grain hardness,
starch and protein content, viscosity (Igrejas et al., 2002) have
been applied also, providing dierent approaches to the eval-
uation of landraces diversity and quality. Although biochem-
ical and molecular markers allow fast screening of landrace
diversity to detect useful variation for breeding programmes
(Gregová et al., 2006) or management of the genetic resources
in germplasm collections (Dreisigacker et al., 2005), they are
complementary to but cannot replace morphological charac-
terisation (Moragues et al., 2006a; dos Santos et al., 2008;
Zeven, 1998). Major diculties arise if one attempts to inter-
pret molecular or biochemical data generated from germplasm
accessions in the absence of their morphological characteri-
sation (Gregová et al., 2006; Zeven, 1998). Overall a signif-
icant decrease of genetic diversity has been observed related
to the replacement of bread wheat landraces by elite cultivars
which appears to be associated with loss of some quality traits
such us protein content and glutenins quality (Gregóva et al.,
“a8259” — 2009/9/14 — 14:36 — page 6 — #6
i
i
i
i
i
i
i
i
6 A.C. Newton et al.
2006; Caballero et al., 2001). At the same there exist enormous
gaps in our knowledge that needs to be fulfilled concerning
landraces structure, within and among landraces diversity and
useful traits.
3.2. Durum wheat landraces diversity
Durum wheat is a primary cereal crop in several regions
of the Mediterranean Basin, including the southern peninsular
Italy (Motzo and Giunta, 2007), southern Anatolia of Turkey
(Akar and Özgen, 2007) and southern Spain (Ruiz and Martín,
2000). Durum wheat has a great economic importance due to
the long tradition of pasta making in Italy (De Vita et al., 2007)
and bulgur making in Mediterranean countries (Akar and Öz-
gen, 2007). Archaeological findings suggested that bulgur and
cracked wheat could be distinguished as two basic ingredi-
ents of Mediterranean cooking and bulgur was known at least
since the 3rd millennium BC. (Valamoti, 2002). A screen of
Madeiran and Canary Islands wheat accessions showed that
durum accounts only for 19 and 15% of the germplasm collec-
tions respectively (Andrade et al., 2007).
In these and other regions, durum wheat landraces are cul-
tivated by farmers to a very limited extent (Moragues et al.,
2006a; dos Santos et al., 2008; Ruiz and Martín, 2000; Zhang
et al., 2007; Cherdouh et al., 2005; Teklu et al., 2005; Kebebew
et al., 2001; Ben Amer et al., 2001).
As in the case of bread wheat, durum landraces have been
replaced in modern farming systems by elite cultivars. The
studies performed to evaluate the diversity of durum landraces
are predominantly based on germplasm collections that pre-
serve accessions of populations abandoned by farmers. Such
studies show the existence of considerable crop heterogene-
ity and genetic variability (Medini et al., 2005; Masum Akond
and Watanabe, 2005; Teklu et al., 2005; Queen et al., 2004;
Pagnotta et al., 2004; Alamerew et al., 2004; Kebebew et al.,
2001; Ben Amer et al., 2001). Among the DNA molecu-
lar markers, Randomly Amplified Polymorhic DNA (RAPD)
(Mantzavinou et al. 2005), microsatellites or Simple Sequence
Repeats (SSRs) (Zhang et al., 2007; Hao et al., 2006), Am-
plified Fragment Length Polymorphisms (AFLPs) (Martos
et al., 2005) and Restriction Fragment Length Polymorphisms
(RFLPs) (Autrique et al., 1996) have been used to genotype
durum wheat landraces. However, it is not always easy to
determine which material authors have analysed. For exam-
ple Ruiz and Martín (2000) showed that from 619 durum
wheat entries of the Spanish Plant Genetic Resources Cen-
ter (CRF-INIA (Spanish: Centro de Recursos Fitogenéticos -
Instituto Nacional de Investigación Agraria)) collection, 428
were turgidum and 126 were durum accessions and claimed
that they were representative of durum landrace diversity in
southern Spain. However, no detailed information about the
origin and sampling methodologies were given.
As a result of the increased economic importance of durum
wheat, several breeding programmes aim to develop new cul-
tivars and release old durum landraces (De Vita et al., 2007;
D’Amato, 1989). In southern Italy local durum landraces of
Mediterraneum typicum were cultivated until the beginning of
the 1950 when a process of their replacement by modern culti-
vars began (Motzo and Giunta, 2007; Ruiz and Martín, 2000).
As with bread wheat landraces, there has been a significant
decrease of genetic diversity in this process which appears to
be associated with some loss of some quality traits (Oak et al.,
2004; De Vita et al., 2007). Overall, it can be concluded that
our knowledge on durum landraces need to be improved.
3.3. Barley landraces diversity
Barley landraces show stability under adverse climatic con-
ditions. This is attributed to the heterogeneity present and that
provides them with a buering capacity. The ‘within’ as well
as ‘among’ landraces diversity has been the subject of many
studies, for example with respect to their agronomic traits,
and morphological characters (Assefa and Labuschagne, 2004;
Abdellaoui et al., 2007). Biochemical markers used for as-
sessing diversity (KolodinskaBrantestam et al., 2003) or com-
plementing morphological data include isozymes (Jaradat and
Shahid, 2006) and hordeins (Demissie and Bjørnstad, 1997).
Molecular markers used include restriction fragment length
polymorphisms (RFLPs) (Bjørnstad et al., 1997; Demissie
et al., 1998; Backes et al., 2003), random amplified poly-
morphic DNA (RAPDs) (Abdellaoui et al., 2007; Manjunatha
et al., 2007; Papa et al., 1998) microsatellites or simple se-
quence repeats (SSRs) (Yahiaoui et al., 2008; Feng et al.,2006;
Hamza et al., 2004; Jilal et al., 2008), amplified fragment
length polymorphisms (AFLPs) (van Treuren et al., 2006;
Assefa et al., 2007) and Inter-Sequence Simple Repeats (IS-
SRs) (Kolodinska Brantestam et al., 2004).
A controversial issue of genetic diversity studies in lan-
draces is the loss of genetic diversity overtime, as dierent
trends in genetic diversity changes have been observed for
dierent countries. A study carried out on barley germplasm
derived from the Nordic and Baltic countries using ISSRs
showed that while there were no significant changes of ge-
netic diversity observed during the last century in the north-
ern parts of this geographical region a significant decrease was
observed in the southern parts (Kolodinska Brantestam et al.,
2004). Similar results were obtained when biochemical mark-
ers were employed (Kolodinska Brantestam et al., 2003).
3.4. Oat landraces diversity
Oat landraces or oat mixtures were widely grown in Europe
until early 1900, while in the USA most cultivars developed
up until the 1970s traced back to only seven landraces intro-
duced from Europe (Coman, 1977). Oat landraces had fairly
uniform morphological characters but were heterogeneous for
alleles that conditioned reactions to diseases. The 14 cultivars
developed as a single-plant selection from the uniform mor-
phologically landrace ‘Kherson’ referred to above (Coman,
1977) diered mainly in reactions to diseases and, to a lesser
extend, morphological characters. Furthermore, from a sin-
gle plant selection in a potato field in Cumberland, United
Kingdom, the short straw ‘Potato Oat’ was obtained from
“a8259” — 2009/9/14 — 14:36 — page 7 — #7
i
i
i
i
i
i
i
i
Cereal landraces for sustainable agriculture. A review 7
which a number of sub-cultivar selections were isolated. We
might therefore expect a rather narrow genetic basis to be
present but the number of cultivars developed from ‘Kherson’
and ‘Potato Oats’ indicates otherwise.
In a number of experiments with many genotypes, mor-
phological characters were shown to possess great pheno-
typic plasticity in their environmentalresponse (Diederichsen,
2008; Katsiotis et al., 2009). Diederichsen (2008) tested
10 105 entries, including landraces, over four years in a sin-
gle location and found eight environmentally stable characters,
while during the RESGEN CT99-106 project (Katsiotis et al.,
2009), 1011 entries were tested in four diverse environments
over three seasons but only two stable characters were found,
namely panicle shape and kernel covering. In both cases the
morphological diversity within the oat accessions did not dif-
fer between landraces and modern cultivars.
Using molecular markers non-landrace entries showed the
same polymorphism as landraces, with the majority of the
AFPLP marker variation (89.9%) residing within accessions
of each country, revealing the success of oat breeding pro-
grammes in maintaining genetic diversity within elite cultivars
(Fu et al., 2005). Finally, in a study among Canadian oat culti-
vars released between 1886 and 2001 using 30 SSRs, a range
of increasing and decreasing patterns of allelic changes was
observed at dierent loci and significant allelic decrease was
detected in cultivars released after 1970 coming from specific
breeding programmes (Fu et al., 2005).
3.5. Maize landraces diversity
Maize domestication resulted from a single event involving
its wild progenitor teosinte (Z. mays subspecies parviglumis),
introgression from other teosinte types and the segregation
into two European germplasm pools (see above) between
which much hybridisation occurred. The idea of hybridisa-
tion rather than a slow northward dispersion accompanied by
selection for earliness is supported in the case of Spain and
Portugal where many maize landraces are still cultivated. The
Iberian maize germplasm display no close relationship with
any American types, sharing alleles with both Caribbean and
North American flints (Rebourg et al., 2003; Vaz Patto et al.,
2004). Other studies have shown that maize landraces can be
distinguished by morphologicaland agronomic traits (Pinheiro
de Carvalho et al., 2008; Brandolini and Brandolini, 2001;
Goodman and Paterniani, 1969), biochemical traits such as
zeins (de Freitas et al., 2005) or molecular markers (Reif et al.,
2005; Rebourg et al., 2001; Gauthier et al., 2002). More re-
cently SNP markers have been increasingly applied to study
useful landraces (Tenaillon et al., 2001). These studies also
show that a significant landraces diversity well adapted to
agro-ecological conditions still exist in several countries (Pin-
heiro de Carvalho et al., 2008; Vaz Patto et al., 2007; Bran-
dolini and Brandolini, 2001; Ruiz De Galarreta and Alvarez,
2001).
In summary there are many germplasm collections of lan-
draces of the major cereals worldwide exhibiting much vari-
ation in valuable morphological, agronomic and biochemical
traits. The germplasm has been characterised to variable de-
grees and in many dierent ways including molecular markers
which can assist selection.
4. GENEBANKS AND CONSERVATION
OF CEREAL LANDRACES
Throughout the centuries farmers have been the major
guardians of genetic diversity. The importance of these crop
resources for agriculture and food security was stressed in the
International Agricultural Congress at Rome in 1927 (Zeven,
1998). The extinction of traditional farming systems, the ag-
ing and exodus of rural population, globalisation, and envi-
ronmental degradation, have led to extinction of many cereal
landraces and much of this diversity has been eroded. As a
consequence, during the last century most of this unique ce-
real biodiversity has disappeared and the information regard-
ing traditional cultivars is presently very scarce. According to
FAO (1998), it is estimated that 75% of the genetic diversity of
crop plants was lost in the last century. The erosion of these re-
sources results in a severe threat to the world’s long-term food
security. Although often neglected, the urgent need to conserve
and utilize landraces genetic resources as a safeguard against
an unpredictable future is evident (Hammer et al., 1999).
The first organised attempts to conserve landrace resources
by growing them on farm, (in situ conservation) were made
in Austria, during the 1930s (Zeven, 1996). Nowadays the
germplasm collections are major guardians of landraces diver-
sity. Increasingly the new tasks of these genebanks are related
to the conservation of plant resources and the need to keep ac-
cessions representing the landraces diversity and genetic struc-
ture (van Treuren et al., 2006). Cereal landraces represents
a group of populations, sharing common morphological and
agronomic traits, geographical origin and history, and uses
(Camacho Villa et al., 2005). These populations are often com-
posed of several genotypes,which together make up the cereal
landraces characteristics (Jaradat and Shahid, 2006). For these
reasons the genebanks need to collect or conserve rare alle-
les and avoid genetic drift when accessions are sampled both
during field collection and sample regeneration (Mantzavinou
et al., 2005). Recently, enormous eort has been made to cap-
ture biodiversity being lost by collecting as many germplasm
accessions as possible from dierent geographic regions, espe-
cially from the rich centres of diversity. However, it seems that
in many cases the sampling strategies were inadequate and the
data collected incomplete and scarce, and insucient attention
was paid to ensuring the maintenance of the collected material
throughout the lifetime of the genebank reducing the utility of
the resource (Sackville Hamilton and Chorlton, 1997). Based
on this evidence concern was expressed about the erosion of
genetic diversity of the landraces held in the genebanks. It
became evident that the management and research on bio-
diversity requires renewed approaches (Hammer and Gladis,
1996; Hammer and Spahillar, 1998). This problem is increased
by the scarcity of knowledge about landrace structure or by
the varying understanding of landraces definitions. During
the last few years genebank management procedures, such
“a8259” — 2009/9/14 — 14:36 — page 8 — #8
i
i
i
i
i
i
i
i
8 A.C. Newton et al.
as collecting strategies and techniques, conservation meth-
ods, monitoring viability, and regeneration strategies aiming
at maintenance of integrity and characterisation of the acces-
sions have been improved by IPGRI (2003). However a major
eort is still needed to standardize methodologiesused by dif-
ferent genebanks (Engel and Visser, 2003).
A second major problem is the undetermined level of du-
plicates within and between collections. In order to improve
conservation eciency there is an urgent need to rationalise
collections by identifying and minimising unnecessary dupli-
cation (Dobrovolskayaet al., 2005) and to develop germplasm
core collections, a concept proposed by Frankel and Brown
(1984). The first step in the identification of probable du-
plicates is based on the available passport data (Hintum and
Knüper, 1995), followed by characterisation data such as
agro-morphological, molecular and protein traits. There is an
absence of research dealing with this problem and little infor-
mation is accessible on global information systems. However,
there are good examples of research where germplasm collec-
tions have been evaluated identifying specific agronomic or
quality traits for breeding purposes (Pecetti et al., 2001; Raciti
et al., 2003) or assessing accession duplications. For example,
Ruiz and Aguiriano (2004) confirmed the existence of 90%
of the duplications among 106 cases in durum wheat collec-
tions with 266 accessions, using their gliadin patterns. Agro-
morphological traits and biochemical or molecular markers
are important tools in accession descriptions which will help
to fill in information gaps which would otherwise diminish the
ability to exploit such material by farmers and plant breed-
ers. In principle, the better managed and more comprehen-
sive a collection, the more valuable it is. The European Co-
operative Programme for Plant Genetic Resources (ECP-GR,
http://www.ecpgr.cgiar.org/) is a good example of a success-
ful collaboration which should help ensure the continued pro-
tection and conservation of cereal landraces through standard-
ised procedures, compatible data documentation systems, and
compatibility with European frameworks for better manage-
ment, study and exchange of resources. Valuable collections
of cereal landraces are held at several European and world
genebanks (see World Information and Early Warning System
(WIEWS) on Plant Genetic Resources for Food and Agricul-
ture (PGRFA), http://apps3.fao.org/wiews/wiews.jsp).
The Convention on Biological Diversity (CBD, 1993) rec-
ognizes the contribution of farmers to the conservation and
development of genetic diversity. Sharing of benefits and the
concomitant increased recognition of the value of the re-
sources are the most eective ways to promote conservation
and to ensure the continued availability of plant genetic re-
sources. There is a need to make an economic evaluation of
genebank conservation including the issue of benefit sharing
but this has not received much attention among those formu-
lating legal measures for the implementation of the CBD. Con-
servation involves both preservation and evolution. Therefore,
ex situ preservation alone cannot provide the lasting benefits
that accrue from the conservation of habitats and ecosystems
rich in biodiversity (Swaminathan, 2002). This interaction
among genebanks and those that are motivated to preserve the
traditional seeds will be fundamentalto overcoming the prob-
lems facing genebanks with seed regeneration, such as genetic
drift, contamination and loss.
4.1. On-farm conservation
Two prevalent methods to conserve plant diversity are in
situ and ex situ conservation. In ex situ conservation the ge-
netic resources are conserved outside of their natural habi-
tat (or cultivation territory) in identified genebanks. Unfortu-
nately, in the ex situ conservation methods the variability that
has been collected remains static because the natural evolution
process is not allowed to continue (Dhillon et al., 2004). In
contrast, in situ approaches to conservation are at the level of
ecosystems and natural habitats, and include the maintenance
and recovery of viable population of species in their natural
surroundings, or in the case of domesticated species, in the
surroundings where they have developedtheir distinctive prop-
erties. This approach involves two methods, (1) the genetic re-
serve and (2) on-farm conservation (Hawkes et al., 2002). The
former is defined as location, management and monitoring of
genetic diversity of natural wild populations within defined ar-
eas for active long-term conservation, for example the natural
conservation sites in the Near-East where the wild Triticum
species are conserved in their place of origin. On-farm con-
servation is the sustainable management of genetic diversity
of locally developed traditional crop cultivars along with as-
sociated wild and weedy species or forms within traditional
agricultural systems. Such in situ methods, including on-farm
conservation, have an advantage over ex situ methods since
they provide a natural laboratory for evolution to continue and
help the continued gradual build-up of traits imparting adap-
tation to specific ecogeographical regions and those matching
the requirements of local tribes, communities and populations.
New and more adapted types evolve and thus diversity is aug-
mented. The need for on-farm conservation of original lan-
draces is one of the most important recent questions in plant
genetic resource management (Dhillon et al., 2004).
The ‘European Plant Conservation Strategy’ (Council of
Europe, Planta Europea, 2001), the ‘European Community
Biodiversity Strategy’ (European Commission 2000), the
‘Convention on Biological Diversity’ (Convention on Biolog-
ical Diversity, 1992) and the ‘International Treaty on Plant
Genetic Resources for Food and Agriculture’ (http://www.
planttreaty.org) all stress the need to improve the eciency
of conservation techniques, particularly those related to in
situ conservation of endangered crops and crop wild relatives.
However, little progress has been made on the methodologies
for on farm conservation of plant genetic diversity (Maxted,
2003), especially in case of cereal landraces. Although there is
a general agreement between conservationists on the fact that
landraces should to be conserved on farm, i.e. in the place of
their natural origin (Kovács, 2006b), in practice this is often
very dicult.
In situ conservation programmes were initiated in most
European countries (Maxted, 2003), using participatory ap-
proaches (see below), but with limited success. The main prob-
lems were that the original places are no longer agricultural
“a8259” — 2009/9/14 — 14:36 — page 9 — #9
i
i
i
i
i
i
i
i
Cereal landraces for sustainable agriculture. A review 9
land, that traditional farming systems and knowledge has dis-
appeared in the given region, or that the in situ conservation is
not economic for the farmers and therefore they do not partic-
ipate in such projects (Kovács, 2006b).
Some very successful in situ conservation strategies have
been established in some neighbouring countries. In case of
cereals, Turkey, Israel and other countries of the Fertile Cres-
cent have active genetic conservation programmes (Jaradat
et al., 2004). One of the first projects was established in Am-
miad to conserve wild wheat and wheat and barley diversity in
its place of origin (Anikster and Noy-Meir, 1991). However,
in Europe there are few such good examples. For example,
the AEGRO (AGRI GEN RES 057) project which attempts to
promote an integrative approach for the conservation of crop
wild relatives and landraces in situ and on farm, such us in
the case of Avena landraces. The Scottish landrace protection
scheme was set up in 2006 to compile an Inventory of iden-
tified landraces and traditional cultivars which are still being
grown and used in agriculture in Scotland providing a safety
net for the continued use of landraces by storingseed produced
by each grower each year (N. Green and G. Saddler, SASA,
Edinburgh, personal communication). The farmers are an ag-
ing population and the future of Scottish landraces depends on
the continued regeneration of the landraces, so if seed harvest
fails, the landrace would otherwise be lost.
Clearly much landrace germplasm is being maintained
across the world both in long-term storage in major collec-
tions and on farm where it continues to evolve, both of which
have their merits and problems. There is much concern about
loss of variation, identification, description and accessibility of
accessions despite international strategies for addressing these
issues.
5. GENOTYPING AND PHENOTYPING
Although the diversity within landraces has been demon-
strated to be a powerful means to improve barley yields in
marginal environments in recent times (Ceccarelli, 1996), to
fully realise and utilise the potential of such resources they
need to be accurately and appropriately genotyped and pheno-
typed and the data made readily available in forms which can
be easily interpreted by the plant breeding community.
The genetic structure of landrace collections, when linked
to geographical and environmental data, may reveal genomic
signatures of selection which are valuable information for
breeding for specific environments, farming methods and
users. For example, a collection of landraces was made in
1981 in Syria and Jordan from farmers who had been using
their own seed for generations (Syrian Jordanian landrace col-
lection (SJLC), Weltzien, 1988). These lines, sampled from a
gradient of agroecological conditions, showed a wide range of
responses to drought stress as established by extensive field
trials. Preliminary morphological evaluation revealed consid-
erable variation between and within collection sites for many
agronomically important characters (Ceccarelli et al., 1987)
and disease reactions (van Leur, 1989). Russell et al. (2003)
assayed genetic variability at 21 nuclear and 10 chloroplast
microsatellite loci for a stratified subset of 125 barley landrace
accessions from the SJLC collection, sampled from five dier-
ent ecogeographical regions. Chloroplast polymorphism was
detected, with most variation being attributable to specific dif-
ferences between the five regions. A total of 244 nuclear al-
leles were detected, only 38 of which were common to the
five regions sampled, most variation being within sites. There
were strong associations between the chloroplast and nuclear
SSRs and linkage disequilibrium, the non-random association
of alleles, at both linked and unlinked SSR loci, clearly show-
ing that these landraces have ‘adapted gene complexes’ which
might be advantageous for breeding programmes and the ge-
netic diversity and population structure was clearly driven by
a drought gradient.
In another study of barley, Bjørnstad et al. (1997), com-
pared cultivated accessions from Europe, north America and
Japan with Ethiopian landraces, and found that the Ethiopian
germplasm was significantly less diverse than the cultivated
germplasm, but that it was also genetically more distinct. Sim-
ilar studies to those in barley have been carried out in wheat
and other cereals. Al Khanjari et al. (2007) surveyed Omani
wheats using SSR markers and Stodart et al. (2007) used Di-
versity Array Technology (DArTr
) markers to examine 705
accessions from the Australian Winter Cereals Collection and
found much diversity, the latter study identifying Nepal as a
unique gene pool of particular value.
An exciting development in the potential exploitation of
landrace germplasm is the considerable interest in using
association-based approaches to identify candidate genes or
regions underpinning complex traits (Gaut and Long, 2003;
Flint-Garcia et al., 2003; Gupta et al., 2005). Large-scale in-
vestigations of sequence variation within genes and across
genomes have only just begun for plant species. Such studies
are required to determine the distribution and extent of link-
age disequilibrium, since this will determine the resolution
power of association-based mapping strategies. From stud-
ies in other plant species it is clear that the natural decay
of linkage disequilibrium with distance occurs at a consider-
ably slower rate in inbreeding systems because eective re-
combination is severely reduced and genetic polymorphisms
remain correlated over longer physical distances (Nordborg
et al., 2002; Morrell et al., 2005).
Recently sequence diversity and patterns of linkage dise-
quilibrium were investigated across a 212 kb region in culti-
vated, landrace and wild barley to determine the impact of in-
breeding and evolution history (domestication and selection)
(Caldwell et al., 2006). High levels of association were found
to stretch across the whole region in the cultivated sample,
with linkage disequilibrium values extended across the en-
tire 212-kb region. In contrast, linkage disequilibrium and its
significance decreased as a function of increasing distance in
both landraces and wild barley (Fig. 1). These contrasting pat-
terns exist despite similar levels of inbreeding and most likely
reflect dierent population histories associated with the oc-
currence of bottlenecks and selection within the domesticated
germplasm. Therefore, large linkage disequilibrium regions in
cultivated, low-resolution whole-genome scans could be de-
ployed to identify candidate gene regions; this would then be
“a8259” — 2009/9/14 — 14:36 — page 10 — #10
i
i
i
i
i
i
i
i
10 A.C. Newton et al.
Cultivated
r2
Median Group Distance
(A) (B) (C)
0 250 500 750 1000 1250 1500 1750
Landrace
r
2
Median Group Distance
0 250 500 750 1000 1250 1500 1750 2000
Wild
r2
Median Group Distance
Figure 1. Plots of linkage disequilibrium as measured with the commonly used statistic (r2) (y axis), which is based on the allele frequencies
at two loci as a function of distance in kilobases (kb) (x axis) for the (A) cultivated, (B) landrace, and (C) wild samples (Caldwell et al., 2006).
complemented by fine-scale, high-resolution linkage disequi-
librium mapping utilising landraces and wild barley to identify
candidate genes.
5.1. Genotyping technology
The developments in association genetics have been facili-
tated through advances in DNA molecular marker technology
reviewed above and by Buckler and Thornsberry (2002). An
example of the latest technology is the high throughput Illu-
mina ‘Golden Gate Assay’ SNP approaches which gives a high
density of markers across many genotypes, enabling associa-
tion genetic approaches to become highly eective. Such sur-
veys provide information on genomic diversity, domestication
and evolution, identify geographic regions, which contain high
levels of diversity, and discriminate between groups of similar
accessions. With the large amount of sequence information in
barley, over 400000 Expressed Sequence tags (ESTs), a plat-
form was set up for high resolution genotyping known as Il-
lumina Oligo Pool Assay (OPA) with 1536 Single Nucleotide
Polymorphisms (SNPs) in each assay. Genes which are tran-
scriptionally responsive to abiotic stress were chosen in par-
ticular for examining landrace populations.
A subset of the 176 landrace accessions from the SJLC re-
ferred to above were subjected to high throughput genotyp-
ing on the OPA platform and 72% of the 1536 SNPs were
polymorphic and well distributed across the 7 barley chro-
mosomes. Using the available phenotypic data also, whole
genome association scans were used to identify and validate
genes and markers linked to performance under drought stress
(J. Russell et al., personal communication). Again five distinct
groups clustering around key ancestors and regions of origin
of the germplasm were identified, the lines from North East
Syria and South Jordan being particularly contrasting. Dier-
ences in the patterns of diversity between regions of origin
were observed along the chromosome, highlighting selection
signatures of adaptation to the environment and/or agronomic
practices prevalent in these regions.
5.2. Genotype-phenotype association
Another example of strategies to combat drought stress us-
ing landrace germplasm is found in the work of Comadran
et al. (2007) who identified barley genomic regions influenc-
ing the response of yield and its components to water deficits
in a collection of 192 genotypes that represented landraces,
old, and contemporary cultivars sampling key regions around
the Mediterranean basin and the rest of Europe. They used a
stratified set of 50 genomic and EST derived molecular mark-
ers, 52 of which were SSRs, and 1131 DArTr
markers which
together revealed an underlying population sub-structure that
corresponded closely to the geographic regions in which the
genotypes were grown. The population was phenotyped for
yield at two contrasting sites in each of seven Mediterranean
sites for two years leading to marker-trait associations to un-
derstand the genetic and physiological dynamics underlying
barley domestication and intensive breeding carried out in
the last century and its relation to adaptation to drought. The
yields observed for individual genotypes ranged from 10 t/ha
to complete failure of individual genotypes to produce any
seed due to the stress. As this was a highly structured sample,
after accounting for this in the analysis, multi-environment
QTLs were detected most frequently on chromosomes 3H,
4H, 5H and 7H (Fig. 2). One of the encouraging findings of
this study was the detection of significant genetic variation for
yield in the eight severely stressed environments where the
mean yield was less than 2 t/ha, with one of the most con-
sistent genomic regions being that on chromosome 7H, where
four out of the five significant associations came from the Jor-
danian sites with mean yield ranging from 0.3 to 1.2 t/ha. The
detection of QTLs in the low yielding environments oers the
“a8259” — 2009/9/14 — 14:36 — page 11 — #11
i
i
i
i
i
i
i
i
Cereal landraces for sustainable agriculture. A review 11
1
Hor2
2
Hor1
3
4
5
6
7
8
Glb1
9
10
11
12
13
14
1H
1
2
3
4
Ppd1
5
6
7
8
9
Vrs1
10
11
12
13
14
15
2H
1
2
3
4
5
6
7
8
9
10
11
sdw1
12
13
14
Glb4
15
16
3H
1
2
3
4
Dhn6
5
RubA
6
7
8
9
mlo
10
11
Vrn2
12
Bmy1
13
4H
1
2
3
4
5
6
ari-e
7
8
9
10
Vrn1
11
12
13
14
15
5H
1
2
3
4
5
6
7
8
Amy1
9
10
11
12
13
14
6H
1
Wx
2
3
4
5
6
7
Amy2
8
9
10
11
12
M4W,I5W I4D M5W I5W,T5D,I4W T5D
J5D J4D,J4W
I4W,M5W
I4D,I5W,M4W ,T5W I4W M4W
J4W,M5W
T4D T5W M4D A4W E5W,M4W,A5W,I5W
T4D
M5D E5W S5W,E5W ,S4W M4D,S5D, J4W T4D
I4D,I5W M4D,E5D
I5D,M5W M4W,S4W,I4W E5D
M4W,M5D S4D,E4W,I4D T4W
S5D M5D,M5W T5D S5D,A5D,T5D
I5W E4D M5D
T5D
M4W,I5W I4D M5W I5W,T5D,I4W T5D
J5D J4D,J4W
I4W,M5W
I4D,I5W,M4W ,T5W I4W M4W
J4W,M5W
T4D T5W M4D A4W E5W,M4W,A5W,I5W
T4D
M5D E5W S5W,E5W ,S4W M4D,S5D, J4W T4D
I4D,I5W M4D,E5D
I5D,M5W M4W,S4W,I4W E5D
M4W,M5D S4D,E4W,I4D T4W
S5D M5D,M5W T5D S5D,A5D,T5D
I5W E4D M5D
T5D
7H
A5W,M4D,M5D,I4W,S4W,E4W
E4W,I4D,I4W,J4D,J5D,S5W
J5D,E5D,E4W,S5W,T5W
J5W,I5D,J4D,J5W,J4W
A5W,M4D,M5D,I4W,S4W,E4W
E4W,I4D,I4W,J4D,J5D,S5W
J5D,E5D,E4W,S5W,T5W
J5W,I5D,J4D,J5W,J4W
Barley bin map showing significant associations with yield
Barley bin map showing significant associations with yield
1
Hor2
2
Hor1
3
4
5
6
7
8
Glb1
9
10
11
12
13
14
1H
1
2
3
4
Ppd1
5
6
7
8
9
Vrs1
10
11
12
13
14
15
2H
1
2
3
4
5
6
7
8
9
10
11
sdw1
12
13
14
Glb4
15
16
3H
1
2
3
4
Dhn6
5
RubA
6
7
8
9
mlo
10
11
Vrn2
12
Bmy1
13
4H
1
2
3
4
5
6
ari-e
7
8
9
10
Vrn1
11
12
13
14
15
5H
1
2
3
4
5
6
7
8
Amy1
9
10
11
12
13
14
6H
1
Wx
2
3
4
5
6
7
Amy2
8
9
10
11
12
M4W,I5W I4D M5W I5W,T5D,I4W T5D
J5D J4D,J4W
I4W,M5W
I4D,I5W,M4W ,T5W I4W M4W
J4W,M5W
T4D T5W M4D A4W E5W,M4W,A5W,I5W
T4D
M5D E5W S5W,E5W ,S4W M4D,S5D, J4W T4D
I4D,I5W M4D,E5D
I5D,M5W M4W,S4W,I4W E5D
M4W,M5D S4D,E4W,I4D T4W
S5D M5D,M5W T5D S5D,A5D,T5D
I5W E4D M5D
T5D
M4W,I5W I4D M5W I5W,T5D,I4W T5D
J5D J4D,J4W
I4W,M5W
I4D,I5W,M4W ,T5W I4W M4W
J4W,M5W
T4D T5W M4D A4W E5W,M4W,A5W,I5W
T4D
M5D E5W S5W,E5W ,S4W M4D,S5D, J4W T4D
I4D,I5W M4D,E5D
I5D,M5W M4W,S4W,I4W E5D
M4W,M5D S4D,E4W,I4D T4W
S5D M5D,M5W T5D S5D,A5D,T5D
I5W E4D M5D
T5D
7H
A5W,M4D,M5D,I4W,S4W,E4W
E4W,I4D,I4W,J4D,J5D,S5W
J5D,E5D,E4W,S5W,T5W
J5W,I5D,J4D,J5W,J4W
A5W,M4D,M5D,I4W,S4W,E4W
E4W,I4D,I4W,J4D,J5D,S5W
J5D,E5D,E4W,S5W,T5W
J5W,I5D,J4D,J5W,J4W
Barley bin map showing significant associations with yield
Barley bin map showing significant associations with yield
Figure 2. Barley chromosome ‘bin’ map showing the location of phenotypes (in red) associated with yield under drought stressed environment
conditions.
prospect of developing Marker Assisted Selection protocols
for yield improvement in such situations.
5.3. Accurate whole-plant field phenotyping
for exploiting variation within landraces
In plant breeding literature, the term “phenotypic selec-
tion” is often used interchangeably with “visual selection”,
even though the two are not synonymous. With the current
rapid advancements in high-throughput molecular genotyp-
ing technologies described above, it is becoming increasingly
clear that the limiting factor in applying those powerful tech-
nologies to molecular breeding programmes is no more the
capacity of genotyping, but the potential for accurate or pre-
cision phenotyping (Campos et al., 2004). Phenotyping be-
comes a particular challenge when moving from qualitative
to quantitative traits, like yield and stability, which have the
greatest interest for breeders (Thomas, 2003). A way to bridge
the genotype-phenotype gap (Miflin, 2000; Parry and Shewry,
2003) is provided when the unit of evaluation and selection
in plant breeding becomes the individual plant (individual
genome) and the confounding eects of competition and soil
heterogeneity on selection eciency are addressed with ap-
propriate experimental designs (Fasoulas and Fasoula, 1995,
2000).
Fasoula (2004) demonstrated a methodology for accurate
whole-plant field phenotyping using the analysis of crop yield
potential, i.e. yield, stability, and responsiveness to inputs
(Fasoula and Fasoula, 2002, 2003) which resulted in extract-
ing superior lines from breeder’s seed of two local barley (cv.
Athenaida) and durum wheat (cv. Kyperounda) cultivars of
landrace origin. An excerpt of this analysis is presented in
Table I .
Developments in genotyping technologies are making the
variation available in landraces ever more accessible. How-
ever, high quality, extensive and detailed, relevant and appro-
priate phenotyping needs to be associated with the genotyping
to enable it to be exploited successfully. We also need to un-
derstand the complexityof the genetics of these desirable traits
in order to develop new germplasm.
6. NUTRIENT UPTAKE AND UTILISATION
Landraces have developed mostly in environmentswith low
nutrients availability, and may therefore represent a source of
variation for selection of varieties adapted to low fertiliser in-
put cropping systems. While the literature on N and P uptake
and utilisation of landraces is relatively rich, little has been
documented for other nutritional elements. Landraces dier
from elite cultivars in their heterogeneous genetic structure as
well as for several typical morpho-physiological traits. How-
ever, the focus below will be on the morphological and phys-
iological aspects associated with uptake and utilisation of ni-
trogen and phosphorus by cereal landraces.
The main factor determining nutrient uptake is the root sys-
tem. This has been shown to be more developed in wheat lan-
draces than in high yielding elite germplasm, especially semi-
dwarf varieties (Siddique et al., 1990; Waines and Ehdaie,
2007). Good soil exploration by roots has been shown to be
“a8259” — 2009/9/14 — 14:36 — page 12 — #12
i
i
i
i
i
i
i
i
12 A.C. Newton et al.
Table I . Ranking of selected lines within the local Cyprus durum wheat cultivar Kyperounda, of landrace origin, based on the analysis of
crop yield potential for accurate whole-plant field phenotyping, using the R-19 honeycomb design, capable to evaluate 19 entries. The original
cultivar Kyperounda (control) was assigned the design code 19 and ranked second last in terms of mean yield per plant.
Yield g/plant Stability of performance Responsiveness to inputs Expected response to selection
Line code x%x/s%(xsel x)/s%x(xsel x)/s2%
7 142.52a100 2.69 97 1.60 80 4.30 78
14 142.12ab 100 2.28 82 1.76 89 4.01 73
5 140.87ab 99 2.58 93 1.72 87 4.44 81
17 138.63ab 97 2.78 100 1.78 90 4.95 90
Other intermediate lines
6 123.26bcd 86 2.44 88 1.77 89 4.32 79
13 123.09bcd 86 2.29 82 1.72 87 3.94 72
3 116.89cde 82 2.15 77 1.70 86 3.66 67
Kyperounda 111.00cde 78 2.03 73 1.67 84 3.39 62
(19-control)
8 101.10e71 2.44 88 1.78 89 4.34 79
essential for absorption of phosphorus(Gahoonia and Nielson,
2004a, 2004b) and nitrogen (Cox et al., 1985; Edwards et al.,
1990; Feil et al., 1990; Laperche et al., 2006; Wieseler and
Horst, 1994), though in the latter case the results are more di-
vergent (Heuberger and Horst, 1995; Kuhlmann et al., 1989;
Van Beem, 1997). Thus, appropriately selected landraces with
well-developedroot systems could be a source of variation for
the improvement of nutrient uptake, but its use would require
suitable methods for the assessment of the root system, which
are still lacking.
Arbuscular mycorrhizas may considerably increase the ac-
tive absorbing surface with minor cost for the plant, com-
pared to the formation of roots and root hairs, enhancing P up-
take and to a certain extent, uptake of other nutritive elements
(Bolan, 1991). The degree of colonialisation has been shown
to depend on the host genotype in wheat (Hetrick et al., 1993;
Kapulnik and Kushnir, 1991; Manske, 1989, 1990; Manske
et al., 1995) and barley (Baon et al., 1993). However, high
colonisation rates are not always associated with correspond-
ingly high symbiosis benefits for the plant as this also depends
on the host genotype (Hetrick et al., 1993; Manske, 1989,
1990; Manske et al., 1995). There is evidence that certain lan-
draces benefit more from symbiosis than high yielding elite
cultivars (Kapulnik and Kushnir, 1991; Manske, 1990). Thus,
improvement of the eciency of symbiosis based on selected
landraces might be possible.
Nitrogen-fixing bacteria in the rhizosphere are also impor-
tant root symbiotic relationships for nutrient uptake, especially
in Azospirillum species. These associations are also influenced
by the host genotype and are particularly developed in sev-
eral wheat landraces and wild ancestors originating from the
southern Mediterranean basin (Kapulnik et al., 1983, 1985,
1987). Although the bacteria contribute to N nutrition of the
host plants, the major benefit for the plant was from stimula-
tion of root growth and thus drought tolerance (Kapulnik et al.,
1983, 1985, 1987).
Landraces and old varieties are often later maturing than
modern cultivars, especially those bred for dryer environments
(Canevara et al., 1994). This may be of importance for the
uptake of N in N-limited environments. If little or no N fer-
tiliser is applied, N supply depends on mineralisation of soil
organic matter, organic fertilisers and crop residuals. The time
course of mineralisation is not always compatible with that of
the crop’s requirements (Panga and Lethaya, 2000) and varies
greatly from year-to-year due to factors such as the weather
and preceding crops. The potential uptake by cereal plants is
mostly higher than the actual mineralisation in the soil, espe-
cially in the later growth stages (Baresel et al., 2008). Late-
maturing genotypes may consequently absorb more N overall,
and may therefore be better adapted to these conditions, if wa-
ter availability is not limiting.
Under N-limited conditions, wheat landraces and varieties
with a taller growth habit and lower harvest index have been
shown to absorb and translocate more nitrogen into the grain
than modern cultivars (Baresel et al., 2005). Figure 3 shows
that old landraces and very old cultivars with similar mor-
phological habit may absorb much more nitrogen under low-
yielding conditions than modern cultivars and breeding lines.
One reason might be greater pre-anthesis uptake and buering
capacity in genotypes with high vegetative biomass (Baresel
et al., 2008), but this aspect has been little investigated. Ni-
trogen absorbed before grain filling is remobilised and then
translocated to the grain after anthesis. Translocation e-
ciency might therefore also contribute considerably to e-
cient N utilisation. However, only small dierences between
genotypes could be detected in the absence of leaf diseases
(Bertholdsson and Stoy, 1995; Johnson et al., 1967; Papakosta
and Garianas, 1991; Pommer, 1990) and therefore genetic
variation of translocation eciency does not appear to oer
much opportunity to improve N eciency.
There are many possible combinations of environmental
factors which determine nutrient availability and landraces
generally reach their genetic equilibrium in environments with
reduced nutrient availability and variable conditions from year
to year. They therefore show many dierent adaptive types
(Attene and Veronesi, 1991) and if maintained through on farm
“a8259” — 2009/9/14 — 14:36 — page 13 — #13
i
i
i
i
i
i
i
i
Cereal landraces for sustainable agriculture. A review 13
260 280 300 320 340 360 380
700 800 900 1000 1100
high−yielding environment
Protein yield (kg/ha)
low−yielding environment
Protein yield (kg/ha)
Pollux
Tassilo
Taca
Dream
Eppweizen
Atar
Diplomat Arus
Jubilar
Kuwerts
Banatian
Naturastar
Old varieties and landraces
Breeding lines
Commercial varieties
Aszita
Wenga
Varieties for organic farming
Figure 3. Protein yields of modern varieties and breeding lines, va-
rieties bred for organic farming and landraces or very old varieties
in an environment with high and low average protein yield, respec-
tively. The latter group have relatively high protein yields in the low-
yielding environments, but low performance in the high-yielding en-
vironments. Results of a linear regression analysis on 70 genotypes
in 9 environments; see Baresel et al. (2005) for details.
conservation, will oer a valuable resource for finding ways of
improving nutrient use eciency without loosing geneticplas-
ticity.
Nutrient use eciency is a very important criterion for
sustainability. Landrace material oers a potential source for
crop improvement although these traits are highly interactive
with their environment, particularly developmental stage, soil
conditions and other organisms aecting roots and their envi-
ronment.
7. NUTRITION AND QUALITY
Early landraces and wild species provide a broad represen-
tation of natural variationnot only in agronomically important
traits, but also in nutraceuticals, which have decreased dur-
ing the breeding of modern cultivars. The nutritional status of
the most important staple foods, such as the cereals is ulti-
mately dependent on their metabolic composition (Galili et al.,
2002). Nevertheless, while traits associated with yield and re-
sistance have been the focus of most research, quality traits
that are dependent on chemical composition are less well stud-
ied. There are some notable exceptions, such as protein content
and structure of wheat cultivars and landraces (Láng, 2006;
Rakszegi et al., 2006). Mineral content in modern wheat cul-
tivars has significantly decreased including copper, iron, mag-
nesiou, manganese, phosphorous, selenium, and zinc (Murphy
and Jones, 2006). Looking for breeding sources for higher lev-
els of iron and zinc has revealed that the highest levels can be
found in landraces and low yielding genotypes (Monasterio
and Graham, 2000). An ancestral wild wheat gene was found
that accelerates senescence and increases nutrient remobilisa-
tion from leaves to developing grains. The gene was found in
all wild emmer accessions and most domesticated emmer but
not in durum lines or hexaploid wheat (Uauy et al., 2006). Ce-
real seed mineral content have been analysed in a wide range
of cereal genebank accessions (Bálint et al., 2003)to find opti-
mal sources to increase mineral content of modern wheat cul-
tivars. Emmer wheat is a promising source of genetic variation
for protein, zinc and iron content.
In bread wheat, the concentration of carotenoids is low, but
they are more abundant in, for example, durum wheat, em-
mer and einkorn cultivars and landraces, having higher con-
centration in landraces then in cultivars (Panfili et al., 2004).
Tocols, in contrast, are abundant both in bread wheat and du-
rum wheat (Panfili et al., 2004). As carotenoid and tocol con-
tents are independent traits (Hidalgo et al., 2006) that were not
subject to conscious human selection and are selectively neu-
tral, the wide range of natural variation still exists in the cereal
genetic resources and landraces. In recent experiments bread
wheat landraces, emmer and einkorn were found to be the best
sources of tocols, while durum landraces and emmer were the
best sources of carotenoids (Hidalgo et al., 2006).
Total phenolics in wheat varied both with cultivar and farm-
ing site (Gélinas and McKinnon, 2006). The phenolic com-
pounds flavonoids, saponins, lignans and sterols are found in
oat grain, but in minor quantities. Their concentrations are
very low compared to avenantramides and tocols but have an-
tioxidant and other bioactivitie properties (Peterson, 2004).
Tocopherols and tocotrienols have higher concentrations in
cultivars with high total lipids (Bryngelsson et al., 2002) and
landraces of black oats have higher concentrations of antioxi-
dants than elite cultivars (Mannerstedt-Fogelfors,2001).
Maize contains appreciable amounts of carotenoids
(Wurtzel, 2004) and has a greater total phenolic content and
total antioxidant activity than wheat, oats or rice (Adom and
Liu, 2002). In the maize kernel, tocopherol and oil content
may be physiologically associated (Kurilich and Juvik, 1999)
while the natural variation detected in maize landraces ker-
nel pigmentation may be associated with increased availabil-
ity of certain antioxidant compounds. White maize polyphe-
nolics have shown to have antioxidant and anti-carcinogenic
eects (Del Pozo-Insfran et al., 2006). Blue, purple and
red-pigmented maize kernels are also rich in anthocyanins
with well-established antioxidant and bioactive properties
(Del Pozo-Insfran et al., 2006). Purple maize, that has been
cultivated for centuries in the Andean Region, is a good exam-
ple of this (Pedrechi and Cisneros-Zevallos, 2006).
In general, cereal landraces and old varieties are among
the best sources of phytonutrients accompanied with opti-
mal micronutrient concentrations. Grains, fruits and vegeta-
bles contain a broad variety of phytonutrients, which show a
significant eect on reducing the incidence of aging-related
and chronic diseases. Among the numerous antioxidant com-
pounds present in these foods, grain fat-soluble antioxidants
and their unique bioactive compounds play an important role
in disease prevention. The additive and synergistic eects of
“a8259” — 2009/9/14 — 14:36 — page 14 — #14
i
i
i
i
i
i
i
i
14 A.C. Newton et al.
such phytochemicals in fruits, vegetables and whole grains are
thought to be primarily responsible for their health benefits.
Landraces are clearly a potential source of traits for im-
proved nutrition of cereal crops, particularly antioxidants,
phenolics in general, carotenoids and tocol in particular. They
also have the potential to improve mineral content, particularly
iron and zinc, if these traits can be successfully transferred to
improved varieties.
8. BIOTIC AND ABIOTIC STRESS RESISTANCE
AND TOLERANCE
The replacement of landraces by homogeneous cultivars
entails a significant loss of genetic variation for resistance
to biotic and abiotic stresses. Elite cultivars may not possess
the combined resistances already present in the landrace that
they are intended to replace. Also, landraces might be a good
reservoir of resistance mechanisms other than the hypersen-
sitive mechanism typically exploited in modern cultivars due
to its simple inheritance and complete expression. Most of
the studies cited and others are either screening for qualita-
tive resistance or more detailed mapping of the major genes
responsible. There are few studies of landraces for partial and
polygenic resistance due to the increased resources needed to
obtain quantitative data. However, such studies are needed to
determine whether multiple sources of partial resistance from
such sources oer genes more likely to be durable than those
in the elite gene pools. Below we will review the biotic stresses
for just wheat and barley where landraces have provided valu-
able sources of resistance, then disease tolerance and some
abiotic stresses for cereals in general.
8.1. Wheat diseases
8.1.1. Septoria leaf blotch
The fungus Mycosphaerella graminicola (sexual stage of
Septoria tritici) or Septoria Tritici Blotch (STB) is currently a
major disease of worldwide distribution. Most currently grown
wheat cultivars are more or less susceptible to M. graminicola.
The Italian landrace Rieti, an ancestor of many modern Euro-
pean wheat cultivars has been identified as very resistant to all
studied isolates (Arraiano and Brown, 2006). The presence of
Stb6 gene in both European and Chinese landraces suggests
that this gene has been present in cultivated wheat since the
earliest times of agriculture (Chartrain et al., 2005). Resistance
has also been found in Czech and Slovak landraces (Vechet
and Vojácková, 2005).
8.1.2. Powdery mildew
Powdery mildew of wheat is a foliar disease caused by the
obligate biotrophic fungus Blumeria graminis f. sp. tritici (syn.
Erysiphe graminis f. sp. tritici) that can cause loss in both
grain yield and quality. Screening of old wheat cultivars, lan-
draces and related species for resistance to powdery mildew
started in the 1930’s (Hsam and Zeller, 2002) and Pm genes
have since been identified in many dierent, widely distributed
wheat cultivars and landraces. Non-major gene resistance has
been sought such as the durable adult plant resistance found in
the landrace line k-15560 (Peusha et al., 2002).
8.1.3. Fusarium Head Blight
There is little resistance available to the major disease
Fusarium Head Blight caused by Fusarium graminearum al-
though some has recently been identified in landraces (Zhang
et al., 2000).
8.1.4. Bunts and smuts
Common bunts (Tilletia foetida and T. caries) are impor-
tant diseases that are easily controlled by seed dressing, but
might become more important in organic and low input agri-
culture. More than 15 resistance genes (Bt1-15) have been
identified in wheat and the landrace PI178383, originally col-
lected in Turkey, carries resistance genes Bt-8, Bt-9 and Bt-
10 plus an unidentified factor (Goates, 1996). Resistance has
also been reported in landraces of bread wheat (Hubert and
Buertsmayr, 2006) and of durum wheat (Mamluk and Nachit,
1994). Karrnal bunt (Tilletia indica) is the most recently de-
scribed smut of wheat and resistance has been identified in
Indian landraces (Anon., 1943).
8.1.5. Rust diseases
Stem rust (Puccinia graminis) resistance transferred to
bread wheat from Yaroslav emmer (Sr2 complex) in combina-
tion with other genes seems to have provided the foundation
for durable resistance to stem rust in CIMMYT germplasm
(Roelfs, 1988) in the last 50 years. Sources of durable resis-
tance to stem rust in durum wheat have been reported, like
the durum wheat Glossy Huguenot, eective in Australia over
the past 100 years (Hare, 1997). Interestingly, its resistance is
also quantitative and based on a reduced number of pustules
in adult plants and a delayed onset of disease. As indicated
by Roelfs (1988), of the 41 known genes for stem rust re-
sistance, 20 originated in species other than T. aestivum and
T. turgidum; of the 35 known genes for leaf rust resistance,
12 originated in species other than T. aestivum and T. turgidum.
Among the genes originating from T. aestivum for resistance
to either rust, a number of these are from landraces (McIntosh
et al., 1998). The recently reported spread of the stem rust
race Ug99 and the dependence of so many elite cultivars of
wheat on the Sr31 resistance gene has spurred renewed interest
in surveying landrace collections for novel resistance sources
(Bonman et al., 2007).
Durable resistance to leaf rust (Puccinia recondita f.sp. trit-
ici and P. triticina) of wheat is thought to be more dicult to
“a8259” — 2009/9/14 — 14:36 — page 15 — #15
i
i
i
i
i
i
i
i
Cereal landraces for sustainable agriculture. A review 15
obtain than to stem rust but resistance against leaf rust has been
identified that appears more durable than the norm. Resistance
in the bread wheat cultivars Americano 44D, and Frontana,
derived from resistant landraces, appears to be particularly
durable. It appears that Lr12 and Lr13, both genes for hyper-
sensitive resistance that is expressed only in the adult plant
stage, in combination with Lr34, are the basis of most of this
resistance (Roelfs, 1988; Rubiales and Niks, 2000). Ameri-
cano 44D, a Uruguayan landrace of unknown origin (called
Universal 2 in Argentina), was used by Klein in breeding
early Argentinean lines and is now considered another impor-
tant source of durable resistance to leaf rust (Van Ginkel and
Rajaram, 1992).
Durable resistance to yellow rust (Puccinia striiformis)has
been described in wheat landraces from China, Italy and the
Netherlands (Van Dijk et al., 1988; Zhang, 1995). Partial re-
sistance combined with temperature-sensitive resistance have
been suggested as the major components of the durable resis-
tance found in old winter wheats in the Netherlands (Van Dijk
et al., 1988). Sources of partial resistance to leaf rust have been
recently reported in landraces from various origins (Fekadu
and Parlevliet, 1997; Martínez et al., 2001a, 2001b; Shtaya
et al., 2006a, 2006b).
Histological studies on interactions between plants and
rusts can help both to discern the various resistance mecha-
nisms and to combine them in a genotype in the hope to in-
crease durability. The commonly used hypersensitivity resis-
tance, typically conferred by single genes with race-specific
eectiveness, is due to a post-haustorial defence mechanism.
This type of resistance is very common in non-host interac-
tions (Niks and Rubiales, 2002). It also is the mechanism re-
sponsible for the partial resistance of some wheat landraces to
wheat leaf rust (Martínez et al., 2001c).
8.1.6. Aphids
Resistance to the Russian wheat aphid (Diuraphis noxia)
has been found in landraces of wheat from Iran and the former
Soviet Union (Du Toit, 1987).
8.2. Barley diseases
8.2.1. Powdery mildew
In most barley growing regions powdery mildew (Blumeria
graminis, (syn.: Erysiphe graminis)(D.C.Speer)f.sp.hordei)
is very common. Intensive studies were carried out on barley
landraces from Ethiopia (Negassa, 1985a, 1985b), Jordan and
Syria (Van Leur, 1989) and other countries of the Near East
(Weltzien, 1988), Europe (Honecker, 1938), India (Freisleben,
1940), Japan (Hiura, 1960) and world-wide (Moseman, 1955;
Nover and Mansfeld, 1955, 1956; Homann and Nover, 1959;
Rigina, 1966; Wiberg 1974a, 1974b; Moseman and Smith,
1976; Czembor, 2002). The main history of incorporation of
powdery mildew resistance genes in cultivated barley and the
exploration of their genetic diversity in Europe is described by
Wolfe and Schwarzbach (1978).
The situation regarding diversity of genes for partial re-
sistance to powdery mildew in cereal landraces is less clear.
However, many assessments of partial resistance to pathogens
in barley have been made with powdery mildew (Wright and
Heale, 1984; Asher and Thomas, 1983, 1984, 1987; Anderson
and Torp, 1986; Carver, 1986; Heun, 1986; Geiger and Heun,
1989; Newton, 1990; Kmecl et al., 1995).
8.2.2. Rust diseases
Three rust species commonly occur on barley: leaf rust
(Puccinia hordei Otth), stripe rust (P. striiformis West. f. sp.
hordei) and stem rust (P. graminis Pers.: Pers. f. sp. tritici
Eriks. et Henn.).
The origins of resistance sources were similar to those
of mildew-resistant landraces from the Mediterranean region
where both the host and the pathogen are indigenous and
have co-evolved (Anikster and Wahl, 1979). Israel in par-
ticular is part of the centre of origin and genetic variation
of wild native Hordeum species H. vulgare ssp. spontaneum
Koch, H. bulbosum L. and H. murinum (Wahl et al., 1988;
Kandawa-Schulz, 1996). Some interesting material has also
been found in Azerbaijan and Turkmenia (Bakhteev collec-
tion) and Iran (Kuckuck collection) (Nover and Lehmann,
1974; Walther and Lehmann, 1980). Many studies described
the activities in evaluationand in breeding for major genes and
partial resistance (Cliord, 1985; Reinhold and Sharp, 1986;
Yahyaou et al., 1988; Khokhlova et al., 1989; Jin et al., 1995;
Lukyanova and Terentyeva, 1997; Alemayehu and Parlevliet,
1997). Resistance against a new pathotype of Puccinia hordei
with virulence for the resistance gene Rph7 has been identified
in barley landraces (Shtaya et al., 2006c).
Stripe rust is known in most of the barley growing re-
gions. Screening for new sources of resistance has been car-
ried out by many groups (Nover and Lehmann, 1966, 1970,
1975; Upadhyay and Prakash, 1977; Stubbs, 1985; Van Leur
et al., 1989; Okunowski, 1990; Luthra et al., 1992; Hill et al.,
1995). As with wheat rusts, much durable resistance is pre-
haustorial and is known to be the mechanism responsible for
the partial resistance of some barley landraces to barley leaf
rust (Shtaya et al., 2006a, 2006b). Sources of partial resistance
to leaf rust (P. hordei) have been recently reported in barley
and wheat landraces from Spain and from Fertile Crescent
(Martínez et al., 2001a, 2001b; Shtaya et al., 2006a, 2006b).
8.2.3. Scald
Evaluations of barley for resistance to scald (Rhynchospo-
rium secalis (Oudem.) J. J. Davis f. sp. hordei) have been car-
ried out in many countries (Fukuyama et al., 1998; Yitbarek
et al., 1998). Recent examples of such reports are: novelalleles
at the Rrs1 and other loci have been foundfor Rhynchosporium
secalis resistance (Grønnerød et al., 2002; Bjørnstad et al.,
2004).
“a8259” — 2009/9/14 — 14:36 — page 16 — #16
i
i
i
i
i
i
i
i
16 A.C. Newton et al.
8.2.4. Net blotch
Studies of landraces for resistance to net blotch
(Pyrenophora teres (Died.) Drechsl. f. teres) have been
carried out by many scientists (Schaller and Wiebe, 1952;
Buchannon and McDonald, 1965; Gaike, 1970; Smirnova
and Trofimovskaya, 1985; Proeseler et al., 1989; Lukyanova,
1990; Faiad et al., 1996). Sato and Takeda (1994) studied the
variation of host resistance of 2233 accessions of the barley
world collection and found sources of resistance in accessions
from Ethiopia, North Africa and Korea. New sources with
resistance to up to eight races of P. t e r e s were found among
Peruvian landrace accessions (Afanasenko et al., 2000).
8.2.5. Barley stripe
Many studies were conducted to identify new sources
of resistance to barley stripe (Pyrenophora graminea Ito &
Kuribayashi) including landraces (Baigulova and Pitonya,
1979; Nettevich and Vlasenko, 1985; Skou and Haahr, 1985;
Van Leur et al., 1989; Su et al., 1989; Lukyanova, 1990; Bisht
and Mithal, 1991; Ceccarelli et al., 1976; Kirdoglo, 1990;
Skou et al., 1992, 1994).
8.2.6. Common root rot and spot blotch
Several germplasm collections have been evaluated and re-
sistance to common root rot and spot blotch (Cochliobolus
sativus (Ito & Kurib.) Drechsler ex Dastur) has been identi-
fied from several resources (Banttari et al., 1975; Velibekova,
1981; Rochev and Levitin, 1986; Lehmann et al., 1988;
Lukyanova, 1990; Gilchrist et al., 1995; Semeane, 1995; Fa-
iad et al., 1996).
8.2.7. The smuts
Three species of cereal smut attack barley: Ustilago nuda
(Jens.) Rostr. (U. segetum var. nuda), U. nigra Tap k e (U. sege-
tum var. avenae)andU. hordei (Pers.) Lagerh. (U. segetum).
Many studies on resistance to these pathogens were con-
ducted (Shchelko, 1969, Nover et al., 1976; Damania and
Porceddu, 1981; Onishkova, 1987; Dunaevskij et al., 1989;
Surin, 1989; Lukyanova, 1990; Dubey and Mishra, 1992) iden-
tifying sources of resistance from landraces from Ethiopia,
Yemen, Tibet, Canada and USA.
8.2.8. Fusarium complexes
Many scientists described dierences in fusarium resistance
between cultivars (Grigor’ev et al., 1988; Van Leur, 1989; Gu,
1989; Corazza et al., 1990; Khatskevitch and Benken, 1990;
Lukyanova,1990; Takeda, 1992; Filippova et al., 1993; Nelson
and Burgess, 1994; Perkowski et al., 1995, 1997). Based on
these reports it can be concluded that valuable sources of re-
sistance were identified in the East Asian region landrace ac-
cessions in particular (Takeda and Heta, 1989).
8.2.9. Viruses
Yasuda and Rikiishi (1997) evaluated a total of 4342 bar-
ley accessions from the world collection on a field in Japan
infected with strain I (Kashiwazaki et al., 1989) of BaYMV
for resistance. The percentage of asymptomatic cultivars was
highest among Ethiopian landraces followed by those from
Japan. Cultivars showing severe disease symptoms were fre-
quently found among Chinese, Nepalese, southeast Asian,
north African, north American and European accessions. Field
resistance to barley yellow dwarf was detected in several
Ethiopian barleys (Schaller et al., 1964).
8.2.10. Aphids
The spring two-rowed barley RWA 1758 has been devel-
oped via selection from CIho 4165, a landrace originally col-
lected in Afghanistan (Bregitzer et al., 2008).
8.3. Disease tolerance
A character much neglected in elite breeding programmes
is disease tolerance, not least because of the varied defini-
tions of the term and the diculty of measuring it (Bingham
and Newton, 2009). However, inter-specific variation has been
found in cereals, for example powdery mildew-infectedleaves
of a wild oat showed a smaller reduction in net photosynthetic
rate than a cultivated oat genotype under comparable infec-
tion severities (Sabri et al., 1997). The wild oat leaves also
showed a slower rate of disease-associated senescence. In a
comparison of wild and cultivated barley genotypes similar ef-
fects were reported (Akhkha et al., 2003) but the eect of this
variation on tolerance at the scale of the crop canopy has not
been determined. Several landrace accessions appear amongst
the accessions screened for tolerance in barley (Newton and
Thomas, 1994; Newton et al., 1998, 2000), though not dispro-
portionately with more modern cultivars.
8.4. Abiotic stresses
Generally the genus Hordeum shows a high degree of adap-
tation to dierent stressful environments.
8.4.1. Drought tolerance
Drought is the most common abiotic constraint for stable
barley production in rain-fed areas. Under Mediterranean con-
ditions, water stress is particularly common at the end of bar-
ley life cycle (Passiuora, 1996). In comparisonto other cereals,
barley is well adapted to arid environments and the immedi-
ate progenitor of cultivated barley H. vulgare ssp. spontaneum
can grow in desert condition (Nevo, 1992; Zohary and Hopf,
1998). Such ecotypes were identified in desert locations in Jor-
dan (Jaradat et al., 1996). The study of drought stress on yield
in Mediterranean environments noted above (Comadran et al.,
2007) identified genomic regions in landraces that may be very
valuable for combating such stress.
“a8259” — 2009/9/14 — 14:36 — page 17 — #17
i
i
i
i
i
i
i
i
Cereal landraces for sustainable agriculture. A review 17
8.4.2. Frost tolerance
Selection of highly frost tolerant lines from Turkish barley
landraces has resulted in conversion of spring based produc-
tion into winter based and enlargement of winter sown barley
production in the Turkish highlands since 1940. Today 60%
of barley production has been provided by winter sown bar-
ley, yield has doubled and these landraces have been routinely
and widely used as parents in many breeding programmes in
Turkey (Akar et al., 1999).
8.4.3. Salinity tolerance
In the investigation of Abo-Elenin et al. (1981) 1163 en-
tries were tested in the field and 777 in lysimeters. In this study
Abyssinia’ was the most tolerant. Mano et al. (1996) screened
6712 accessions for salt tolerance at germination. Accession
variation showed a normal distribution and the most tolerant
ones could germinate in sea water. However, six-rowed culti-
vars were more tolerant than two-rowed, hull-less than hulled,
normal than semi-dwarf ‘uzu’, and winter than spring. In an-
other study Mano and Takeda (1995) evaluated 5182 barley
cultivars for salt tolerance at seedling stage. Generally the
geographical dierentiation among tested accessions was not
clear.
Tolerance to salinity is more frequent in bread than is du-
rum wheat, as bread wheat has a salt-exclusion mechanism but
durum does not. This was found in landraces originating from
saline areas of the Middle East and is conferred by Nax genes
which therefore could be incorporated into both durum and
bread wheat (Munns, 2005).
8.4.4. Acid and alkaline soils and tolerance to heavy
metal toxicity
Cereal adaptation to acid and alkaline soils is limited by
two major problems, the aluminium and manganese toxic-
ity in acid environments and boron toxicity in alkaline ones.
The soil acidity is a serious agricultural problem, aecting as
much as 40% of the world’s arable land and up to 70% of
the world’s potentially arable land (Kochian et al., 2005; Hede
et al., 2001). Aluminium toxicity is a main growth and yield-
limiting factor on soils with pHs below 5.0 (Davies, 1994),
and can directly reduce yield by up to 60% (Tang et al., 2003).
Amelioration of soil surface layer is not a reasonable solution
in low input and organic agriculture, and because plant roots
develop in lower acid layers to reach critical water and nu-
trient supplies. Selection and development of genotypes with
enhanced tolerance to acid soils and toxic levels of aluminium
is considered to be a more eective solution to this problem.
Cereal crops show very dierent responses to aluminium
toxicity and soil acidity. The highest aluminium tolerance is
detected among rye (Little, 1988), followed by oat (Slaski,
1992), wheat (Aniol and Madej, 1996), barley (Foy et al.,
1965), and corn (Horst et al., 1997). Some experimental ev-
idence shows a dramatic variation in aluminium tolerance
among cultivars, which can be related to their genetic vari-
ability (Carver and Ownby, 1995). Landraces are important
sources of this variability to improve the aluminium and
acid soils tolerance in breeding programmes, but most re-
search on aluminium tolerance has been carried out on elite
cultivars or isogenic lines (Kochian et al., 2005) and there
are very few studies to identify landrace tolerance for these
traits (Pinheiro de Carvalho et al., 2003; Gudu et al., 2001;
De Sousa, 1998; Cosic et al., 1994). Forty-eight accessions
representing 16 Madeiran wheat landraces were screened for
their aluminium tolerance using erichrome staining and root
elongation (Pinheiro de Carvalho et al., 2003, 2004) and the
accumulation of callose in the root types (dos Santos et al.,
2005). The variability of landrace responses to the presence
of aluminium and the existence of high performing accessions
with better performance has been shown by comparison with
elite cultivars such as Maringa. Durum wheat landraces show
less variability and are moderately sensitive or tolerant to the
presence of aluminium (Pinheiro de Carvalho et al., 2003;
Cosic et al., 1994). De Sousa (1998) published a classifica-
tion of aluminium tolerance of 76 wheat cultivars, including
several landraces introduced in earlier twentieth century. The
major sources of aluminium tolerance in wheat are considered
to be originated from Brazil (Zhou et al., 2007). However, Sto-
dart et al. (2007) and Zhou et al. (2007), through the screening
of wheat accessions from dierent countries showed the exis-
tence of potential new sources of aluminium resistance among
the landraces germplasm originated from Bulgaria, Croatia,
India, Italy, Nepal, Spain, Tunisia, and Turkey. The sources
of aluminium tolerance in barley are limited to old cultivars
and landraces, and represent multiples alleles of a single locus
(Nawrot et al., 2001). The evaluation of corn germplasm, in
two dierent studies screening of 76 accessions of unknown
number of Kenyan maize landraces (Gudu et al., 2001) and
40 accessions of five Madeiran maize landraces (Pinheiro de
Carvalho et al., 2004) also showed their high variability in
aluminium tolerance, with several accessions presenting bet-
ter performance then commercial standards.
The soil alkalinity also aects agricultural crops, growing
in arable soils with pHs between 8 and 10. In barley, boron
toxicity is directly responsible for yield penalties of up to 17%
(Cartwright et al., 1984). The screening of 444 accessions of
winter barley and 19 accessions of durum wheat, including
landraces from Europe, west Asia and north Africa showed
that boron tolerance is associated with geographic origin (Yau,
2002; Yau et al., 1995).
Landraces have long been assumed to be valuable sources
of resistance to pathogens and the literature demonstrates that
there is much to be gained from such sources. Transfer of re-
sistance genes from landraces to modern cultivars is likely to
be less problematic than from wild accessions.There is clearly
also potential, largely unrealised, for disease tolerance and re-
sistance or tolerance of pest and various abiotic stresses too
including to toxic environments.
“a8259” — 2009/9/14 — 14:36 — page 18 — #18
i
i
i
i
i
i
i
i
18 A.C. Newton et al.
9. BREEDING: CONVERSION OF LANDRACES
INTO MODERN CULTIVARS
In this section the advantages of breeding cereal landraces,
the methodology of improvement and modifications of breed-
ing assumptions will be proposed.
9.1. Yield-based selection
Breeding from landrace accessions is a strategy being used
to improve yield and yield stability in less favourable agricul-
tural system with lower input levels. Heritabilities are higher
in more favourable than poor environments (Blum, 1988). The
stagnation of yields in these areas (Annicchiarico and Pecetti,
1993) is mainly related to the narrow genetic base of the
more recently bred, high-yielding cereals (Pecetti et al., 2002).
Two d i erent approaches may be followed to raise yields in
the long-term: one is based on increasing yield potential of
broadly adapted cultivars, while the other relies on the bet-
ter exploitation of the adaptive features of genotypes by fit-
ting cultivars to specific target environments (Acevedo and
Fereres, 1993). Improvementof grain yield potential in small-
grain cereals has traditionally relied on direct selection for this
trait (Annicchiarico and Pecetti, 1998). Traditional breeding is
based on a combination of bulk-pedigree method of selection,
applied selection in the presence of stress, and use of adapted
germplasm (Ceccarelli and Grando, 1997).
The opportunity to complement traditional breeding with
use of indices of indirect selection for yield including sets
of morpho-physiological traits, also known as an analytical
breeding, has been put forward, especially for less favourable
regions (Richards, 1982). There are two strategies in analytical
breeding for identifying morpho-physiologicaltraits usable as
tools for selection (Fischer, 1981; Jackson et al., 1996). The
first, called the ‘black box’ strategy, consists of assessing a
germplasm pool for correlated response to yield gain deriving
from selection for sets of putatively useful traits. The second,
defined as the ‘ideotype strategy’, is based on the assessment
of traits chosen a priori, through comparison in isolines or pre-
diction of performance in crop growth models of dierent trait
levels (Annicchiarico and Pecetti, 1998).
9.2. Adaptability
In landraces an understanding of the relationship between
amount of genetic diversity expression of morphological
and agronomic characters and adaptation to stress environ-
ments may elucidate whether the success of landraces in less
favourableareas is due to a population buering mechanism or
to a particular architecture of morpho-physiological traits, or
both. This may in turn clarify whether ‘pure line breeding’ is
the correct approach for less favourable areas (Ceccarelli et al.,
1987). Pure line breeding can be successful only if genotypes
with a very high degree of phenotypic plasticity are identified.
Evans (1980) pointed out that selection for adaptation may
result in yield increases but may not represent selection for
greater yield potential. However, it has already been shown
for durum wheat (Pecetti et al., 1992) and barley that some
of the material selected under unfavourable conditions is able
to retain its superiority in a more favourable environment
(Ceccarelli et al., 1991). For barley the proportion was about
20% of the selected genotypes, and for durum wheat about
30%. In both cases such a proportion was higher than the pro-
portion of lines selected under favourable conditions which
were also able to perform well in a less favourable environ-
ments, and this is in agreement with previous observations
(Pecetti et al., 1994).
Breeding for specific adaptation is particularly important
in the case of crops predominantly grown in unfavourable
conditions, because unfavourable environments tend to be
more dierent from each other than favourable environments
(Ceccarelli and Grando, 1997). The specific adaptation strat-
egy may be explored on the basis of yield response of the
germplasm pool that is representative of the available genetic
base tested across a representative sample of sites within the
target region (Annichiarico, 2002).
9.3. Conversion into density-neutral modern cultivars
Inter-plant competition, i.e., the unequal sharing of growth
resources due to genetic (pre-existing) or acquired dierences
among plants, can be quantified by the drastic increase in
the coecient of variation (CV) of individual plant yields
in the crop stand (Fasoula and Fasoula, 1997). The dier-
ence between the genetically heterogeneous landraces and
the genetically highly homogeneous modern cultivars means
that a landrace stand involves genetic competition among
plants, whereas a modern cultivar stand is devoid of genetic
competition. A systematic study of the relationship between
yield and competitive ability within a bread wheat cultivar
(Fasoula, 1990) found a high and significant negative corre-
lation (r =0.94). The study demonstrated that highly compet-
itive plants, i.e., those yielding less at the ultra-low planting
density (1 plant/m2; absence of competition), out-yielded the
low competitors in mixed stands, i.e., plants yielding more at
the ultra-low density. Conversely, in pure stands, the perfor-
mance of highly competitive genotype plants lagged behind
that of the poor competitors.
Genetic heterogeneity, such as that found in landraces, in-
volves genetic competition among plants, in addition to ac-
quired competition, which is also encountered in stands of
single genotypic cultivars. In principle, genetic competition
can be eliminated when landraces are converted into desir-
able homozygous lines, but the acquired competition is more
dicult to control. The intensity of acquired competition in-
creases in marginal or low input environments, which possess
inherent heterogeneity in the distribution of resources. The
above oers an insight into the reasons that led to the even-
tual replacement of landraces by pure single genotype mod-
ern cultivars in favourable environments and their persistence
in marginal environments. In favourable environments, com-
petition in the crop stands is reduced because of elimination
of both the genetic (single genotype) and the acquired (ample
“a8259” — 2009/9/14 — 14:36 — page 19 — #19
i
i
i
i
i
i
i
i
Cereal landraces for sustainable agriculture. A review 19
resources) components. This results in increased crop yields,
reflected in the reduced CV of individual plant yields (Fasoula
and Fasoula, 1997; Tollenaar and Wu, 1999).
Yields of modern single genotype pure cultivars can be
either density-dependent as in the case of maize hybrids
that yield optimally under high plant densities only (Duvick,
1992), or density-neutral, i.e., remain optimal under a wide
range of plant densities. With appropriate breeding methodol-
ogy, it is eventually possible to convert the density-dependent
into more density-neutral cultivars (Fasoula and Fasoula,
2000; Tokatlidis et al., 2001). To eectively exploit landrace
diversity for breeding purposes, it is important to appreciate
the significance of creating more density-neutral genotypes,
particularly for marginal environments, which usually suer
from drought stress. In drought-prone environments, the use
of more density-neutral cultivars allows use of lower seeding
rates, limiting the damage due to drought. A subtle point is the
understanding that density itself is not a stress; it simply en-
hances the existing dierences that lead to competition in the
stand. Therefore, particularly in more uniform environments,
it is possible to have a high yielding dense stand with reduced
inter-plant competition, as measured by the reduced CV of in-
dividual plant yields.
There is a lot of evidence that mixtures can be valuable,
especially when combining various disease resistances of the
components. A word of caution is, however, presented as
to the interpretation behind the phenomenon. In favourable
environments, the evidence of yield stability due to hetero-
geneity is often counter-balanced by experiments and the-
ory, indicating that increasingly homogeneous cultivars have
higher yield potential across sites and years. A case in point
is the well-known, superior stability of performance of the
genetically homogeneous modern single-cross maize hybrids
over the older genetically heterogeneous double-cross hybrids
(Duvick, 1992), because the adverse eects of genetic com-
petition are restricted by genetic uniformity. This hints to the
non-universality of the superiority of populational (Allard and
Bradshaw, 1964) buering. However, the requirement for su-
perior individual buering is that genes conferring tolerance
to the biotic and abiotic stresses are being gradually incorpo-
rated into a few or a single individual(s). Further, the cause
of the superior stability of certain mixtures has been indicated
(Fasoula and Fasoula, 1997) to mainly rest in the reduced sta-
bility of the individual components (quantified by the higher
CV of single plant yields in pure stands), appearing as reduced
under-compensation in the mixtures, mimicking the eects of
true over-compensation (pseudo-overcompensation).
9.4. Seed degradation – cultivar degeneration
An additional detrimental eect of the negative correlation
between yield and competitive ability relates to the observed
landrace or cultivar “seed degradation”. Zeven (1999) provides
an interesting array of evidence about what he calls the often
inexplicable seed replacement by traditional farmers. A tradi-
tional practice to combat seed degradation has been the peri-
odic seed replacement of farmers’ own seed with seed from
elsewhere. Zeven (2000) reports a widely existing belief that
the home-grown cultivar degenerates after several generations
of re-sowing. He further states (Zeven, 2000) that most farm-
ers do not actually perform traditional maintenance breeding;
as they and their ancestors probably have experienced that tra-
ditional maintenance breeding does not result in a better crop.
Apparently, “farmers must have thought that seed replacement
was a better method to maintain the yielding capacity of their
crops”.
A proposed explanation for the practice of “inexplicable”
seed replacement and the avoidance of traditional mainte-
nance breeding by farmers relates to the consequences of
the existing, but mostly unsuspected, negative correlation be-
tween yield and competitive ability. The problem of cultivar
or landrace degeneration (Fasoula, 1990) can be addressed
by applying the concept of non-stop selection (Fasoula and
Fasoula, 2000) for superior lines at ultra-low plant densi-
ties. The outcomes of non-stop selection exceed that of con-
ventional maintenance breeding (Fasoula and Boerma, 2007;
Tokatlides, 2005). Experimental data of honeycomb selection
within breeder’s seed of the old barley cultivar Athenais and
the old durum wheat cultivar Kyperounda, both of local lan-
drace origin, demonstrated the existence of useful adaptive
variation persisting within homozygote lines, as well as the
potentialities of non-stop selection (Fasoula, 2004).
Adaptive variation is genome-monitored, de novo and her-
itable across generations. It is directional and constantly re-
leased by the sensory mechanisms of the genome in response
to environmental stimuli (McClintock, 1984; Rasmusson and
Philips, 1997). This epigenetic variation stems from the in-
teraction between genotype and environment (Goldberg et al.,
2007) and allows profitable exploitation of limited resources
and continual incorporation of gene variants for resistance to
changing biotic and abiotic stresses. Continuous exploitation
of adaptive variation is synonymous with the continuous ge-
netic upgrading of landraces and cultivars.
Because of the reported negative correlation between yield
and competitive ability, integration of yield and stability genes
into fewer, improved genotypes is more ecient when the unit
of evaluationand selection becomes the individual plant grown
at ultra-low planting densities. When this negative correlation
is considered at the level of the individual plant, it means that
a plant possessing genes for high yield potential will also pos-
sess genes for low competitive ability. At the level of the vari-
ety crop stand, the negative correlation means that the greater
the inter-plant competition in the stand, quantified by the yield
CV of individual plants, the greater the crop yield reduction.
At the level of selection and landrace/variety maintenance, it
means that high competitors are selected at the expense of
higher yielding genotypes. As a result, the variety eventually
degenerates. This phenomenon oers a novel explanation for
the previously reported practice of landrace seed replacement
amongst traditional farmers and their avoidance of traditional
maintenance breeding.
In summary, single gene traits are generally easily trans-
ferred from landrace germplasm to modern cultivars, but most
of the desirable traits characteristic of landraces are com-
plex and dicult to express in dierent genetic backgrounds.
“a8259” — 2009/9/14 — 14:36 — page 20 — #20
i
i
i
i
i
i
i
i
20 A.C. Newton et al.
Maintaining these characteristics in heterogeneous landraces
is also problematic. Breeding,selection and deployment meth-
ods appropriate to these objectives should be used rather than
those used for high input intensive agriculture plant breeding.
10. PARTICIPATORY BREEDING
Participative approaches to agricultural research and de-
velopment are now extensively used throughout the world to
help define and address the practical research needs of farm-
ers. They have proved useful in solving practical problems in
complex and diverse farming systems characteristic of organic
farming and low input systems. During the last few years par-
ticipatory research divided into several dierent topics and in
the field of plant breeding there are already many very dif-
ferent strategies like Participatory Varietal Selection, true or
complete Participatory Plant Breeding (Witcombe et al., 1996)
and an intermediate approach, Ecient Participatory Breed-
ing (Morris and Bellon, 2004). The essential advantages of
participatory plant breeding over conventional plant breed-
ing involve: better targeting of local environmental conditions,
better definition of selection criteria important to the end-
users, faster and greater adoption of improved cultivars by the
farmer, and increase or maintenance of genetic variability. Par-
ticipatory plant breeding also gives voice to farmers and ele-
vates local knowledge to the status of science (Ceccarelli and
Grando, 1997).
Very e ective participatory plant breeding projects in ce-
reals are active all over the world. In Europe in particular,
a few participatory plant breeding projects are running with
success either in cross-pollinated and self-pollinate cereals. In
Portugal, the VASO (Vale do Sousa - Sousa Valley) project,
running since 1984, is a maize Participarory Plant Breeding
project developed to cover the needs of small maize farmers,
with scarce land resources, in poly-cropping systems for hu-
man uses, particularly bread production (Moreira, 2006). Lo-
cal germplasm is used, adapted to the local conditions over
centuries of cultivation, quality being the first priority over
quantity. This Participarory Plant Breeding project concerns
mainly flint-type open-pollinated landraces with quality for
the production of the traditional maize bread called ‘broa’.
This quality depends on traits not present in the available com-
mercial hybrid cultivars (Brites et al., 2008). ‘Broa’ production
still plays an important economic and social role in Central and
Northern Portuguese rural communities. Selection takes place
at the farmers’ field by the farmer in close collaboration with
the breeder.
The genetic diversity evolution through this participatory
breeding project was evaluated using molecular markers and
it was concluded that though interesting phenotypic improve-
ments were achieved, the level of genetic variability was
not significantly influenced and diversity was maintained
(Vaz Patto et al., 2007). Several maize open-pollinated lan-
draces were selected within this project with the joint collab-
oration of the breeder and the farmers. In Sweden, the Al-
lkorn project, running since 1995, is a cereal Participarory
Plant Breeding project for organic farming (Larsson, 2006).
The goal of the project is to identify and find interesting cereal
cultivars for organic farming in Sweden. Cultivars should be of
high quality for human nutrition and be well adapted to local
soils and climate. They should have good weed-suppressive
traits as well as tolerance to diseases and pests. The farmer
can select the best cultivar, from his point of view, from sev-
eral quality cultivars that he has chosen from the project and
tested in his own fields. The main aim of this breeding is not
to conserve the cultivars but to develop them for the future use
in organic farming all over the country. Cultivars of nearly all
species of cereals which have historically been used in Swe-
den are tested in this project, namely Triticum monococcum,
T. dicoccum,T. spelta, winter and spring wheat T. aestivum,
winter and spring rye Secale cereale, hulless barley, spring
barley, oats, and black oats. These cultivars have a broad di-
versity and cover primitive cultivars, older landrace cultivars
and early Swedish breeding cultivars from 1900 to 1950, since
in this early breeding period one of the parents of the cultivars
was often a local landrace. The cultivarsare selected each year
for better adaptation to organic conditions. The idea is to form
province groups of farmers that could help each other with
seed supply, seed cleaning, shelling of spelt wheat, milling
and selling. Regional production groups are formed for the lo-
cal market with help of local mills and local bakers and the
consumer can then find heritage cultivars from each region,
country or province. The goal of this type of participatory re-
search is empowering the farmers: supporting the formation
of groups capable of assessing their own needs and addressing
them either directly or through demands on research organisa-
tions.
In Hungary several Participarory Varietal Selection projects
are running both in on farm cereal landrace conservation and
organic farming research. The on-farm conservation Partici-
parory Varietal Selection strategy is mainly connected with the
regeneration of traditional Hungarian landraces and old culti-
vars which were maintained only in ex situ collections over
the last 20 years. Several organic farmers are involved in such
projects and their main interest is related to the traditional use
of special local food and heritage farming in relation to agro-
tourism. In such Participarory Varietal Selection programmes,
the Hungarian National Genebanks provide accessions to sev-
eral farmers participating in the in situ conservation project,
and the locations for long-term maintenance are chosen by all
participants in order to try to find the optimal place (environ-
mental condition) and landrace to maintain, considering the
social background of the region too (Holly, 2000). In recent
years new projects have been initiated in the field of Partici-
parory Varietal Selection in Hungary. One of the biggest ones,
entitled “Selection of suitable cultivars for organic farming”
was carried out in a real Participarory Varietal Selection sys-
tem in which most of the important stakeholders - the breeders,
seed producers, farmers and end users - were involved in cul-
tivar selection for several field crops, including wheat, barley,
sunflower, pea, etc. (Kovács, 2006a). In this project breeding
institutions provided both landraces and old and modern cul-
tivars and farmers and end user made the cultivar evaluation
independently of the providers’ opinion in each locality tak-
ing into consideration the end users priorities (Kovács et al.,
“a8259” — 2009/9/14 — 14:36 — page 21 — #21
i
i
i
i
i
i
i
i
Cereal landraces for sustainable agriculture. A review 21
2006). A similar research programme is running for under-
utilised cereal species, such as einkorn, emmer and macha
wheat in the country regions where such species were tradi-
tionally used. In this case the re-introduction of the forgot-
ten species into modern agriculture it is important to exploit
the genetic diversity which still exists in such cereals (Kovács,
2008).
A participatory breeding programme in tetraploid wheat
was initiated in 2001 at INRA-Montpellier, France, based on
a demand for organic pasta by the industry (Desclaux, 2005).
The quality of the wheat cultivars being produced under or-
ganic conditions did not meet the requirements of the pro-
cessing industry. To identify the main causes, a multidisci-
plinary public research team, bringing together plant breeders,
soil scientists, ecologists,agronomists and sociologists was as-
sembled to work in close collaboration with the farmers and
end-users. The two main French territories involved were Ca-
margue and Pays Cathare (Desclaux et al., 2002). The lack of
cultivars adapted to low nitrogen conditions became apparent
rapidly in such organic crop systems. The main aim of the on-
farm and participatory breeding is to take on-board farmers
preferences and to better target local environmental conditions
by increasing and managing the genetic variability.
Old durum wheat cultivars, segregating or advanced pure
lines and populations resulting from crosses between durum
wheat and emmer or wild species were provided to the farm-
ers by the breeders. Selection and evaluation are done in close
collaboration and results are discussed between all the partici-
pants. Information is feed back and can lead to re-examination
of the objectives of the breeding scheme (Desclaux, 2005).
This French project is neither a farmer-led nor a formal-led
project but it is led by both professionals and researchers and
requires farmer’s critical participation right from the first steps
of the breeding scheme, i.e. it is collegiate and decentralised
(Desclaux, 2005).
In the UK, a project for developing appropriate participa-
tory methodologies for cereal seed production and cultivar se-
lection under organic conditions has been established (Wolfe
and Hinchslie, 2005). This project aimed to overcome the
deficiencies of the UK ocial Recommended List of Cereals
for organic farmers. In winter wheat comparative trials using
cultivars and mixtures of cultivars were established at 19 UK
farms the 2003/04 and 2004/05 seasons (Clarke et al., 2006;
Jones et al., 2006). A collegiateparticipatory method was used
to balance statistical rigour with farmer’s objectives of man-
aging a whole farm system. Participating farmers sowed the
seed in large marked plots using their standard methodology
within a field containing wheat. Researchers gained informa-
tion from each of the farmers about their farming system, field
and trial. These experiments showed the large variability of
organic systems in the UK and the diculty of selecting a sin-
gle cultivar suitable for them all. This project also helped to
develop a small core of trained farmers and researches which
can be exploited in further participatory projects.
The diversity available within landraces is useful for breed-
ing purposes in at least four dierent ways (Ceccarelli et al.,
1987), including the release of the highest yielding lines as
pure line cultivars, the utilisation of superior lines as parents,
the evaluation of multilines built with a variable number of
pure lines, and the identification of lines showing extreme ex-
pression of specific attributes.
Participatory plant breeding and variety selection has
proven more successful than the approach used in high in-
put breeding programmes for landrace improvement in stress-
prone environments where sustainable approaches are a high
priority. Despite being more complex to carry out, it not only
delivers improved germplasm, but also aids uptake and com-
munication between farmers, researchers and advisors for the
benefit of all.
11. LEGAL ISSUES
Landraces as an important genetic resource have been in-
cluded in international treaties and national decrees that pro-
tect and enhance their use in their local environment. The
objectives of the convention on biological diversity are the
conservation of biological diversity, the sustainable use of its
components and the fair and equitable sharing of the benefits
arising out of the utilisation of genetic resources. The objec-
tives of the International treaty on plant genetic resources for
food and agriculture (http://www.planttreaty.org) are the con-
servation and sustainable use of plant genetic resources and the
fair and equitable sharing of benefits derived from their use, in
harmony with the Convention on Biological Diversity, for sus-
tainable agriculture and food security. The treaty promotes or
supports, as appropriate, farmers and local communities’ ef-
forts to manage and conserve on-farm their genetic resources
for Food and Agriculture.
In Europe the marketing of seeds of landraces and conser-
vation of cultivars is ruled by the strict seed trade and cultivar
protection laws. Cultivars have to meet the Distinctiveness,
Uniformity and Stability standards to be registered legally.
Agricultural crops have to meet the Value of Cultivation and
Use criteria and to be listed on the “European common cata-
logue of cultivars of agricultural plant species” to be tradable
within Europe. Combined with the extension of the intellectual
property rights in the UPOV (French: Union internationale
pour la protection des obtentions végétales) 91 act, the mainte-
nance of agro-biodiversity on-farmand the conservation of lo-
cal cultivars are generally threatened. According to the Biodi-
versity Action Plan for agriculture presented by the European
Commission in 2001 “the conservation and improvement of in
situ /on farm plant genetic resources also depends on the ef-
fective possibility of sustainable uses and on legislation which
makes it possible to market diversified genetic materials”.
The directive of conservation cultivars is still limiting the
diversity of seed. The most recent draft of a directive of con-
servation cultivars has the same goals as the UPOV standards:
Uniformity, stability and distinctness, but less documentation
is needed to fulfil the requirements. Landraces are never uni-
form, neither are they stable. Therefore, landraces will not be
allowed according to the directive of conservation cultivars.
The directive of conservation varieties therefore applies for
old out-dated approvedcultivars, and will not contribute to the
biodiversity of arable land (SANCO, 2006). An international
“a8259” — 2009/9/14 — 14:36 — page 22 — #22
i
i
i
i
i
i
i
i
22 A.C. Newton et al.
commission has written a manifest on the future of seed in
order to strengthen the movement toward sustainable agricul-
ture, food sovereignty, biodiversity and agricultural diversity,
to help defend the rights of farmers to save, share, use and
improve seeds and enhance our collective capacity to adapt to
the hazards and uncertainties of environmental and economic
change (Shiva, 2007). The most recent European Union Com-
mission Directive 2008/62/EC published on 20th June 2008
does provide for certain derogations for acceptance of agricul-
tural landracs and varieties which are naturally adapted to the
local and regional conditions and threatened by genetic ero-
sion and for marketing of seed of those landraces and varieties.
Clearly legislation was designed primarily to protect trade
and return royalty income to modern plant breeders with ex-
pensive programmes to fund. As the desirability of using lan-
draces becomes more apparent to achieve greater sustainabil-
ity, legislation changes are being made to to facilitate this trade
too. However, more changes are needed to promote the ex-
ploitation of diversity in landraces and encourage their use.
12. CONCLUSIONS
The position of cereal landraces as valuable genetic ma-
terial in contemporary agriculture is gaining renewed impor-
tance. A lot of recent research eort has gone into collecting,
organising, studying and analysing them with a primary goal
being to incorporate their positive qualities in new cultivars
or populations for a more sustainable agricultural production,
particularly in response to recent climate changes. Positive at-
tributes include landraces being a source of novel resistance
genes or combination of genes with a good deployment strat-
egy. Particularly important is the fact that resistance found in
cereal landraces has been durable in many instances, in con-
trast to the gene-for-gene type of resistance that is usually en-
countered in modern cultivars. However, the durability may
be associated with and dependent on deployment in heteroge-
neous populations.
A major part of this valuable landrace diversity is conserved
in the world’s gene banks network and should be exploited
systematically for traits such as quality and specific adapta-
tion to stress environments. The available genetic variation in
adaptive responses to soil and climatic conditions conserved in
landraces is little understood known and even less used. More
uniform and user-friendly documentation about collection and
characterisation of landraces, either morphologically or with
molecular tools, is needed to access this variation more eec-
tively. Gene banks should aim at adopting a common concept
of landraces and plan special inventories for them. The level of
diversity should be monitored during their conservation so that
the original level of variation is maintained. More studies are
needed in order to investigate if their long-term maintenance
by farmers resulted in increasing genetic variation.
New high-throughput genotyping platforms and phenotyp-
ing data in common databases will enable powerful associa-
tion genetics approaches to be utilised for improvement and
utilisation of landrace resources. Knowledge of linkage dis-
equilibria in landraces compared with elite germplasm will
help focus breeding for stress-adapted cultivars or popula-
tions. With the current rapid advancements in high-throughput
molecular genotyping technologies, it is becoming increas-
ingly clear that the limiting factor in applying those powerful
technologies to molecular breeding programmes is no more
the capacity of genotyping, but the potential for accurate or
precision phenotyping. Particularly important and challenging
is phenotyping for the so-called quantitative traits.
The renewed focus on cereal landraces for breeding pur-
poses is also a response to some negative consequences of
modern agriculture and conventional breeding, such as the
liberal use of high inputs, the loss of genetic diversity, and
the stagnation of yields in less favourable areas. To deliver
this, participatory plant breeding and variety selection prac-
tices have emerged as a powerful way to merge breeders’
knowledge and farmers’ selection criteria, emphasizing de-
centralised selection in the target environments with the active
participation of local farmers. Location-specific adaptation of
these diverse landraces will be important for further selection
at the farmer level. New strategies are emerging to produce
“modern landraces” based on multiple cross populations of
einkorn, emmer and bread wheat in combination with on farm
site-specific selection to obtain highly adaptable populations
for local and regional production.
The reported practice of the “inexplicable seed replace-
ment” by traditional farmers is connected to the gradual seed
degradation during landrace maintenance, brought about by
the established negative correlation between breeding and
competitive ability. Further enhancement of productivity and
stability is achieved through practicing “non-stop selection”
within landraces across the marginal production environments,
to exploit the constantly released by the genome useful adap-
tive variation. The procedure and results of non-stop selection
exceed those of conventional maintenance breeding, combat-
ing seed degradation and resulting to a constant cultivar up-
grade. An additional essential breeding consideration is the
creation of density-neutral landrace germplasm, since its use,
particularly in marginal environments, permits lower seeding
rates, limiting the damage due to drought.
The issues of conservation and sustainable use of landraces
have been so far included in international convention treaties
and directives which need to be implementated on European
and national levels. In this review we have highlighted the
value of landraces as resources for the future sustainability of
cereal crop production, the methods to enhance their genetic
makeup and avoid seed degradation, and emphasised the level
of coordination and resourcing needed to realise their great
potential.
Acknowledgements: We thank the European Union for funding the
COST Action 860 “Sustainable variety development for low-input and organic
agriculture” (2004-2008) through which this review was initiated.
REFERENCES
Abdellaoui R., M’Hamed H.C., Naceur M.B., Bettai¨eb-Kaab L., Hamida
J.B. (2007) Morpho-physiological and molecular characterization
of some Tunisian barley ecotypes, Asian J. Plant Sci. 6, 261–268.
“a8259” — 2009/9/14 — 14:36 — page 23 — #23
i
i
i
i
i
i
i
i
Cereal landraces for sustainable agriculture. A review 23
Abo-Elenin R.A., Heakal M.S., Gomaa A.S., Moseman J.G. (1981)
Studies on salt tolerance in barley and wheat. Source of tolerance
in barley germplasm, Barley Genet. IV, 402–409.
Acevedo E., Fereres E. (1993) Resistance to abiotic stresses. in: Hayward
M.D., Osemark N.O., Romagosa I. (Eds.), Plant Breeding and
Prospects, Chapman & Hall, London, UK, pp. 406–421.
Adom K.K., Liu R.H. (2002) Antioxidant activity of grains, J. Agr. Food
Chem. 50, 6182–6187.
Afanasenko O.S., Terentyeva I.A., Makarova I.N. (2000) Landraces from
Peru – new sources of resistance to net blotch of barley, Proceedings
of the 8th International Barley Genetics Symposium, Adelaide,
October 22–27 2000, Vol. II, pp. 71–72, Adelaide University,
Australia.
Akar T., Özgen M. (2007) Genetic Diversity in Turkish Durum
Wheat Landraces, in: Wheat Production in Stressed Environments,
Springer, Netherlands, 12, pp. 753–760.
Akhkha A., Clarke D.D., Dominy P.J. (2003) Relative tolerances of wild
and cultivated barley to infection by Blumeria graminis f.sp. hordei
(Syn. Erysiphe graminis f.sp. hordei). II – the eects of infection
on photosynthesis and respiration, Physiol. Mol. Plant. Path. 62,
347–354.
Alamerew S., Chebotar S., Huang X., Röder M., Börner A. (2004)
Genetic diversity in Ethiopian hexaploid and tetraploid wheat
germplasm assessed by microsatellite markers, Genet. Resour. Crop
Ev. 51, 559–567.
Al Khanjari S., Hammer K., Buerkert A., Ro¨der M.S. (2007) Molecular
diversity of Omani wheat revealed by microsatellites: II. Hexaploid
landraces, Genet. Resour. Crop Ev. 54, 1407–1417.
Alemayehu F., Parlevliet J.E. (1997) Variation between and within barley
landraces, Euphytica 94, 183–189.
Allard R.W., Bradshaw A.D. (1964) Implications of genotype-
environmental interactions in applied plant breeding, Crop Sci. 4,
503–508.
Almekinders C.J.M., Louwaars N.P., de Bruijn G.H. (1994) Local seed
systems and their importance for an improved seed supply in devel-
oping countries, Euphytica 78, 207–216.
Anderson J.B., Torp J. (1986) Quantitative analysis of the early pow-
dery mildew infection stages on resistant barley genotypes, J.
Phytopathol. 115, 173–185.
Andrade V., dos Santos T.M.M., Afonso Morales D., Costa G., Pinheiro
de Carvalho M.A.A. (2007) Evaluation of wheat germplasm at
the Madeira and Canary Archipelagos, using a single molecular
marker. A rapid screening method for durum wheat identification,
Cereal Res. Commun. 35, 1397–1404.
Anikster Y., Noy-Meir I. (1991) The wild-wheat field laboratory at
Ammiad, Israel J. Bot. 40, 351–362.
Anikster Y., Wahl I. (1979) Coevolution of rust fungi on Gramineae and
Liliaceae and their hosts, Ann. Rev. Phytopathol. 117, 367–430.
Anikster Y., Manisterski J., Long D.L., Leonard K.J. (2005) Leaf rust and
stem rust resistance in Triticum dicoccoides populations in Israel,
Plant Dis. 89, 55–62.
Aniol A., Madej L. (1996) Genetic variation for aluminum tolerance in
rye, Vortr. Pflanzenz, Chtg. 35, 201–211.
Annichiarico P. (2002) Defining adaptation strategies and yield stability
targets in breeding programmes, in: Kang M.S. (Ed.), Quantitative
genetics, genomics and plant breeding, CABI, Wallingford, UK, pp.
165–18