Kew Bulletin

, Volume 65, Issue 4, pp 561–576

A global approach to crop wild relative conservation: securing the gene pool for food and agriculture


  • Nigel Maxted
    • School of BiosciencesUniversity of Birmingham
  • Shelagh Kell
    • School of BiosciencesUniversity of Birmingham
  • Álvaro Toledo
    • FAO Commission on Genetic Resources for Food and Agriculture
  • Ehsan Dulloo
    • Bioversity International
  • Vernon Heywood
    • School of Biological SciencesUniversity of Reading
  • Toby Hodgkin
    • Bioversity International
  • Danny Hunter
    • Bioversity International
  • Luigi Guarino
    • Global Crop Diversity Trust
  • Andy Jarvis
    • International Centre for Tropical Agriculture, Recta Cali-Palmira
  • Brian Ford-Lloyd
    • School of BiosciencesUniversity of Birmingham

DOI: 10.1007/s12225-011-9253-4

Cite this article as:
Maxted, N., Kell, S., Toledo, Á. et al. Kew Bull (2010) 65: 561. doi:10.1007/s12225-011-9253-4


In light of the growing concern over the potentially devastating impacts on biodiversity and food security of climate change and the massively growing world population, taking action to conserve crop wild relatives (CWR), is no longer an option — it is a priority. Crop wild relatives are species closely related to crops, including their progenitors, many of which have the potential to contribute beneficial traits to crops, such as pest or disease resistance, yield improvement or stability. They are a critical component of plant genetic resources for food and agriculture (PGRFA), have already made major contributions to crop production and are vital for future food security; their systematic conservation in ways that ensure their continuing availability for use is therefore imperative. This is a complex, interdisciplinary, global issue that has been addressed by various national and international initiatives. Drawing on the lessons learnt from these initiatives we can now propose a global approach to CWR conservation, the key elements of which are: (1) estimating global CWR numbers, (2) assessment of the global importance of CWR diversity, (3) current conservation status, (4) threats to CWR diversity, (5) systematic approaches to CWR conservation, (6) CWR informatics, and (7) enhancing the use of CWR diversity.

Key Words

conservationcrop diversitycrop wild relativesgenetic diversityplant genetic resources for food and agriculture


In light of the growing concern over the predicted devastating impact of climate change on global biodiversity and food security, coupled with a growing world population, taking action to conserve crop wild relatives (CWR) has become a high priority. Crop wild relatives are species with a close genetic similarity to crops and many of them have the potential or actual ability to contribute beneficial traits to these crops, such as resistance to biotic and abiotic stresses, and higher, more stable yields (Prescott-Allen & Prescott-Allen 1986; Hoyt 1988; Maxted et al.1997a; Tanksley & McCouch 1997; Meilleur & Hodgkin 2004; Stolton et al.2006). A pragmatic approach not without problems places all species of the same genus as a crop in this category (Heywood 1994; Maxted et al.2006). CWR have already made major contributions to crop production, and are vital for future food security. The systematic conservation of this critical component of plant genetic resources for food and agriculture (PGRFA) in ways that ensure their continuing availability for use is therefore imperative.

Darwin (1868) observed “… it appears strange to me that so many of our cultivated plants should still be unknown or only doubtfully known in the wild state”. It was the Russian botanist N. I. Vavilov who fully recognised and championed the potential of CWR for crop improvement in the 1920s and 30s, referring to the use of wild Aegilops L., Secale L., Haynaldia Kanitz and Agropyron Gaertn. species in wheat breeding, for example (Vavilov 1949). CWR were first routinely used by agricultural scientists to improve major crops in the 1940s and 50s and by the 1960s and 70s this was leading to some major breeding improvements (Meilleur & Hodgkin 2004).

There has been increasing interest in CWR conservation and use in recent years, arising from increased recognition of their value as well as increasing ease of use. These are complex, interdisciplinary, global issues that have been addressed by various national and international initiatives, including two Global Environment Facility projects (‘In situ Conservation of Crop Wild Relatives through Enhanced Information Management and Field Application’ and ‘Design, Testing and Evaluation of Best Practices for in situ Conservation of Economically Important Wild Species’), the European Community-funded project ‘European Crop Wild Relative Diversity Assessment and Conservation Forum (PGR Forum)’, the FAO commissioned ‘Establishment of a global network for the in situ conservation of crop wild relatives: status and needs’, the IUCN Species Survival Commission Crop Wild Relative Specialist Group and the European ‘In Situ and On Farm Network’. The need to address CWR conservation is also highlighted in international and regional policy instruments, such as the Convention on Biological Diversity (CBD 1992), FAO Global Plan of Action for the Conservation and Sustainable Utilization of PGRFA (FAO 1996), European Community Biodiversity Strategy (EC 1998), CBD Global Strategy for Plant Conservation (CBD 2002), International Treaty on Plant Genetic Resources for Food and Agriculture (FAO 2001), European Plant Conservation Strategy (Planta Europa 2001), Global Strategy for CWR Conservation and Use (Heywood et al.2008), and most recently in the European Strategy for Plant Conservation (Planta Europa 2008). This last recommends the establishment of 25 CWR genetic reserves in Europe and undertaking gap analysis of current ex situ CWR holdings, followed by filling of diversity gaps. Drawing on the experience of these initiatives, and to help meet international and regional obligations, this paper aims to outline a coherent policy for CWR conservation and use to be implemented over the next 10 years. CWR conservation and use provides an excellent exemplar of how to address the dual Millennium Development Goals of combining biodiversity conservation with poverty alleviation.

Estimating global CWR number

Recent studies have found that using the broad definition of a CWR (i.e. all species in the same genus of the crop) the number of CWR species of interest may be much larger than previously recognised. For example, using this definition, Kell et al. (2008) found that 17,495 (8,624 of them endemic) out of approximately 20,590 species, or 85% of the European flora, comprises crop and CWR species. Practically, with such large numbers there is a need for more effective prioritisation. Maxted & Kell (2009) calculated that the 77 major and minor food crops as defined by Groombridge & Jenkins (2002) contained around 10,700 species using the numbers of species per genus from Mabberley (2008). However, this number, like the European estimate, is likely to be inflated by the inclusion of remote as well as closely related CWR found in the same genus as the crop. Maxted et al. (2006) argued that a more precise target could be obtained by focusing on the crop gene pool GP1B or taxon groups TG1b and TG2 alone, which contain the closest CWR species. Based on an initial sample of 14 food crop groups (see Table 1) Maxted & Kell (2009) estimated that globally, for the 77 major and minor food crops, there are 221 very close wild relatives and 471 close wild relatives. Thus, as a working estimate, we may need to deal globally with around 700 closely related CWR species (i.e., less than 0.26% of the world flora) in order to ensure that the highest priority genetic diversity of major and minor food crops is conserved and made available for use in crop improvement programmes as a contribution to future worldwide food security. In addition, CWR of forage species which also contribute to food security will need to be identified and conserved. A case can also be made for similar action for the CWR of fuel and fibre crops and industrial and ornamental crops although they are not the primary concern of this paper.
Table 1.

Numbers of primary and secondary CWR species in 14 crop gene pools (Maxted & Kell 2009).


Crop taxon

Species in genus

Primary CWR species

Secondary CWR species

% Priority in genus

Finger millet

Eleusine coracana Gaertn.






Hordeum vulgare L.





Sweet potato

Ipomoea batatas (L.) Poir.

600 – 700





Manihot esculenta Crantz






Musa acuminata Colla






Oryza sativa L.





Pearl millet

Pennisetum glaucum (L.) R. Br.

80 – 140




Garden pea

Pisum sativum L.






Solanum tuberosum L.






Sorghum bicolor (L.) Moench






Triticum aestivum L.

6 + 22




Faba bean

Vicia faba L.






Vigna unguiculata (L.) Walp.






Zea mays L.







2117 – 2277








Therefore, a goal over the next ten years is to refine the working estimate of global highest priority CWR for food and agriculture. Once such a list is available it would greatly facilitate the targeting of in situ and ex situ conservation actions. In the longer-term, however, it would be unwise to restrict conservation to these species alone, as we know those traits desired by germplasm users are not exclusively located in the closest CWR species, so a secondary priority would be the genetic conservation of all 10,739 CWR species congeneric to the 77 major and minor food crops.

Assessment of global importance of CWR diversity CWR, like other wild species, may be valuable components of ecosystems but many of them also have additional specific value as gene donors for plant breeding. Prescott-Allen & Prescott-Allen (1986) calculated that the yield and quality contribution by CWRs to US grown or imported crops was over US$350 million a year, while Phillips & Meilleur (1998) estimated that potential losses associated with endangered food crop wild relatives were around US$10 billion annually in wholesale farm values. Further, Pimentel et al. (1997) estimated that the contribution of genetic resources to yield increase is about 30% of production, and much of this results from CWR species, so the introduction of new genes from wild relatives contributes approximately $20 billion toward increased crop yields per year in the US and $115 billion worldwide.

Despite their known value as gene donors, Tanksley & McCouch (1997) argued that breeders were not fully exploiting the potential of CWR because, historically, they relied on searching for specific beneficial phenotypic traits associated with particular CWR taxa, rather than searching for beneficial genes. Although it would be very difficult to give a precise estimate of CWR use by breeders because the data are likely to be commercially sensitive and therefore not readily available, Maxted & Kell (2009) recently reviewed the use of CWR in crop improvement and cited 291 articles reporting the identification and transfer of useful traits from 185 CWR taxa into 29 crop species. They found that the degree to which breeders had used CWR diversity varies markedly between crops, being particularly prominent in barley, cassava, potato, rice, tomato and wheat. Of these, rice and wheat are the crops in which CWR have been most widely used, both in terms of the number of CWR taxa used and successful attempts to introgress traits from the CWR to the crop (Fig. 1). The most widespread CWR use has been and remains in the development of disease and pest resistance, with references citing disease resistance for 39% of inter-specific trait transfers, pest resistance 17%, abiotic stress 13%, yield increase 10%, cytoplasmic male sterility and fertility restorers 4%, quality improvers 11% and husbandry improvement 6% of the reported inter-specific trait transfers. It is also notable that the number of publications detailing the use of CWR in breeding has increased gradually over time (Maxted & Kell 2009). The exploitation of the potential diversity contained in CWR species remains poorly directed as the approach by breeders to CWR use has not been systematic or comprehensive; therefore, most CWR diversity that might be used for breeding remains untapped. Hajjar & Hodgkin (2007) comment that although quantitative trait loci have been identified in many CWR species, the potential to exploit them as a breeding resource using new molecular technologies has yet to be fully realised. Although this situation is likely to improve with time, it does underpin the need for the continued availability of a broad range of CWR diversity, also emphasising the conservation-use linkage and the need for the conservation community to meet the evolving needs of the users. Therefore, the goal over the next ten years is to take advantage of novel technological advances in trait recognition and inter-specific breeding to extend the breadth of CWR use to a broader range of crops and systematically review the potentially useful diversity in CWR gene pools.
Fig. 1

The number of references reporting the identification and transfer of useful traits from 185 CWR taxa to 29 crop species, showing the number of CWR taxa used in each crop (Maxted & Kell 2009).

Threats to CWR diversity

CWR are biologically no different from other wild plant species, and, like them, many are currently threatened with loss of diversity and/or extinction, as a result of anthropogenic influences: habitat destruction and fragmentation, unsustainable resource exploitation, changes and intensification of land management, and invasive species. In addition to these factors, and interacting with them, accelerated climate change is likely to present a step-shift in terms of extinction and genetic erosion (IPCC 2007) and this threat requires urgent action to avoid food insecurity (Lobell et al.2008). In Europe, Thuiller et al. (2005) modelled projections of the future distribution of 1350 plant species and found that by 2080 more than half of them could be vulnerable or threatened by climate change. In one of the few studies of the effects of climate change on CWR, Jarvis et al. (2008) projected that 16 – 22% of wild Arachis L., Solanum L. and Vigna Savi species would go extinct by 2055. Similar modelling of climate change scenarios in Mexico by Lira et al. (2009) highlight that most of the eight wild Cucurbitaceae taxa studied are predicted not to survive under accepted climate change models. A further threat that is specific to CWR diversity is that many CWR of major crops are found in disturbed, pre-climax communities (Jain 1975), which are the habitats that are likely to be subject to increasing levels of anthropogenic change and destruction.

The most commonly applied means of assessing threats to wild taxa is the application of the IUCN Red List criteria (IUCN 2001) but CWR have thus far not been specifically prioritised for assessment, although it should be noted that a Red Book of Crop Wild Relatives in Bolivia (Libro Rojo de Parientes Silvestres de Bolivia) has recently been published (Mora et al.2009). Of the 14 food crop groups studied by Maxted & Kell (2009), only two crop groups had been assessed — one sweet potato relative, Ipomoea pulcherrima Ooststr. was assessed as being vulnerable and 44 potato (Solanum) species were assessed as being threatened to varying degrees.

Although it is difficult to quantify the loss of genetic diversity within CWR species, it is likely to be very much greater than the loss of species given that most of the species that are able to survive the threats to which they are exposed will suffer some genetic erosion (loss of genetic diversity (Maxted et al. 1997b). It therefore seems likely that virtually all CWR species are currently suffering some degree of loss of genetic diversity. Maxted et al. (1997b) estimated that 25 – 35% of plant genetic diversity would be lost between the ratification of the CBD in 1993 and the 2010 Biodiversity Target date. The magnitude and rapidity of climate change, coupled with other threats, is likely to impose extreme selection pressure on the surviving populations of CWR over the coming 50 – 100 years. Loss of any genetic diversity means that species may not be able to adapt to changing conditions quite so readily in the future, but also that vital diversity necessary to underpin our future food security will not be available to breeders.

A goal over the next ten years should be to undertake systematic threat assessment for as wide a range of CWR taxa as possible, using IUCN or national criteria, or both. This is a critical element of the IUCN CWR Species Survival Commission Specialist Group’s Operational Strategy (see As a first step, the global priority 700 CWR species should be assessed, then as a secondary priority all 10,739 CWR species related to the 77 major and minor food crops. The IUCN Red List criteria are not entirely suitable for assessing the threat to genetic erosion or extinction within species, and so a subsidiary goal will be the development of a practical means of assessing threat to genetic diversity within species. Likewise, the current IUCN criteria do not take climate change specifically into account, although plans are in hand to do so.

Another urgent goal should be to undertake bioclimatic modelling of as many CWR as possible, so that we can obtain as accurate a picture as possible of their likely adaptation, migration or loss.

Active CWR Conservation

A complementary approach to CWR conservation should be adopted, encompassing in situ conservation in natural habitats as well as ex situ measures focused mainly on seed storage. As for other wild species, the preferred means of conserving CWR is in situ, as living populations in protected areas or genetic reserves1 or as seed samples dried and stored at sub-zero temperatures in gene banks (Maxted et al.1997c; Heywood & Dulloo 2006; Stolton et al.2006; Iriondo et al.2008a). Currently, the highest proportion of CWR diversity is actively conserved ex situ, although the coverage is far from comprehensive. Although CWR diversity undoubtedly occurs in existing protected areas, the CWR species are rarely actively managed and monitored in them, so their conservation status is uncertain and populations or entire species could be at risk without the protected area management being aware of the threat or consequences.

The First Report of the State of the World’s PGRFA (FAO 1998) listed 4% of governmental, 14% of CGIAR and 6% of private gene bank holdings as wild species, while the Second Report of the State of the World’s PGRFA (FAO 2009) listed 10% of gene bank holdings as wild species and concluded “Interest in collecting and maintaining collections of CWR is growing as land-use systems change, concerns about the effects of climate change grow and techniques for using the material become more powerful and more readily available”. So it appears that ex situ collections and interest in conserving CWR has increased, yet FAO (2009) ultimately concludes that “For several major crops, such as wheat and rice, a large part of the genetic diversity is now represented in collections. However, for many other crops, especially many neglected and under-utilized species and CWR, comprehensive collections still do not exist and considerable gaps remain to be filled”. Those CWR collections that do exist, because of the ad hoc manner in which they have largely been collected, are unlikely to constitute genetically representative samples of global CWR diversity.

As an indicator of the coverage of ex situ collections, analysis of the data in the European crop gene bank portal (EURISCO) revealed that CWR taxa account for 5.6% of total germplasm holdings, and that the 1,095 CWR species included represent only 6% of the 17,495 CWR species found in Europe (O’Regan 2007). The ratio of cultivated to wild species in ex situ collections is 12:1, though most diversity is located in wild species (Maxted et al.2008a). In contrast, analysis of European ex situ seed collections held in botanic garden gene banks via the ENSCONET portal revealed that CWR taxa account for 61.8% of total germplasm holdings, and that the 5756 CWR species included represent 33% of 17,495 CWR species found in Europe. Initially it would appear that botanic garden gene banks do a much better job of conserving CWR diversity than crop gene banks. However, care should be taken in drawing too crude a conclusion from this stark difference, as it is likely that the 1,095 CWR species held in crop gene banks are more likely to include high priority CWR species from the primary and secondary gene pools. Moreover, many of the botanic garden accessions are small and genetically poorly sampled (Heywood 1999).

The effectiveness of in situ CWR conservation is, if anything, more uncertain than ex situ, and although there has been considerable attention paid to the theory of design, establishment, management and monitoring of CWR diversity in genetic reserves (see Jain 1975; Hoyt 1988; Gadgil et al.1996; Hawkes et al.1997; Maxted et al.1997a, c; Safriel et al.1997; Heywood & Dulloo 2006; Stolton et al.2006; Iriondo et al.2008a), full practical implementation remains limited (Maxted et al.1997a, c; Meilleur & Hodgkin 2004). In practice, conservation of CWR is often planned within existing protected areas because: a) by definition they are protected and managed (although not always effectively) and already have an associated long-term conservation ethos and are less prone to short-term management changes, b) in some cases, it is relatively easy to negotiate changes in the existing site management plan so as to facilitate genetic conservation of CWR species, and c) the creation of new conservation sites can be avoided along with the problems and cost of acquiring new land for conservation (Iriondo et al.2008a). There are some notable examples of activities that have made a significant contribution to the process of conserving CWR in situ (see Table 2), but even in some of these cases the sites are not managed in the most appropriate manner to conserve CWR genetic diversity.
Table 2.

Examples of CWR conserved in protected areas.


Protected Area



Wild emmer wheat (Triticum turgidum subsp. dicoccoides)

Ammiad, Galilee


Anikster et al. (1997); Safriel et al. (1997)

Teosinte (Zea diploperennis Iltis, Doebley & R. Guzman)

MAB Sierra de Manantlán Biosphere Reserve


Sanchez-Velasquez (1991)

Wild wheats (Triticum turgidum subsp. dicoccoides, T. monococcum L., Aegilops tauschii Coss, A. speltoides Tausch)



Karagöz (1998)

Chestnut (Castanea sativa Mill.), wild plum (Prunus cerasifera Ehrh. var. divaricata)



Kűçűk et al. (1998)

Medicago L. spp., Vicia L. spp., Trifolium L. spp., Lathyrus L. spp., Lens Mill. spp., Triticum L. spp., Avena L. spp., Hordeum L. spp., Aegilops L. spp., Allium L. spp., Amygdalus L. spp., Prunus L. spp., Pyrus L. spp., Pistacia L. spp. and Olea L. spp.

Abu Taha


Al-Atawneh et al. (2008)






Wadi Sair




Wild wheats (Triticum boeoticum Boiss., T. urartu Thumanjan ex Gandilyan, T. araraticum Jakubz.)



Avagyan (2008)

Wild bean populations (Phaseolus L. spp.)

Central valley

Costa Rica

Zoro Bi et al. (2003); Baudoin et al. (2008)

Wild coffee (Coffea mauritiana Lam., C. macrocarpa A. Rich., C. myrtifolia (A. Rich. ex DC.) J.-F.Leroy)

Black River Gorges National Park


Dulloo et al. (1998)

Wild onions (Allium columbianum (Ownbey & Mingrone) P. M. Peterson, Annable & Rieseberg, A. geyeri S. Watson, A. fibrillum M. E. Jones)

Great Basin, Washington State


Hannan & Hellier in Pavek et al. (1999); Hellier (2000)

Wild grapevine (Vitis rupestris Scheele, V. shuttleworthii House, V. monticola Buckley)

Witchita Mountains and Ouachita National Forest, Oklahoma, Clifty Creek, Missouri


Pavek et al. (2003)

There are a number of potential approaches to systematic CWR conservation, but three distinct (though complementary) approaches may be adopted — individual, national and global (Maxted & Kell 2009). It is important to recognise, however, that conservation strategies are more likely to be successful if national governments, on-the-ground agencies and local people set the agenda, for it is they who will be responsible for their implementation, with international NGOs and IGOs playing a supporting role (Smith et al.2009). The individual approach involves a protected area or gene bank manager actively promoting CWR conservation within the site or facility that they manage. The national approach requires countries to develop CWR conservation strategies, which when implemented over time would result in the systematic representation of the nation’s CWR diversity in an in situ network of genetic reserves or other conservation areas, with complementary ex situ storage of genetically representative population samples in national gene banks (Fig. 2). This model for establishing a national CWR inventory was recently tested in the UK and Portugal (see Maxted et al.2007; Magos Brehm et al.2008) and showed that in the UK 17 sites contain 152 (67.3%) of the priority UK CWR species. However, so far, very few countries have developed a national CWR conservation strategy and the United Nations Environment Programme (UNEP)/Global Environment Facility (GEF)-supported project described below is exceptional in including such a national strategy for each of the five countries involved as one of its outputs. The global approach entails the establishment of a worldwide network of in situ genetic reserves that is independent of national political borders and focuses on worldwide priority crop gene pools, again with complementary ex situ storage of genetically representative population samples.
Fig. 2

Model for the development of national CWR strategies (Maxted et al.2007).

A global approach to the in situ gap analysis, that involves the comparison of naturally occurring CWR diversity with the diversity sampled and conserved either in situ or ex situ (see Maxted et al. 2008b), for 14 globally important food crop groups (finger millet, barley, sweet potato, cassava, banana/plantain, rice, pearl millet, garden pea, potato, sorghum, wheat, faba bean, cowpea and maize) was undertaken by Maxted & Kell (2009) and suggested priority locations for CWR genetic reserve establishment (see Map 1). Although further crop groups should be added, these priority sites can be used to begin recommendations for establishment of the global network. For each food crop group the identified priority sites are primarily within the boundaries of existing protected areas, although this is not always the case and the results do indicate the need to establish new protected areas. The results of the 14 crop complex analyses (with the exception of the Middle East and Eastern Congo) show few obvious opportunities for conservation of multi-crop gene pools in single genetic reserves,2 further research is likely to identify additional potential multi-taxon CWR genetic reserves locations where limited conservation resources could be effectively targeted.
Map 1

Global priority genetic reserve locations for wild relatives of 14 food crops (FAO 2009). The ‘centres of crop diversity’ (indicated by the enclosed lines) are likely to contain further priority sites for other crop gene pools.

A complementary global approach to ex situ gap analysis of 13 globally important food crop groups (chickpea, common bean, barley, cowpea, wheat, maize, sorghum, pearl millet, finger millet, pigeon pea, faba bean, and lentil) has also been undertaken. Jarvis and colleagues (Jarvis et al.2008) analysed 28,751 herbarium and gene bank specimen/accession occurrences accessible through the Global Biodiversity Information Facility (GBIF) for 643 CWR taxa belonging to the 13 gene pools (see They used a maximum entropy climate envelope model to create distribution maps for each species under current climates using the WorldClim database. The results show the species and geographic regions where high priority for genetic resource collection exists (see Map 2). The results show that significant gaps still exist in ex situ conservation, and that protected areas do not currently provide adequate conservation of the species. Priorities for collection occur especially in Africa, northern Australia, Central America and the Andes. As many as 39 species occur sympatrically in parts of Africa, where multiple gaps could be filled with targeted collecting or establishment of genetic reserves.
Map 2

Locations of priority for ex situ and/or in situ conservation based on the identification of gaps in our current ex situ conservation system for 13 important crop wild relative gene pools. A species level richness in gaps, and B genus level richness of gaps, showing sites where gaps exist in multiple gene pools.

Over the last 10 years the GEF has supported a number of projects that seek to enhance the conservation and use of crop wild relatives. One such project is the UNEP/GEF-supported project, ‘In situ conservation of crop wild relatives through enhanced information management and field application’, coordinated by Bioversity International. Five countries — Armenia, Bolivia, Madagascar, Sri Lanka and Uzbekistan — are involved in the project through their governments and other agencies. This project has expanded substantially the previously limited body of knowledge on in situ CWR conservation in developing countries. The project facilitated the prioritisation of CWR species from 36 different genera for ecogeographic assessments and threat assessment through Red Listing. Over 310 CWR species were assessed according to the current IUCN criteria and Bolivia has produced the first ever national Red Book of CWR. This is probably the largest set of such assessments undertaken for CWR and represents a major contribution to practice. Further, the partnership, through the involvement of protected area authorities and other relevant stakeholders such as local and indigenous communities, has seen the development of CWR species management plans for implementation in protected areas, as well as the adaptation of protected area management plans themselves so as to take into account the management needs for CWR conservation. Species management and monitoring plans have been developed to manage CWR diversity in protected areas in each project country (Table 3). In addition the project has identified areas outside protected areas for in situ CWR conservation, including those for Oryza nivara S. D. Sharma & Shastry and O. rhizomatis D. A. Vaughan in Puttalam District, Sri Lanka and Malus sieversii M. Roem., Juglans regia L. and Pistacia vera L. in Uzbekistan. Therefore, although significant progress has been made in recent years, CWR species, as well as the diversity within them, are seriously under-conserved both ex situ and in situ.
Table 3.

Examples of CWR conserved in protected areas in Armenia, Bolivia, Madagascar, Sri Lanka and Uzbekistan.

Crop Gene Pool


Protected Area



Dioscorea maciba Jum. & H. Perrier, D. bemandry Jum. & H. Perrier, D. antaly Jum. & H. Perrier, D. ovinala Baker and D. bemarivensis Jum. & H. Perrier.

Ankarafantsika National Park



Cinnamomum cappara-coronde Blume

Kanneliya Forest Reserve

Sri Lanka


Amygdalus bucharica Korsh.

Chatkal Biosphere Reserve



Triticum araraticum, T. boeoticum, T. urartu and Aegilops tauschii

Erebuni State Reserve



Theobroma L. spp.

Parque Nacional y Territorio Indigena Isiboro-Secure


It is critical that in the next ten years a more strategic approach is taken to targeting ex situ CWR conservation. First the initial strategic target should be to conserve ex situ genetically representative samples of the highest priority 700 globally important CWR food crop-related species, and second to ensure that there is systematic representation of the 10,739 priority CWR species related to the major and minor food crops conserved in gene banks. Further, the extension of the gap analysis of food crop groups by Jarvis and colleagues is strongly recommended to help prioritise the collection of the 77 major and minor food crop gene pools.

The establishment of genetic or other kinds of reserve for CWR in times of rapidly rising human population, climate change and ecosystem instability as well as a global economic crisis is a complex goal, which necessitates a carefully researched strategic approach. Sites competing for reserve status would need to be assessed and prioritised for their longer-term sustainability in terms of the predicted impact of climate change on the site and the development plans associated with local communities (Brooks et al. 2006). Some CWR populations conserved in vulnerable locations are unlikely to be able to adapt sufficiently quickly to climate change or even in doing so may go through a genetic bottleneck reducing the long-term viability of sustaining their diversity in situ. On the other hand, basing a strategy on ex situ action alone is unlikely to be economically feasible or practical given (a) the very large number of species involved (possibly as many as 229,500 spp.; Maxted & Kell 2009; see also discussion in Heywood 2009a), (b) the need to sample and conserve ecogeographically and genetically diverse populations for each species, and (c) the fact that few species have immediate use potential. As such, the strategic establishment of a global network of in situ CWR genetic reserves to complement targeted ex situ collections should be regarded as a priority. In situ CWR conservation has been primarily associated with genetic conservation in protected areas. However, CWR species are just as likely to be found outside existing protected areas as within them. In fact, as is argued above, many CWR of major crops are found in disturbed, pre-climax communities and these are more rarely designated as protected areas. Therefore, another target in the next ten years will be to further elaborate methodologies for the in situ conservation of CWR diversity outside of conventional protected areas that are still able to meet the objective of maximising long-term sustainability of CWR diversity. Various public and private initiatives to provide some degree of protection to species outside protected areas exist, especially in Australia, Brazil, China, Costa Rica, Mexico, South Africa, the USA and several European countries (Maxted et al.2008c; Heywood 2009b). One approach is to promote CWR in situ conservation in less formally designated protected areas such as Indigenous and Community Conserved Areas (ICCAs) (Pathak et al.2004; Kothari 2006, 2008; see also, where indigenous peoples and local communities have for millennia conserved natural environments and species for economic as well as cultural, spiritual and aesthetic reasons independent of more formal conservation sector interventions.

CWR Information Management

CWR conservation requires effective information management. There have been many initiatives to develop information management systems for PGRFA, one of which specifically addressed CWR information management. The FP5 funded “European Crop Wild Relative Diversity Assessment and Conservation Forum (PGR Forum)” (Maxted et al.2008a) developed the Crop Wild Relative Information System (CWRIS) (Kell et al.2008) with the goal of providing a model for collation and management of CWR conservation and sustainable use data and a system for accessing this information. This system makes available the PGR Forum Crop Wild Relative Catalogue of Crop Wild Relatives for Europe and the Mediterranean (, containing more than 25,000 species records and in excess of 273,000 records of taxon occurrences in 130 geographical units across the region, with a limited number of detailed case studies giving the Catalogue biological depth. The structure of CWRIS is generic and independent of the exemplar data initially included, and therefore could be used to meet the information management requirements of any CWR conservation and sustainable use data. An XML (Extensible Markup Language) schema corresponding to the data model is available (see Moore et al.2008). Recently CWRIS has been further developed and extended by the EC Gen Res project “An Integrated European In Situ Management Work Plan: Implementing Genetic Reserves and On Farm Concepts (AEGRO)”. CWRIS was used to manage population level data for gene pool based studies of Avena L., Beta L., Brassica L. and Prunus L. spp. (Frese & Maxted 2009). In the next ten years, CWRIS should be populated with other gene pool data and further developed and extended using data sets from other regions.

The GEF-supported project ‘In situ conservation of crop wild relatives through enhanced information management and field application’, has contributed to CWR information management through the establishment of national information systems on CWR in Armenia, Bolivia, Madagascar, Sri Lanka and Uzbekistan. The systems have facilitated the mapping of CWR distribution and identified priority areas for their conservation. In addition to the national information systems, a global portal (see, maintained by Bioversity International, has been developed through which national CWR inventories can be searched.

While progress has been made in CWR information management this progress has been largely focused on short-term research-based projects and there remains little coherent vision. There is a need to unite the various elements already existing, such as CWRIS ( for the taxonomic backbone and EURISCO ( for gene bank holdings, and to globalise their application to form a seamless information system from CWR diversity in the field, in situ and ex situ conserved diversity, through to characterised and evaluated CWR diversity, in order to ensure the highest priority CWR diversity is conserved and available to use by the broad user community. Meeting the needs of the latter is the raison d’être of CWR diversity conservation, so the system needs to be developed jointly with the germplasm user community to provide the quality of service that they require.

Enhanced CWR use

Conservation is not an end in itself. To be effective, conservation should be linked to use. In fact, sustainable use is seen as the long-term means of sustaining active conservation. Therefore, to encourage CWR conservation there is an associated need to promote use of the conserved diversity. There are numerous ways in which CWR use in breeding can be promoted, but traditionally this has focused on trying to identify traits of interest through phenotypic characterisation and evaluation. This has in many cases proved prohibitively expensive. The Report on the State of the World’s PGRFA (FAO 1998) highlights the fact that two thirds of globally conserved ex situ germplasm lack basic passport data, 80% lack characterisation data and 95% lack evaluation data, making the use of such germplasm, including CWR germplasm, much more difficult than it need be. Since the publication of the first SoW report, the Second Report on the State of the World’s PGRFA (FAO 2009) details several new international initiatives that support the increased characterisation and evaluation of germplasm, including the fairly widespread adoption of core collections that are adequately characterised and evaluated. However, it still concludes that “The country reports were virtually unanimous in suggesting that one of the most significant obstacles to a greater use of PGRFA is the lack of adequate characterization and evaluation data and the capacity to generate and manage such data”.

The situation regarding access to in situ conserved germplasm is likely to be even less conducive to exploitation — there are currently no known genetic reserves where the conserved species are fully characterised and evaluated. The bottleneck over systematic characterisation and evaluation has been acknowledged almost since the need for their conservation was recognised in the late 1960 s and early 1970 s (Frankel & Bennett 1970). It could be argued now that simply increasing the amount of ‘traditional’ characterisation and evaluation is unlikely to result in the required step change in the exploitation of CWR. This is despite the increasing knowledge of useful genes that can be found in CWR and what can be achieved by their introgression into crops (Maxted & Kell 2009). For these reasons there is a major incentive to embrace innovative technologies to improve CWR utilisation.

‘Next generation technologies’ offer a way of screening thousands of samples of germplasm for those interesting gene variants that are adaptively important, making them available for use in conventional breeding. Given that such useful forms of genes will essentially represent ‘natural’ genetic variation, this approach could even help to alleviate the concerns associated with GM technology. There are now many candidate genes shown to be involved in some way in drought tolerance, for instance, and large scale transcriptomics and resequencing will allow us to identify these genes and all the variants that can be found in the CWR related to many crops. A recent study of wild Brassica nigra (L.) W. D. J. Koch, B. rapa L. and B. oleracea L. populations from Northern and Southern Europe (Mitchell 2008) searching for genes associated with climate change adaptive characters (i.e. rainfall and temperature) located 42 genes which showed differential north-south expression across all three species (e.g. water channel-like protein, low temperature and salt responsive protein, wax synthase-like protein). The clear advantage of this approach is that rather than select CWR genetic diversity in general, the potential user can focus more directly on adaptive capacity. Genomic databases containing such information must increasingly be linked to web-enabled databases of ex situ conserved CWR germplasm, such as the European Internet Search Catalogue of Ex Situ PGR Accessions (EURISCO) (, System-wide Information Network for Genetic Resources (SINGER) ( and Germplasm Resources Information Network (GRIN) (

There is another, complementary approach to enriching data on conserved germplasm, and therefore stimulating its use. ‘Predictive characterisation’ is the use of spatial analysis to predict which germplasm might have desired traits. For example, GIS can be used to identify germplasm likely to be drought or frost tolerance by overlaying environmental data with the locality of populations (see Pollak & Pham 1989; Chapman & Barreto 1996; Guarino et al.2002; Kaur et al.2008). Similar approaches can be adopted for other abiotic traits (e.g., other climatic variables, salinity, soil mineral excesses or deficiencies, and day length requirements). For desired biotic characteristics, the distribution of the CWR taxon can be overlaid with the known distribution of pests or diseases. The CWR populations found to be coincident with high levels of pests and diseases are likely to have evolved resistance over time; therefore, these populations might be predicted to harbour the required pest or disease resistance. Further, if the pest or disease distribution is not precisely known, it is also possible to overlay the CWR taxon’s distribution with the climatic conditions suitable for that disease. For example, Bhullara et al. (2009) used this Focused Identification of Germplasm Strategy (FIGS) approach and starting with 16,089 accessions of bread wheat, reduced the number to 1,320 then isolated 7 new resistance alleles to powdery mildew (genePm3) from these accessions. However, caution is necessary, as contrary examples are found, where resistance alleles are discovered in area where the disease is unknown, such as the location of Rhizomania resistance in wild beet populations from Kalundborg Fjord, Denmark (Lothar Frese pers. comm.). It is more difficult to imagine how predictive characterisation might be used to predict yield or quality traits — therefore, for these traits, traditional characterisation and evaluation will remain necessary. While a literature review currently reveals limited use of GIS analysis for predictive characterisation, with an increased emphasis on the need to link in situ CWR conservation to use, these techniques offer innovative opportunities (Guarino et al.2002).

Ultimately, unless the professionals involved with CWR conservation can ensure that conserved germplasm is held in a form better suited for breeders and other user groups and that there is less of a barrier between conservation and utilisation, then the use of conserved CWR is not likely to improve. The rapid advancement of biotechnological techniques for transferring traits between species is also likely to mean that inadequate characterisation and evaluation is likely to prove an even greater barrier to CWR use in the future. Therefore in the next ten years there is a requirement for collection managers, whether the collections are held in situ or ex situ, to promote the use of their conserved germplasm to the user community by adopting the approaches outlined above or other novel technologies associated with pre-breeding. Even though it seems likely that the various technical challenges remaining in trait transfer between CWR and elite breeding material will be overcome in the next ten years or soon after, breeders are likely to still largely obtain the diversity they need from crop gene banks. This means that ex situ collections are likely to act as a conduit between in situ conserved diversity and the breeders, so the unification of the in situ and ex situ CWR conservationists must also be seamless if we are to give users what they will demand.


The central proposition of this paper is that climate change presents a new and rapidly developing threat to global food security. CWR contain the genetic diversity that can at least partially mitigate this threat, yet CWR themselves are in turn threatened. Conservation of CWR has not been immune to the ‘research-implementation gap’ between science and real world action that has been identified elsewhere for conservation assessment. This is a real dilemma that has to be addressed urgently. We already have the techniques and experience to conserve and use CWR diversity effectively for the benefit of humankind — the onus is now to strengthen the weak existing links between biodiversity and agrobiodiversity conservationists to act systematically between themselves and in concert with crop breeders to help guarantee the basis of global food security. It is very important, therefore, that governments build on recent knowledge and protocols and incorporate responsibility for maintaining and implementing the conservation of CWR into their national biodiversity and PGR systems.

Further suggestions on how the systematic conservation of CWR diversity might be achieved and how the conservation-utilisation link might be improved and expanded are summarised in Appendix 1. Although the points are extensive, they are not exhaustive. It is also likely that they are already being applied by many proactive genetic reserve and gene bank managers to help meet the germplasm users’ demands.


Synonymous terms include ‘genetic reserve management units’ (GRMUs), ‘gene management zones’ (GMZs), ‘gene sanctuaries’ or ‘genetic sanctuaries’ and ‘crop reservations’.


It should be noted that both multi-species and single species reserves have advantages and disadvantages and much more experience is needed to be able to draw conclusions as to which is preferable (see discussion in Heywood & Dulloo 2006: 40 – 41).


Copyright information

© The Board of Trustees of the Royal Botanic Gardens, Kew 2011