by GRAIN | 20 Jun 1996
Seedling - June 1996

June 1996



After many years of relying almost exclusively on ex situ conservation of plant genetic resources, the international community is at last beginning to face the shortcomings of genebank fridge conservation. The United Nations Food and Agriculture Organisation (FAO) has just prepared a draft Report on the State of the World's Plant Genetic Resources that analyses how we are doing in this field. The bulky report, which comes in three volumes, contains a wealth of information, statistics and assessments. This article briefly looks into the "highlights" of the report - especially of its ex situ section - and shows how some of the NGO's past fears are finally being confirmed by FAO.


Farming communities everywhere in the world have adapted local crops and plants to their needs. But it has been in the tropical and sub-tropical belts of the planet where the largest wild plant biodiversity is concentrated and thus where farmers have had access to a wider variety of material to meet their needs. As a result, the centres of origin and most of the diversity of the 30 crops that provide 95% of human dietary energy or protein are to be found in the South. The same is true for the 120 plant species important for food and agriculture at the national level and for the 7000 plants that have been cultivated or collected by humans for food at one time or another.

It is this treasure of farmer genius which has been under severe and increasing threat for several decades now. As the modernisation of agriculture and the Green Revolution crop varieties were promoted for one staple crop after the other, hundreds of farmers varieties of those crops were either relegated to marginal areas or just got lost. The response by the institutional sector and the international community was slow in coming, but did get underway in the 1960s and 1970s. But rather than addressing the causes of genetic erosion (by challenging the Green Revolution approach) scientists started to address the symptoms: they swarmed out to collect the remaining farmer varieties of major food crops and put them into genebanks. By doing this, the local cultivars and wild relatives could be saved for a long period while still being available to breeders. This approach is called ex situ conservation (see box). In situ conservation, on the other hand, means that plants are preserved in their natural habitat or - even better - on farmers' fields (also known as on-farm conservation).

What is the problem?

During recent decades, NGOs have frequently questioned the ex situ approach as the way to go about plant genetic resource conservation and use. Their concern focused on three main issues: the methodology as such, the political implications and the technical constraints.

The methodology: a conservation system that is based on genebanks relies on a crop-by-crop, fragmentary approach, which ignores the interactions of the different elements in the agro-ecosystem, as well as the role of farmers' knowledge. It cuts the genetic materials developed by farmers off from their evolutionary process and thus forms a static approach to conservation. As far as the use of the material is concerned, it is mainly directed to official plant breeders, with farmers merely being on the receiving end of the `improved' varieties. In a sense, it is a `band-aid' approach to conservation, that does not provide solutions for the causes of genetic erosion.

Political implications: the main issues raised are the lack of control of accessions by those who provided the material in the first place (local communities or countries). During much of the 1970s and 1980s, the International Board on Plant Genetic Resources (IBPGR) — an institute of the Consultative Group on International Agricultural Research (CGIAR) — promoted a genebank system that effectively channelled genetic resources from the South to genebanks in the North, or to the genebanks of the CGIAR itself. Other political biases were the predominance of the main commodity crops in the genebanks (and the lack of minor crops and wild species important to subsistence farmers); the almost absolute focus on servicing breeders and the biotechnology industry; and a constant trend towards the privatisation of plant genetic resources. Some of these issues are being addressed in the current negotiations under the Convention on Biological Diversity and the International Undertaking on Plant Genetic Resources.

How Ex situ conservation works

Genebanks are compartmentalised cold storages in which seeds are theoretically kept in controlled conditions of temperature and humidity. The banks work on the principle that dehydrated seeds are capable of remaining viable for long periods of time in cold conditions. Seeds may be kept in long-term storage (from 0 to -18ºC), medium-term storage (0 to 10ºC), or short-term storage (more than 10ºC). Orthodox seeds (those that can be dehydrated) account for most of crops, including all major cereals. But so-called `recalcitrant' plants have seeds that do not stand up to such a process, such as coconut, avocado, mango and tea. The ex situ conservation of these crops and vegetatively propagated plants relies largely on field banks. Recently, in vitro conservation techniques have also been used for the conservation of recalcitrant crops. Wild crop relatives and large species (such as trees) are often conserved in botanical gardens.

For ex situ conservation a number of seeds have to be collected, because even under the most stringent long term conservation conditions seeds eventually loose viability and die. Accessions have to be regenerated regularly by planting them out in order to obtain new seed. However, the process of regeneration also results in a loss of genetic diversity, especially if it is done in conditions different from those of the site of origin of the accession. This is because the new environment may result in a different kind of selective pressure.

Genebanks may hold different types of collections. Active and working collections allow easy access to breeders: they are raw materials used for the development of new varieties. However, the ability of any genebank to conserve agricultural biodiversity for current and future generations, and the trust it may be given, depends on how it manages its base collections. Base collections can only be maintained under long- to medium-term storage conditions.

But even the best conserved accession will be of limited use unless basic information on it is available. Information on accessions is classified in three categories:

- Passport: this includes basic data such as the sampling date and site.

- Characterisation: this includes data on taxonomic, environment-independent properties that describe the variety.

- Evaluation: it includes data on agronomic properties of the accession, which are normally closely related to the environment.


Technical questions: Many samples in genebanks are easy-to-access-from-the-road collections and do not represent the full spectrum of what farmers developed. The long-term security of genebanks for conservation is questioned. Problems with power supply and consequent temperature fluctuations, seed desiccation, and poor regeneration are frequent. NGOs have suggested that genetic erosion in the genebanks may be as high as it is in the field.

The State of the World Report

This year FAO published the first (draft) Report on the State of the World's Plant Genetic Resources for Food and Agriculture. The Report is the result of an unprecedented effort to review and assess the state of Plant Genetic Resources in agriculture. It tries also to identify the causes of their erosion and to analyse the areas where action can be taken in order to conserve genetic resources. The Report identifies the intensification of agriculture as the leading cause of genetic erosion. It formally recognises the role of resource poor farming communities in the conservation and development of biodiversity.

The Report, based on 151 country reports, 12 sub-regional meetings and FAO's World Information and Early Warning System (WIEWS) database, contains a lot of interesting information. It also confirms some of the fears NGOs and farmers have been expressing for a long time. Besides the political issue of who has the control over the accessions, the Report talks about problems related to the security of genebanks, available information on the accessions, the danger of deterioration due to lack of regeneration, and the use of existing accessions.

Ex situ today

By the end of the 1970s, there were 54 seed keeping facilities, of which 24 had long-term storage capabilities. After a widespread but largely uncoordinated effort by both countries and the CGIAR in the late 1970s and early 1980s to build up a genebank system and to collect existing agricultural biodiversity, the world now has a total of 1,308 national, regional and international germplasm collections, according to the FAO Report. Of these, 397 are maintained under long- or medium-term storage conditions, most of the rest being active working collections for use by researchers and plant breeders. For the purpose of conservation, the base collections — held under long- or medium-term storage conditions — are the most important ones, as these supposedly contain the unique material for use sometime in the future.

According to FAO, there are a total of about 6.1 million accessions under ex situ conditions: 600,000 in the CGIAR system and 5.5 million in national collections and regional genebanks. Some 50% to 65% of the accessions (3-4 million) are in base collections. Ninety per cent are kept under cold storage, while 527,000 are kept in field banks and 37,000 in vitro.

More than 80% of the accessions are held in national banks, and 45% in just twelve countries (Brazil, Canada, China, France, Germany, India, Japan, Korean Republic, Russia, Ukraine, UK and USA).

High tech, high problems.

All these figures might seem overwhelming, and give a sense of security that most of the diversity is stored safely away. FAO itself is quick in denouncing this false sense of security that the big numbers might suggest. A lot of the materials is quickly losing its viability and there is a lot of uncoordinated duplication. Some countries have expensive excess storage space and find it hard to keep up with the growing maintenance costs, while other countries lack proper long term facilities.

In the process leading to the Leipzig conference all but six participating states reported to have ex situ conservation facilities. Some 75 countries have seed storage facilities suitable to medium/long-term storage, but only 35 countries have secure long-term seed storage facilities, if measured according to internationally accepted criteria. To these `secure' national collections, one would have to add 9 of the CGIAR genebanks and 4 regional genebanks which are in decent state. The FAO report points out with concern that one of the major genebanks in the world — the Vavilov Institute (VIR) in Russia, which holds the world's third largest wheat collection and important collections of other crops (including barley, rice and corn, cotton, lentil and sugarbeet) -- does not meet these criteria. The VIR has only medium-term storage facilities. The need for frequent regeneration is a strong threat to the diversity in its accessions.

Technical genebank problems reported by countries include:

- Equipment problems, particularly in cooling units (reported by a number of African countries, Vietnam and Romania), lack of seed cleaning and humidity control equipment.

- Irregularities in the supply of electricity (reported by a number of African countries, Cuba, Bangladesh, Iraq and Turkey).

- Difficulties in seed drying, especially in the humid regions of Africa (where the problem is most pronounced) Asia and Latin America.

In addition, the FAO notes that in many of even the few `secure' facilities "full safety duplication is lacking, and there is often a large backlog of accessions for regeneration". So the impressive figure of 1308 collections of 6.1 million samples ends up boiling down to perhaps a dozen or two truly secure, safe and up to date conservation facilities.

Also, a cursory glance at the data shown in Figure 1 suggests that the genetic booty is spread fairly evenly across the world: 50.4% of all accessions are in the North, 38% in Southern collections and 11.6% in CGIAR genebanks, located in the South. However, both the Southern origin of most genetic diversity and continued Northern control over the CGIAR bias these figures to the North's benefit.

The security of the system

In addition to the above mentioned technical problems with the safety of the collections, there are two additional ones: lack of regeneration, and chaos in duplication.

One of the main problems genebanks around the world are facing is regeneration. In order to stay alive, seeds stored in genebanks need to be grown out and harvested — regenerated — once in a while. The frequency of this regeneration depends on the crop. As the FAO report states, if a genebank had to regenerate its collection once every 10 years, one would expect routine regeneration needs to be 10% annually. However, the FAO report finds the reality quite different: `some 95% of the countries report a far higher level of need'. Of the 95 countries providing information about regeneration activities, at least 71 (together holding nearly three million accessions) "experience some difficulties in regenerating their collections".

FAO concludes that almost half (48%) of all stored seeds worldwide now need to be regenerated. But its report also warns that some of these "may already have lost their viability or genetic integrity, or they may be from populations where re-collecting may prove more cost-effective than regeneration". However, for many of the accessions, re-collecting may prove impossible, because of extinction in the field.

In addition, many countries do not have either the funds, facilities or staff necessary to conduct their needed regeneration activities. Although countries in the South are most affected by this backlog, both the CGIAR genebanks and some countries in the North (such as the USA and Japan) are also affected. Part of the problem comes from the fact that when the genebank system was set up, nobody really took into account the needs and costs of the long-term maintenance of the accessions. One notorious example has been the spread of expensive genebanks in some Asian countries — built with Japanese aid money — several of which are now virtually empty or simply not functioning. The global picture, according to the FAO Plan of Action, is "a steady deterioration of many facilities and their ability to perform even basic conservation functions". This is serious language, especially if it comes from an agency that is coordinating a Global System on Plant Genetic Resources, which takes the ex situ genebank approach as the starting point.

Duplication of the existing collections is both a problem and a need. The FAO report notes that there is a lot of uncoordinated and unknown duplication amongst the world's stored seeds. One study published in 1987 estimated that only 35% of the stored seeds were unique, the rest being duplicates. This study was based on the then 2.5 million accessions, and the FAO notes that with the steep rise the amount of stored seeds worldwide since then, "it must be assumed that inadverted duplication is now even higher". It states that this over-duplication is a waste of money and should be minimised.

But on the other hand duplication of unique accessions and their storage in other genebanks is crucial to ensure their security in the face of unexpected losses (because of fire, earthquakes, war, etc.). Also in this area, the results of the FAO surveys are alarming: only half of the countries provided information on their duplication effort (probably meaning that they do not have a structural approach to duplication). Of the other half that did respond, only 11 countries indicated that their collections (430.000 accessions) were fully duplicated somewhere, 51 countries reported partial duplication, and 10 countries reported no duplication at all. Even the IARC genebanks, with their high tech resources and funding, report high levels of duplication lacking in several crops.

The security and duplication problems in the current genebank system lead to a situation where genetic erosion in the banks might be higher than that in the field. For example, FAO reports that a review of the USA genebanks (based on tests conducted between 1979 and 1989) showed that 29% of the US collection had seed germination rates that were either unknown (21%) or less than 65% (8%). Furthermore, 45% of the accessions had less than 550 seeds. In layman's terms this means that an important part of the USA collection (one of the most sophisticated in the world) was either dead or dying, or might have too few seeds to guarantee survival of genetic diversity. Another example is provided by the Latin American Maize Programme (LAMP), where in 1991 only half of the accessions could be evaluated due to lack of viable seed. Also, the lack of reliable storage facilities had resulted in the total loss of a large number of accessions and severe genotype deletions in many more.

What is being conserved?

Figure 2 shows the representation of crops in genebanks. The high proportion of cereals (40%) reflects not only their key role in global food security, but also to Northern agriculture and to international agricultural trade. In contrast, tubers and roots, which include important crops for subsistence farmers like sweet potato, potato and cassava, only account for 4% of global accessions. Bananas and plantains are represented by a mere 10,500 accessions. Many locally-valuable crops have only just started to be collected, and regional base collections are being kept by national genebanks. FAO admits that there are very few accessions of medicinal plants. The information on the type of accession (whether wild relatives, local varieties or breeding lines) is only available for one third of global accessions. Where this information is known, the proportion of each kind of accession varies according to the crops and to the genebank system, with IARCs focusing on landraces and the private sector on advanced breeding lines. Globally, 48% of accessions are advanced cultivars or breeding lines, 36% are old cultivars or farmers' varieties and only15% are wild or weedy plants or crop relatives.

Again, the picture is a biased one. When the early plant hunters went out collecting, they obviously focused on the ones they were interested in, something still reflected in genebank holdings. Grossly lacking are crops that do not enter international trade and are `only' important at national or local levels. Also, species providing wild foods and other products important to local livelihood systems are notably absent. As far as the type of material is concerned, almost half of the stored seeds worldwide consist of breeders' germplasm, while only just over one third are the original farmer varieties. This poses a fundamental question on what the current genebank system is doing: storing material chewed out by the commercial breeders, or saving the original diversity developed by farmers? It is also is another bias which points more to the needs of commercial plant breeders than to those of small farmers.

Information and use

The information on the accessions held ex situ may be almost as important as the accessions themselves. This information may aid rapid identification of the required accessions and/or characteristics and may help to prevent excessive duplication of efforts, while showing where the gaps are. From a political and equity perspective, information on the origin of the accessions is also relevant. Ease of access to the information may make it easier for both breeders (as is the case) and also farmers (as it should also be) to use the materials.

Yet news from the Report on what we know about the stored seeds is, again, not good:

- Although 37% of national collections and nearly all the CGIAR genebanks accessions have passport data, in most collections these data refer only to the country of origin. Plant breeders often develop their own collections because of the lack of information on collections in the genebanks.

- The rates of characterisation and evaluation are also very low. As a result, the accessions of national collections are not fully utilised even by current genebank clients: breeders. The only exceptions to poor characterisation seem to be most countries in Europe, East Asia, North America, Ethiopia, India and the Philippines.

- While some genebanks have their collections fully documented, computerised and even put in the Internet (as it is the case of the Vavilov Institute and the USA base collections), others have not documented any of their accessions.

These factors clearly limit the use of the accessions stored in the genebanks. Nevertheless, a large number of accessions are exchanged around the world. For example, over the last three years the CGIAR centres have distributed an annual average of over 120,000 accessions to national programmes all over the world (see box). And in that time the USA has distributed 116,897 samples to 126 countries. However, these figures do not necessarily show the net NorthSouth flows. For example, in its country report Bangladesh states that the outgoing germplasm packages (to USA, Japan, West Africa, India and IARCs) "have surpassed remarkably over those that have been brought inside the country". Of the over 8000 samples the country shipped out, 4000 were rice varieties sent to IRRI (the International Rice Research Institute, based in the Philippines). In turn, IRRI contributed with only 150 rice varieties to the Bangladesh collection. Similarly, Japan received from Bangladesh 400 rice samples, while only sending in 150 samples.

IARCs accessions distribution

The CGIAR, which often proudly describes the contribution of the germplasm its sends out to developing countries, in fact mostly exchanges seeds with itself, or with the host countries of the individual IARCs: between 1992 and 1994, 55% of the accessions sent out went to sister institutes, 34% to developing countries, 9% to developed countries and 2% to the private sector. But, at a closer look, much of the distribution to developing countries is done to the IARCs' host countries (one third of total distribution between 1987 and 1991). And, although, only 11% of CGIAR germplasm goes to the North or to the private sector, this is a major contribution. FAO reports that estimates of the value of CIMMYT-based wheat germplasm to agriculture in OECD countries ranges from US$300 million to US$11,000 million per year!


Towards a useful genebank system?

After its review of the ex situ conservation around the world, FAO makes some proposals to improve the current genebank system. This should be done mainly through rationalisation, regional and international collaboration and the filling of the existing gaps, both in the plants covered and in the information on them. These ideas have been further developed in the Global Plan of Action, in the form of priority activities.

There is no doubt that ex situ conservation is needed and that, if the existing collections are to be saved, an improvement of the security conditions and the efficiency in the management of genebanks is needed, especially since a large part of the diversity they contain no longer exists in the field.

However, the unavoidable erosion of genetic diversity in the genebanks in the long term, even under the best long-term storage conditions, questions the appropriateness of this technology as a primary conservation tool for agricultural biodiversity. Diversity originated as the result of continuous interaction of plants, animals, environmental conditions and human selection. Storage of the seeds away from the fields and forests breaks this critical cycle.

A conservation system that leaves aside farmers, peasants, pastoralists and their knowledge, and expropriates their materials — making them available as raw materials to breeders and biotechnologists that may claim intellectual property rights over their `creations' — is doubly perverse. Not only do farmers lose control over their heritage -- the seeds -- but they also loose the possibility of continuing with their sustainable and highly productive agricultural practises which build and build upon diversity.

Contrary to what had been assumed by most of the international scientific community, ex situ conservation on its own cannot ensure the conservation of precious plant genetic resources for future generations. The primary conservation effort has to focus on farmers and their fields. It should go hand in hand with actually making use of that diversity in the complex agro-ecosystems that many of these farmers manage. Genebanks, in the context of well defined national genetic resources programs, can and should play an important role in facilitating the movement of germplasm between farming communities and safety backup sites, as well as from these sites back to the farming communities. A genebank as a dynamic partner in not only conservation, but also development of genetic resources — in which not only official plant breeders but also local farming communities can participate — is still a far away dream in the FAO documents. Until FAO and the international community shifts towards this perspective, NGOs and peasant organisations have to continue to do a lot of work to conserve and improve their biological resources and convince policy makers at all levels of the value of their approaches.



* CGIAR (1994), Stripe Study of Genetic Resources in the CGIAR, The Consultative Group on International Agricultural Research Technical Advisory Committee. AGR/TAC:IAR/94/2.1

* FAO (1994), Survey of Existing Data on Ex situ collections of Plant Genetic Resources for Food and Agriculture, Commission on Plant Genetic Resources, FAO, Rome CPGR-Ex1/94/5 Annex September 1994.

* FAO (1996) Report on the State of the World's Plant Genetic Resources, Commission on Genetic Resources for Food and Agriculture, CGRFA-EX-2/96/2

* FAO (1996) Draft Report on the State of the World's Plant Genetic Resources (Full Background Documentation).

* Pat Mooney, Cary Fowler (1990) Shattering; Food, Politics, and the Loss of Genetic Diversity, The University of Arizona Press, Tucson, Arizona.

Author: GRAIN