Development Policy Review, 2001, 19 (2): 181-204

Unequal Exchange? Recent Transfers of Agricultural Resources and their Implications for Developing Countries

Cary Fowler, Melinda Smale, and Samy Gaiji ∗

Plant genetic resources constitute the biological basis for plant breeding and future agricultural development. Their transfer from developing to developed countries over centuries has sometimes been viewed as an example of exploitation, if not ‘biopiracy’. Modern gene flows are different in character and magnitude from historic exchanges, however. This article examines current patterns and finds that developing countries are major net recipients of germplasm samples from CGIAR centres, particularly if ‘improved materials’ are considered. Potentially problematic, intellectual property rights do not currently present major barriers to the availability and use of genetic resources by developing countries. Proposals to restrict flows and redress perceived injustices may reduce the benefits accruing at present to developing countries from germplasm exchanges.

Introduction

It is well known that most major agricultural were domesticated over a period of thousands of years in what are now termed ‘developing’ countries (de Candolle, 1886; Vavilov, 1926; Harlan, 1975; Simmonds, 1976). Historically, the greatest concentration of genetic diversity has consequently been found in these countries (Vavilov, 1926; Zeven and Zhukovsky, 1975). The ‘flow’ of genetic resources from these developing countries to Europe and North America – a process which took place over hundreds, and in some cases, thousands of years – unquestionably provided much of the early biological foundation for in today’s developed countries, and a platform for development and expansion (Fowler, 1994). The implications of 1492 for development, including the massive botanical transfers that took place afterwards, have been the subject of a number of histories (Crosby, 1972, 1986). Commentaries have tended to focus on the economic and developmental advantages of early transfers to ‘gene-poor’ Europe and North America, and the lack of benefits which accrued to the Latin American, Asian and African ‘donors’ of this genetic material (Mooney, 1983). Little attention has

∗ Respectively, Honorary Fellow/Senior Adviser to the Director-General, International Plant Genetic Resources Institute (IPGRI) and Associate Professor, Centre for International Environment and Development Studies, Agricultural University of Norway; economist at IPGRI and International and Improvement Centre (CIMMYT) and Visiting Research Fellow, International Food Policy Research Institute (IFPRI); and Project Leader, System-wide Information Network for Genetic Resources, IPGRI.

 Overseas Development Institute, 2001. Published by Blackwell Publishers, Oxford OX4 1JF, UK and 350 Main Street, Malden, MA 02148, USA. 182 Cary Fowler, Melinda Smale and Samy Gaiji been paid to the magnitude and direction of recent flows of genetic resources and their implications. Rather more attention is currently focused on what has been termed ‘biopiracy’, and charges that intellectual property rights are being used to siphon off materials previously considered to be ‘public’ (Shiva 1997, 1999). The impact of flows of genetic materials and the benefits associated with such transfers are inextricably linked with recipients’ capacity to use the materials (de Souza Silva, 1989). Legal restrictions and funding constraints, as well as the condition of the physical capital and research infrastructure, may impede use. Nevertheless, despite the explosion of public concern,1 most attention has been focused on individual cases, with relatively little analysis of the different impacts that different forms of intellectual property rights might have on various kinds of biological materials in the public domain. With the rise of commercial agriculture and the rediscovery of Mendel’s laws of heredity in 1900, modern ‘scientific’ plant breeding began to take hold, and conditions began to emerge for establishing legal protection for the products of this particular form of innovative endeavour (Fowler, 2000). Intra-species diversity provides the raw materials which breeders use to select, recombine and fashion new crop varieties. As crops co-evolve with pests and diseases, continued evolution and ‘improvement’ are necessary for any stable and prosperous agricultural system. Plant genetic resources for food and agriculture (PGRFA) are thus the biological foundation of this effort (Hoisington et al., 1999). The insights of Mendel, together with the ‘modernisation’ of agriculture and the development of markets for crop varieties with particular qualities, combined to shift the focus of attention and interest from the species level to the variety level. The new biotechnologies have prompted a further shift to the gene level. For the most part, interest in acquiring or transferring crop species no longer exists. Public and private sector interests are related to materials at the variety and gene level necessary for breeding new varieties of crops.

Interdependence and genetic resources

Like developed countries, developing countries have also come to rely heavily on their own production of non-indigenous crops (from other developing country regions) to meet food needs. This ‘dependency’, which evolved over a long period of time in many cases, implies that both developed and developing countries rely on imported germplasm.2 At present, levels of dependence (measured in terms of caloric contribution to nutrition contributed by crops whose ‘centre of diversity’ is outside the country in question) are not markedly different between developed and developing countries,3 a situation that has not been fully appreciated in recent international negotiations over access to biological diversity or reflected in proposals for national legislation

1. A search for the recently coined term, ‘biopiracy’, through one internet search engine (www.hotbot.com), yielded 3,600 citations. 2. ‘Germplasm’ refers to seeds, plants or plant parts that are useful in crop breeding, research, or conservation because of their genetic attributes. 3. This article does not examine the important and related issue of the extent to which developing countries have the capacity to develop these genetic resources. Unequal Exchange? Recent Transfers of Agricultural Resources 183

(OAU/STRC, 1998). Countries in southern , for example, fall between 65% and 100% in their dependence on main food crops that originated outside the region, with most countries exceeding a 90% dependency level (Palacios, 1998).4 Since countries in this region lack both large ex situ (genebank) collections as well as a broad base of modern varieties of their major food crops, future agricultural development will clearly require secure access to the germplasm of non-indigenous crops such as maize (centre of diversity: Central America), (South America), wheat (Near East), (Indochina), (Central and South America), plantain and (Indochina), and (South America).5 Africa is not unique, however, in having a food system based on crops with foreign origins. Virtually every region finds itself in a similar situation (FAO, 1998a). The question facing both developed and developing countries today is whether to encourage the flow and use of genetic resources through a multilateral system with a single set of rules agreed by most or all countries and codified in a treaty, or to restrain their exchange as would inevitably occur were access to be handled through myriad bilateral agreements, with each country making separate rules and procedures for access. Under such a bilateral system, those wishing to gain access to all the existing diversity of rice would thus be required to negotiate access for genetic resources separately with each of the more than 100 countries that possess such resources. Alternatively, countries could gain access to the materials through a single genebank, as would probably be the case were genetic resources held by centres of the Consultative Group on International Agricultural Research (CGIAR) placed within the scope of an agreed multilateral system. Some commentators have pointed to the historic transfer of genetic resources from South to North as evidence of ‘biopiracy’, as noted above, implying that the North’s gain has come from the South’s loss (Shiva, 1990). Some argue that developing countries might, even today, be better served through bilateral systems of access with strict conditions established by each individual country and/or a multilateral system of restricted scope, i.e., one applying only to a few crops. Greater control over access would, they contend, ensure greater benefits to developing countries, which are seen as rich in genetic resources of agricultural crops, despite the greater costs which might be associated with the acquisition (or failure to access) materials needed by the developing countries themselves. It is tempting, and common, to acknowledge no distinction between the historical transfers of crops (and the economic and developmental effects of the transfers) and current germplasm flows and their effects. Likewise, it is common to assume that all

4. Palacios measured the degree to which an individual country’s food energy supply (measured in calories) was dependent on crops whose primary area of genetic diversity was found outside that particular country. 5. It is also interesting to note that countries that are not dependent on non-indigenous crops are typically countries that are not normally considered well-endowed in terms of PGRFA of interest to the international community. Niger, for instance, is between 13 and 27% dependent according to Palacios (1998). However, it might not be able to take advantage of this lack of dependency, because it would still have few marketable genetic resources of interest to other countries. It should also be noted that Niger is the least dependent African nation. The median level of minimum dependency on non-indigenous major crops is 73%. In terms of maximum dependency, as defined by Palacios, , Comoros, Djibouti, , Madagascar, Malawi, Mauritius, Seychelles, South Africa, Swaziland, and Zambia all fall above 97%. Ethiopia, the country generally considered the richest in PGRFA in Africa, is estimated to be between 28 and 56% dependent. 184 Cary Fowler, Melinda Smale and Samy Gaiji forms of intellectual property rights have a similar and deleterious effect on the use and availability of public materials. Recent flows are generally directed to crop improvement, while historical transfers, in many cases, aimed at crop introduction. While aware of historical patterns, this article examines the nature and extent of recent flows of germplasm, which we contend are more indicative of current needs and interests. It also looks at the likely effect of patents and breeders’ rights forms of intellectual property and how they affect the use of the genetic material being transferred. By elucidating present relationships and interests, and by specifying the ways in which intellectual property rights are pertinent, it is hoped that this article will contribute to more informed and rational policy-making in regard to access and benefit-sharing from PGRFA.

Ongoing policy debates

An examination of germplasm flows as well as intellectual property rights is timely, given the current negotiations at the Food and Agriculture Organisation of the United Nations over the possible establishment of a multilateral system to govern access to and benefit-sharing from PGRFA. It is also timely in regard to ongoing discussions by the Conference of Parties to the Convention on Biological Diversity concerning access and benefit-sharing, as well as the more general public discussion of ‘biopiracy’. Policy-makers and negotiators in the various fora might appropriately ask a number of straightforward questions. Whose interests would be served by a multilateral system facilitating access to genetic resources, and whose, if any, would be served by more restrictions being placed on access by the current holders/owners of this germplasm? Is the flow of germplasm still from developing to developed countries? Do developing countries ‘contribute’ more germplasm than they receive? Would they benefit more by regulating access and trading it for agreed benefits on a sample-by-sample basis? To what extent is the public domain an empty shell, robbed of its character and usefulness by privatisation via intellectual property rights? In summary, what can data on the magnitude and direction of germplasm flows, measured in seed samples, tell us about the advisability of organising future transfers through either (i) a multilateral system such as that under discussion at FAO, or (ii) a Convention on Biological Diversity-styled bilateral system which would be likely to emerge by default if the FAO negotiations fail? And what can a dispassionate examination of current intellectual property practices tell us about the future availability of genetic resources once considered to be the ‘common heritage of mankind’?

Types and sources of data

No complete record of historical or even recent germplasm transfers exists. This study takes advantage of all data currently available publicly on genebank transfers, as well as data acquired and analysed specially for the study. Various institutions are now in the process of compiling additional data from their records, thus more will be available in the future. Data for this study were available through the SINGER database6 from the

6. SINGER, the System-wide Information Network for Genetic Resources, is the genetic resources information exchange network of the International Agricultural Research Centres of the CGIAR. It Unequal Exchange? Recent Transfers of Agricultural Resources 185

International Rice Research Institute (IRRI), the International Crops Research Institute for the Semi-Arid Tropics (ICRISAT), and the International Centre for Agricultural Research in Dry Areas (ICARDA).7 The International Maize and Wheat Improvement Centre (CIMMYT) provided some summary data on genebank transfers, newly assembled raw data on distributions from nurseries, and information on CIMMYT contributions to variety releases in the developing world. In addition, data on germplasm exchanges from 1972 to 1990/1 between fifteen developing countries and the centres of the Consultative Group on International Agricultural Research (CGIAR) plus the Asian Vegetable Research and Development Centre (AVRDC) were also available.8 Thus, in some cases we can look at flows into and out of a selected group of countries over a 19-year period – an approach that can provide some indication of what countries have provided in relation to what they have received from genebanks. In other cases, with the available data we can examine flows to virtually all countries, for a particular crop. An additional source of information for this study was the FAO’s State of the World’s Plant Genetic Resources for Food and Agriculture, a volume based on 154 country reports and published in 1998. In the current political and negotiating context, an examination of germplasm transfers into and out of the research centres of the CGIAR is particularly relevant. The CGIAR is the largest consortium of (mainly) crop-oriented research facilities in the world. It concentrates on the major crops of importance to world . The centres are part of the ‘public sector’, and the germplasm they hold is firmly part of the ‘public domain’ (FAO/CGIAR, 1994). In addition, they are among the most active collectors and largest suppliers of germplasm in the world.

Sources of supply of genetic resources

Conscious and concerted efforts to collect and conserve genetic resources date back at least to the 1920s. Harlan and Martini first called attention to the loss of potentially valuable plant genetic resources in 1936. By the late 1950s genebanks were established in a number of countries, most notably the United States and the former Soviet Union (FSU) (Fowler and Mooney, 1990). Only several hundred thousand accessions (samples) of seed are thought to have existed in conservation programmes in some 54 genebanks in the early 1970s (FAO, 1998a). The number now exceeds 6 million, as shown in Table 1, and there are more than 1,300 genebanks. The CGIAR holds over 500,000 accessions ‘in trust’ for the international community under the auspices of FAO.

provides access to information on the collections of genetic resources held by the CGIAR centres. Together, these collections comprise over half a million samples of crop, forage and tree germplasm of major importance for food and agriculture. SINGER can be accessed through the internet at: www.singer..org . The present authors express their appreciation to the staff of the System-wide Genetic Resources Programme for their assistance in providing data through SINGER. 7. For the collections at IRRI, data were available for 1985-97; for ICARDA, for 1990-97; for ICRISAT, for 1974-97; and for CIMMYT, some summary information was available for 1987-97. 8. The term ‘international agricultural research centres’ (IARCs) will be used when reference is made to data that include the AVRDC, a centre that is not a part of the CGIAR. 186 Cary Fowler, Melinda Smale and Samy Gaiji

Table 1: Genebanks and accessions in ex situ collections

Region /Institution No. of genebanks No. of accessions Africa 124 353,523 Latin America & Caribbean 227 642,405 North America 101 762,061 Asia 293 1,533,979 Europe 496 1,934,574 Near East 67 327,963 CGIAR 12 593,191 TOTAL 1,320 6,147,696 Source: FAO (1998a).

As indicated by the dramatic increase in accessions, most materials stored in genebanks today were collected during the 1970s and 1980s. The greatest outflow of materials from developing countries took place during this period of intense collecting. Significantly, this is the period for which we have the best and most complete data.

Importance of genebanks

According to FAO and other observers, a large portion of the genepool of major crops has been collected and placed in genebanks (FAO, 1998a; Jackson et al., 1997).9 , 10 or ’ varieties, make up a substantially larger portion of CGIAR collections (59%) than they do in governmental (12%) or private (9%) collections (FAO, 1998a). Thus, much of the diversity that can be found in farmers’ fields today, and a great amount of diversity that no longer exists on , can be accessed through genebanks,11 particularly those of the CGIAR. Even if genebank collections do not contain the total diversity of a particular crop (some areas or plant types will always remain to be sampled), plant breeders strongly prefer to access materials from genebanks rather than through collecting expeditions. It is easier, quicker and less costly, and the materials generally come with useful associated information (characterisation and evaluation data) (ten Kate and Laird, 1999). Searches for the genetic diversity needed will begin with a look in their own working collections, followed, perhaps, by database searches for other modern and breeding lines. Use of such material is preferred, because fewer deleterious traits

9. No precise data exist, nor is it ever likely that they will. Due to the nature of this biological material and the fact that it is in a dynamic and evolutionary state, any percentage provided of the portion of existing diversity which has been collected would necessarily be an estimate, or a guess. Nevertheless, as FAO has documented, there is considerable agreement that genebank collections of major crops contain a significant amount of the total diversity that exists within those crops. 10. Landraces are traditional or farmers’ varieties which are the product of breeding or selection carried out by farmers and communities, usually over many years. Unlike commercial cultivars, landraces tend not to be genetically uniform. They often exhibit high levels of genetic diversity. 11. Genebanks are facilities for the storage and conservation of genetic resources. In crops such as rice, wheat and maize, seeds are kept under conditions of very low temperature and humidity. Properly conserved, such samples will remain viable and available for use for decades or longer. See Ford-Lloyd and Jackson (1986). Unequal Exchange? Recent Transfers of Agricultural Resources 187 have to be eliminated in order to get the desired genetic material into the new variety. If the desired material cannot be found in other cultivars, then the breeder might turn to older varieties or landraces in genebank collections. If this fails, a collecting expedition might be organised if it seems likely that the material needed can be found. But given the size of current genebank collections and the lack of documentation about the characteristics of unsampled farmers’ materials, such expeditions are rare. More often, expeditions are organised for more general purposes, such as completing the sampling of a crop in a given region for long-term genebank conservation. Although genetic resources obviously did not ‘originate’ in genebanks, the main source of genetic resources today is ex situ, despite the fact that impressive diversity still exists in fields and gardens around the world, where it continues to evolve and be developed (Berg et al., 1991; Brush, 1999). The genebank has become a modern-day ‘centre of diversity’. The principal suppliers of genetic resources are no longer traditional farmers but genebank directors and breeding programme managers. This is dramatically illustrated by the 60 to 1 ratio of samples distributed by the CGIAR genebanks to developing countries, compared with accessions received in 1992, not an atypical year.12 In recent years, many national and international programmes have reduced collecting efforts. According to FAO:

The decrease in international collecting activities has developed in parallel with a general improvement in the conservation infrastructure at the national level. With the development of national plant genetic resources programmes and of national conservation facilities, the collecting of germplasm is now becoming more of a country-driven process based on national priorities (FAO, 1998a).

Today, wheat genetic resources, for example, are more likely to be acquired from a genebank in the CGIAR system (e.g., CIMMYT) or a national genebank in the US or Russia, than from a country in the original centre of origin or diversity. will be obtained from the CGIAR system (e.g., CIP), the US or Japan, holders of the three largest collections. Countries in what is known as ‘Vavilov Centres of Origin’ are no longer the principal suppliers of such materials. In most cases, they ceased filling this role decades ago. In addition, while most major agricultural crops were domesticated in what are now considered developing countries, many of these countries were never a ‘centre of diversity’ for any/many important crops. A large number (like most developed countries) have had comparatively little to offer historically from either in situ or ex situ sources.13 They have been ‘importers’ more than ‘exporters’ of plant genetic resources.

12.The 60:1 ratio may, in fact, be an underestimate. A single sample sent to a national genebank might then be acquired by different users within the same country. In the data presented here, it counts as only one incident of access. Furthermore, the data do not account for flows of improved materials (e.g., breeding lines, cultivars) which are provided by the research centres to developing countries. 13. Many developing countries fall outside all Vavilov centres. In other cases, countries may be a historical ‘centre’ for only one or two crops. CGIAR data indicate that a number of countries have supplied relatively few accessions currently stored in CGIAR genebanks. In other words, not all developing countries are well endowed in terms of intra-species crop diversity. 188 Cary Fowler, Melinda Smale and Samy Gaiji

The original Vavilov Centres, for example, encompassed parts of only three African countries: Algeria, Tunisia and Ethiopia. In this section we have seen that a considerable portion of existing crop diversity has been collected and stored in genebanks. Breeders will typically turn to well organised and documented genebank collections before going to the field to search for the same genetic material. Whereas farmers’ fields in developing countries were the chief source of unique breeding materials historically, the genebank plays that role today.

Germplasm flows between selected countries and international centre genebanks

In this section we examine the flow of genetic material between specific developing countries and international agricultural research centres (IARCs) and document how much material those countries supplied to, and how much they received from, the centres over an 18-19-year period.

Assessing flows of accessions and samples

Much (though not all) of the material (accessions) provided by developing countries to genebanks in the past would be considered ‘unique’. In contrast, multiple samples of a single accession might be provided by a genebank to a country over time. The data do not allow for a comparison of flows of ‘unique’ accessions, much less of ‘genetic diversity’. Few countries, however, are interested in acquiring diversity per se.14 In this article, requests for a sample from a genebank are considered to be indicative of demand. Multiple requests for the same sample are no less indicative of need. Indeed, it could be argued that such requests represent attempts to acquire ‘new’ diversity, in the sense that previous samples were not retained or were lost. The value of the material in a breeding programme is thus not diminished, according to whether the particular accession is being acquired as a germplasm sample for the second time. On the contrary, this would indicate its usefulness. Thus, we have not attempted to construct a balance sheet of exchanges of unique accessions, an impossible task of dubious utility. We have chosen, instead, to compare flows of germplasm samples, as illustrative of need and benefit.15

14. If countries mainly acquired materials from CGIAR genebanks in order to build up collections in their own genebanks, then we would be more interested in monitoring flows of unique accessions. We assume, however, that flows to developing countries are primarily flows into breeding programmes, i.e., are flows aimed at immediate practical ends. Were countries more interested in expanding their genebank collections, we would expect to see even larger numbers of requests for unique accessions than the data indicate that there are. 15. Were it possible to monitor flows of unique alleles, or unique accessions, we would speculate that developing countries would be net providers as a group. It is important to note, however, that developing countries do not access materials ‘as a group’. They do so individually. Therefore, while developing countries, in total, might be net suppliers, it is equally likely that many, perhaps nearly all, would be net recipients on an individual basis. In assessing the benefits of open versus restricted access regimes, countries might be expected to be more concerned with the benefits that accrue to themselves individually. Unequal Exchange? Recent Transfers of Agricultural Resources 189

It is also possible (but cannot be proven with the data at present available) that individual countries are also net recipients of unique germplasm. For example, Zambia may provide maize accessions collected throughout the country to an international centre such as CIMMYT. In return, it may receive materials from Central and South America, as well as other African countries – materials representing a broader range of genetic diversity than could be found among the more limited, albeit still diverse, Zambian maize materials.

Genetic transfers during the peak period of collecting

In 1994, the International Fund for Agricultural Research (IFAR) produced a series of case studies on the role of IARCs in particular developing countries. The studies were prepared by Dr Trevor Williams, the first executive secretary of the International Board for Plant Genetic Resources (later, the International Plant Genetic Resources Institute) at FAO. Each study provided data on germplasm exchanges between the centres and the country in question between 1972 and 1990/1. The analysis provided in this article is based on data from fifteen countries covered by IFAR: Chile, , India, Indonesia, Kenya, Madagascar, Pakistan, Peru, Philippines, Rwanda, Saudi Arabia, Syria, Tanzania, Uruguay, and Zimbabwe. Exchanges were broken down by categories of crops. The results are shown in Table 2.

Table 2: Exchanges of germplasm samples between selected developing countries and IARC genebanks, 1972–90/1

Crop category No. of samples provided by No. of samples distributed to 15 developing countries to 15 developing countries IARCs 63,479 247,386 Roots & tubers 17,726 15,470 Legumes & 33,031 202,130 pulses Vegetables 2,712 47,502 Forages 7,381 16,928 TOTAL 124,329 529,416 Source: IFAR (1994).

The summary data in Table 2 show that the fifteen countries surveyed were net recipients of samples, at a ratio greater than 4 to 1, during the 18-19-year period. In other words, on average, the number of germplasm samples they received from the collections was many times more than they contributed.16 This is significant in several respects. The period 1972-90/1 is unique in that it was during this period that international efforts were mobilised to collect plant genetic

16. It is important to recognise that data of this type do not reveal how many unique accessions are contained in the samples, or the genetic content of these accessions. Samples from the same accession can be distributed many times, making its potential use value in breeding even greater. Further research will address some of these issues. 190 Cary Fowler, Melinda Smale and Samy Gaiji resources. IPGRI (then IBPGR) was founded, collecting expeditions were financed, genebanks were established, and a system of ‘base collections’ of major crops was put into place. It is during this period that the greatest concentrated outflow of genetic resources from developing countries took place. Nevertheless, these countries still received substantially more germplasm samples from the IARCs (CGIAR plus AVRDC) than they contributed even during this period of intense collecting. Total collections of plant genetic resources for food and agriculture (PGRFA) by the CGIAR have been declining since 1987 (SGRP, 1996). In 1994, they hovered around 2,000 accessions for all crops, a figure roughly equal to the number of CIMMYT genebank distributions of one crop – maize – made to developing countries that year. One might expect – and the data provided in the previous section indicate – that, with the slowdown in collecting, the ratio of samples provided to those requesting them to the new accessions collected, might broaden.

Germplasm flows in an ‘open access’ system

The explanation for the above data is relatively straightforward. Countries providing materials to the informal multilateral system have access not only to their own, but also to those of other countries. It is important to note, therefore, that the fifteen countries studied by IFAR did not ‘lose’ the 124,000 accessions provided to the IARCs. Presumably, they still had access to those accessions, either in their own genebanks or their farmers’ fields, or, if needed, from the centres themselves. What they gained, however, was access to additional materials provided by other countries, ‘insurance’ provided by the centres’ storing the material in safe conditions, and access to information generated by centre scientists in their research work with the genetic materials. The particular advantage of this system as compared with one of bilateral ‘trade’ is that the individual country provides something that may be both plentiful and renewable in exchange for something it does not have, but needs. Data provided in this article indicate that, had access to needed genetic material been sought from other countries under a bilateral regime, individual countries would have had to negotiate to acquire or ‘buy’ a considerable number of samples, a situation which would clearly have been problematic for many of them. Simple transaction costs associated with negotiating for access with more than 125 countries (the number represented in the collections of these crops) would have been substantial for any country wishing to gain access bilaterally to the range of materials currently offered through the centres (Visser et al., 2000) A second important observation can be made concerning Table 2. The fifteen countries were net recipients of germplasm samples in all but one of the categories of crops. Negotiations at FAO indicate that countries are willing to establish a multilateral system of facilitated access only for a smaller list of crops (FAO, 1999). Such a list, as currently conceived, would omit many vegetable crops, and possibly even forages. The countries surveyed by IFAR received almost eighteen times as many samples of vegetables as they provided, and more than twice as many samples of forages. The data indicate a historically high degree of reliance of developing countries on the existing ‘open access’ system of the IARCs for precisely those crops least likely to be included in a facilitated access agreement in the future. One must conclude that a Unequal Exchange? Recent Transfers of Agricultural Resources 191 multilateral system of facilitated access that does not include such materials would endanger the favourable flow of samples that has taken place in recent years. The case studies analysed in this article show that, during a peak period of plant collecting, distributions to developing countries still far exceeded flows from developing countries. This pattern held for all crop categories except roots and tubers,17 which in this case might be explained by the particular composition of the countries sampled. Interestingly, future access to certain crop categories (e.g., vegetables and forages) might be restricted and limited if proposals advanced by some countries (including a number of developing countries) are approved at FAO. Historically, developing countries, considered individually, have been major net recipients of these materials in terms of quantity of samples transferred. Restrictions may alter this pattern to their disadvantage.

Germplasm flows of selected crops from international centre genebanks

The fifteen case studies highlighted above provide one means of assessing germplasm flows for a broad range of crops over a long period of time. It is also possible to gain further insight into the nature of these flows by looking at data concerning specific crops handled by the CGIAR. Not surprisingly, these crop-specific data show the same patterns: developing countries are substantial recipients of germplasm samples from CGIAR genebanks, and the bulk of germplasm flows are currently towards these countries. All crops with relevant and reasonably ‘complete’ data in the SINGER database as of March 2000 were analysed. The resulting list is composed of crops (and in the case of Aegilops, a wild relative) of interest to a broad range of countries: rice, barley, , faba , chickpea, Aegilops, groundnut, and pigeon . Some data for maize and wheat have also been included. The data show that developing countries received the majority of samples of each crop distributed by CGIAR genebanks (Table 3): over 90% of the transfers of pigeon and chickpeas, and close to 100% of the groundnuts. The percentage was lowest for maize and barley; nevertheless, over 60% each went to developing countries. It should be noted that the flows do not need to be overwhelmingly in the direction of developing countries to indicate that they are benefiting. Likewise, the fact that developed countries received only a small percentage of the total transfers of groundnuts does not mean that they were hurt by, or failed to benefit from, this allocation. Furthermore, the SINGER data demonstrate that a large proportion of flows occur within regions and among the developing countries, with the CGIAR centres serving as intermediaries for these transfers. Detailed data from the ICRISAT and IRRI genebanks reveal that over the past 25 years, samples sent to commercial companies represented less than 5% of all transfers for rice, chickpea, groundnut, and pigeon pea. While sorghum seed samples distributed to commercial companies reached 31% of transfers in 1991 (a year of relatively low total transfers), they averaged less than 5% annually from 1973 to 1997. Over that period, over 90% of seed samples from ICRISAT collections went to national programmes, CGIAR centres, and universities (Table 4).

17. In this category the ratio was 1.15 samples provided by countries to every sample received. 192 Cary Fowler, Melinda Smale and Samy Gaiji

Table 3: Numbers and percentage of seed samples distributed from CGIAR collections to developed and developing countries, by crop

Crop or Genebank Time Developing countries Developed countries Species collection perioda No. % No. % Rice IRRI 1985-97 167,937 89.8 19,078 10.2 Wheat CIMMYT 1987-98 30,569 76.9 9,201 23.1 Maize CIMMYT 1987-98 12,938 63.0 7,602 37.0 Barley ICARDA 1990-97 17,625 67.1 8,623 32.9 Sorghum ICRISAT 1974-97 171,071 86.7 26,240 13.3 Faba ICARDA 1990-97 13,280 82.8 2,750 17.2 Chickpea ICARDA 1990-97 16,805 89.6 1,951 10.4 ICRISAT 1974-97 160,319 93.6 11,047 6.4 Aegilops ICARDA 1990-97 13,414 88.5 1,746 11.5 Groundnut ICRISAT 1974-97 131,251 99.0 1,275 1.0 Pigeon pea ICRISAT 1974-97 61,205 97.7 1,466 2.3 a) Refers to time period for which data are currently available from CGIAR genebanks. Sources: Singer databse for IRRI, ICARDA and ICRISAT. B. Skovmand and S. Taba, CIMMYT genebank for CIMMYT.

Table 4: Percentage of seed samples distributed from ICRISAT collections, by user and crop, 1993-7

Chickpea Groundnut Pigeon pea Sorghum Total NARS 48.3 61.0 51.1 48.5 50.8 Genebank/network 2.6 2.1 4.9 3.3 3.0 Regional organisation 0.0 1.0 0.1 0.0 0.2 /individual 0.2 3.8 0.5 0.2 0.9 NGO 0.7 0.1 0.5 0.2 0.3 University 47.5 29.6 40.2 39.8 39.9 Company 0.2 0.1 1.4 5.5 2.9 NA/unknown 0.5 2.1 1.1 2.4 1.8 Source: ICRISAT data in SINGER database.

Flows of improved germplasm from centre breeding programmes

While transfers of materials from CGIAR genebanks are significant both in terms of volume and in value to crop breeding programmes, transfers of breeding lines and other improved materials are substantially larger both numerically and, one might argue, in terms of economic importance. Breeding time and costs are reduced, and varieties are released to farmers more quickly. Furthermore, recipients of the improved germplasm Unequal Exchange? Recent Transfers of Agricultural Resources 193

(e.g., national research programmes) also benefit from the high level of expertise and the modern technologies used by the centres in producing the materials. Unfortunately, very few data on the transfer of improved materials are currently available, so that a few isolated, but highly revealing, examples will have to suffice. They are for rice, wheat, and maize - the world’s three major cereals. The data come from IRRI and CIMMYT, the centres historically most associated with and responsible for these crops (though other CGIAR centres also work on them). From 1994 to 2000, CIMMYT distributed 1.2 million seed samples of bread wheat, durum wheat, triticale, barley and other lines. In 1998-9 alone, the weight of seed shipped was over 11,000 kg. In each year, 65-77% of samples went to developing countries, making 71.3% of the total over the six-year period (Table 5).

Table 5: Distribution of breeding lines sent by CIMMYT international nurseries to developed and developing countries, 1994-9 (%)

FSU/Eastern Europe Other Developed Developing Bread wheat 5.8 19.5 74.7 Durum wheat 7.3 26.1 66.6 Barley 7.9 24.4 67.7 Triticale 7.4 24.9 67.7 Special nurseries 6.7 20.9 72.5 All nurseries 6.6 22.1 71.3 Source: Calculated from data provided by CIMMYT international nurseries.

The greatest share was transferred to public co-operators. About a quarter of the recipients in the developed world were in the private sector. Though the total transfers to developing countries were much greater for each type of material, less than 10% of co-operators in these countries were recorded as being from the private sector (Table 6).

Table 6: Percentage of private co-operators receiving breeding lines from CIMMYT international nurseries, 1994-9

Barley Bread Durum Triticale Special wheat wheat nurseries Developing countries 6.0 9.9 7.5 1.0 1.0 FSU/Eastern Europe 10.0 6.6 1.5 1.0 0.9 Developed countries 20.2 27.7 23.9 0.8 0.8 Note: Co-operators may be institutions distributing seed to many individuals. Source: Ibid.

What historical role has this transfer played in international wheat improvement? What contribution does it make to the varieties released by developing countries and sown by their farmers? From 1966 to 1997, 85% of all spring bread and 86% of all spring durum wheat varieties released in developing countries had CIMMYT ancestry (Table 7) (Heisey et al., 1999). Most of the wheat grown in the developing world is spring habit, 194 Cary Fowler, Melinda Smale and Samy Gaiji and most of the spring habit wheat sown in the fields of developing country farmers is CIMMYT-related..18 The most recent global data for maize are currently being assembled, but have not yet been published. Results for Latin America demonstrate that, while the maize seed industry is increasingly dominated by private companies, both public and private breeders have made extensive use of CIMMYT materials. From 1966 to 1997, approximately 55% of all varieties and hybrids released by public breeding programmes in Latin America contained CIMMYT germplasm. An estimated 75% of all seed sold by private companies in Latin America in 1996 contained CIMMYT-derived germplasm. Public sector breeders now tend to subject CIMMYT materials to additional cycles of selection before using them to form finished cultivars. Private companies use CIMMYT seed in different ways, depending on their size and structure (Morris and López-Pereira, 2000). Up to May 2000, more than a thousand maize trials had been requested from CIMMYT, compared with 658 in 1999 and 188 in 1998. Each trial may contain from 50 to as many as 500 distinct lines. Moreover, CIMMYT projects that it will produce more than 500 new varieties for testing and use by developing country national programmes in the next 2-3 years (Córdova and Pandey, 2000). In 1999, IRRI sent out 24,831 samples of IRRI-developed materials directly to 43 countries. An additional 4,600 samples of improved materials originating at IRRI were distributed through the International Network for Genetic Evaluation of Rice (INGER), which is co-ordinated by IRRI (E. Javier, INGER co-ordinator, pers. comm. 2000). INGER provided a total of 18,533 samples in 1999 (70% originating from national research programmes). Countries participating in INGER gain access to materials from other countries, and in some cases they have been able to import them directly from the INGER nurseries. Evenson and Gollin (1997) have estimated that ending the INGER programme would reduce the number of rice varieties released by 20 varieties per year, entailing an economic loss of $1.9 billion per annum. A bilateral system for access and benefit-sharing for crop genetic resources would imply that each country would be responsible for regulating access, conserving indigenous materials, and improving germplasm for use by farmers.19 CGIAR programmes, dependent as they are on the easy exchange of genetic materials and the provision of research funds by donors, might appear out of place or unimportant in such a context. Failing to recognise and understand the CGIAR’s role may threaten its access to both funding and germplasm and therefore compromise the continued production of public goods by the centres.

18. A substantial CIMMYT contribution does not necessarily imply increased or excessive genetic uniformity, however. CIMMYT-related varieties represent a vast array of germplasm constituted by genetic recombination of diverse sources of materials from throughout the wheat-growing world. A recent summary of scientific findings concludes that genealogical diversity and molecular genetic diversity have increased over time in CIMMYT parents, even while yield potential, yield stability, nitrogen use efficiency, and tolerance to heat and drought have improved. This ‘modern’ form of diversity depends very much on the exchange of germplasm in a multilateral context (Smale et al., forthcoming). 19. A shift to a predominantly bilateral system would also be likely to increase conservation (genebank) costs dramatically. Countries acquiring materials through bilateral transactions would face the alternative of investing in genebank conservation of those genetic resources, or in the costs associated with resupply, when needed. Table 7: CIMMYT contribution to varieties of spring bread and durum released in developing countries, 1966-97

1966-70 1971-5 1976-80 1981-5 1986-90 1991-7 1966-97 bread durum bread durum bread durum bread durum bread durum bread durum bread durum Selected from37.7 25.0 48.2 65.2 40.9 76.9 45.8 67.6 52.3 75.0 52.7 76.8 47.5 70.3 CIMMYT cross With 20.4 0.0 23.9 0.0 38.5 7.7 32.0 17.6 28.6 20.5 29.2 19.6 29.3 14.4 CIMMYT parent With 6.0 8.3 6.2 0.0 6.1 0.0 9.1 2.9 10.1 2.3 8.3 1.8 7.9 2.1 CIMMYT ancestor Improved 35.9 66.7 21.7 34.8 14.6 15.4 13.1 11.8 9.1 2.3 9.8 1.8 15.2 13.3 tall, , other semi- dwarf, or unknown Source: Heisey et al. (1999). 196 Cary Fowler, Melinda Smale and Samy Gaiji

Restoration of germplasm to national programmes

According to FAO:

The incidence of disasters caused by war, civil strife, drought, flood, and fire has increased in the past decades and is at record levels in the 1990s. Such calamities have profound effects on agricultural systems, involving farmer seed systems as well as the social structures which bind communities together....During disasters, countries and communities are often caught off- guard, and they frequently lack the capacity to restore their seed supply systems, often leading to long-lasting seed and food insecurity (FAO, 1998b).

Sustainable agriculture requires, at the very least, seeds and planting materials adapted to the natural environment (and ideally, to the cultural and economic ‘environment’ as well). Individual accessions or entire collections can be lost through accidents or mismanagement even in ‘normal’ times. Following natural disasters, war and civil strife, the large-scale loss of adapted seed and genetic resources can pose a serious threat to countries, economies and food security (Sperling, 1997; Richards and Ruivenkamp, 1997). In such situations, restricted-access systems are ill-equipped to meet the needs of affected countries. Like the provision of improved materials, the use of international collections to restore germplasm is seldom fully appreciated in discussions of germplasm transfers. The scale of restoration activities is substantial, however. Between 1981 and 1995, CGIAR centres restored germplasm to 38 countries (SGRP, 1996): Argentina, Bolivia, Botswana, Brazil, Cambodia, Cameroon, Chile, Dominican Republic, Ecuador, Eritrea, Ethiopia, Gambia, Guatemala, Guinea, Guinea- Bissau, Honduras, India, Iran, Kenya, Liberia, Mali, Mexico, Myanmar, Nepal, Nigeria, Pakistan, Panama, Paraguay, Peru, Philippines, Rwanda, Senegal, Sri Lanka, Sudan, Tanzania, Turkey, Uruguay, and Zambia. Since 1995, there have been efforts to restore germplasm and planting materials in at least two additional countries: Mozambique and Somalia. The quantity of materials supplied to individual countries is not known. Restoration of germplasm following accidental loss, or in the wake of natural disasters and civil strife, is more common than most people realise. Restoration to a particular country depends on samples being stored outside that country or area of danger. In constructing access regimes, countries may want to consider the effect that restrictions will have on the ability of other countries or the international community to restore germplasm in disaster situations.

Intellectual property rights and germplasm flows

As noted above, developing countries seek genetic resources for a variety of purposes: conservation, restoration, plant breeding, and for direct introduction and use. The data presented here indicate that they have received large quantities of genetic resources through the CGIAR system. With the advent of new and expanded forms of intellectual property rights applicable not only to plant varieties but to their genetic components as Unequal Exchange? Recent Transfers of Agricultural Resources 197 well, one might ask whether and to what extent developing countries are actually free to use what they have obtained.

FAO-CGIAR Agreements

The CGIAR collections discussed in this article were assembled during a time when PGRFA were considered ‘the common heritage of humankind’ (Guarino et al., 1995). There was an implicit, and sometimes explicit, guarantee that the collections would remain in the ‘public domain’, cared for by the CGIAR, but available for all to use. In keeping with this history, the individual CGIAR centres signed identical agreements with FAO in 1994 ‘placing collections of plant germplasm under the auspices of FAO’. These agreements formalised the status of the materials, establishing the CGIAR collections as the only collections with a recognised international character and formally in the international ‘public domain’. The centres agreed with FAO:

…to hold the designated germplasm in trust for the benefit of the international community, in particular the developing countries in accordance with the International Undertaking on Plant Genetic Resources and the terms and conditions set out in this Agreement (FAO/CGIAR, 1994).

Specifically, they also reaffirmed that they made no claim to the materials and that they would not take out intellectual property rights (IPRs) over the ‘germplasm or related information’. Furthermore, they agreed to bind those to whom they distributed germplasm to the same conditions through the use of FAO-sanctioned ‘Material Transfer Agreements’. All samples covered by the FAO/CGIAR agreements are now distributed under contracts designed to maintain the public nature of the collections. Contracts made, however, are contracts that can be broken either formally or in spirit.20 Also, technological developments and changes in law can effectively redefine the status and circumscribe the use of that which is ‘public’ (Auerbach, 1980).21 Civil society organisations have been particularly vocal in expressing concern that samples (including their genetic components) held ‘in trust’ by the CGIAR might be patented in some countries and thus effectively removed from the public domain in those countries. Indeed, they have charged that specific attempts have been made to do so.22 An assessment of this situation and its impact on germplasm transfers and use requires an examination of both the powers and limitations of intellectual property laws, and the available experience with ‘public domain’ resources such as those held by the CGIAR.

20. It is important to note that germplasm – including a great deal of materials essentially ‘identical’ to those held ‘in trust’ by the CGIAR – may be obtained from sources that are not party to the FAO-CGIAR agreements. Thus, not all recipients or users of germplasm are bound to observe the conditions of the Material Transfer Agreements. 21. For example, many countries are now promulgating national laws governing access to genetic resources. While this topic is beyond the scope of this article, it is likely that many of these laws will have the effect of restricting access and use to materials that were formerly available with few restrictions. Indeed, these laws may limit availability and use more than the application of intellectual property rights. 22. See, for example, materials produced by the Rural Advancement Foundation International, available through its website: www.rafi.org 198 Cary Fowler, Melinda Smale and Samy Gaiji

A wide range of IPRs is now available for use in protecting plant-related innovation in many countries: plant breeders’ rights (also known as plant variety rights), patents, trade secrets and copyrights. The latter two will not be considered here, because, by their nature, they cannot be used to ‘lock up’ or prevent the future use of genetic materials already in the public domain. Plant breeders’ rights and patents, however, deserve attention, though a thorough treatment is beyond the scope of this article.

Plant breeders’ rights

As of September 2000, 46 states were party to the International Convention for the Protection of New Varieties of Plants (UPOV), including 25 developing countries or countries with economies in transition. The breeders’ rights protection offered under UPOV-style laws is aimed at the variety level. In effect, it confers rights to a particular combination of genes manifested as a distinct, uniform and stable variety. The protected variety may, however, be used by others in basic research or as a parent in a breeding programme without recourse or obligation to the right-holder. Such protection could encroach on the public domain if rights were granted to a particular variety or sample found in the international collections. In practice, the threat is not great, arising principally when protection is improperly granted. The vast majority of materials in the collections will fail to meet most, if not all, of the criteria for the grant of plant breeders’ rights; they will not be ‘distinct, uniform, and stable’. According to UPOV, ‘The variety shall be deemed to be distinct if it is clearly distinguishable from any other variety whose existence is a matter of common knowledge…’ (UPOV, 1991). Assuming that materials held ‘in trust’ by the CGIAR are a matter of ‘common knowledge’, it would appear that they would be ‘off limits’ to protection. The seeker of rights must also be the one who has bred, or discovered and developed, the variety. It is unlikely that a court would accept the premise that ‘discovery on the shelves of a public genebank’ meets the standards (T. Roberts, pers. comm. 2000). Moreover, little commercial incentive exists for privatising and marketing such varieties, and commercial disincentives (e.g., bad publicity) for trying to do so are significant. In the first five years since the agreements with FAO, the CGIAR distributed more than 500,000 samples of ‘in trust’ germplasm (Hawtin, 2000). Fewer than 200 cases of improper IPR applications/protection by recipients were alleged during this period.23 And, of these, 10 at most were found to warrant detailed investigation. Fewer than half this number are thought to have constituted violation of either the letter or spirit of the agreements. In these cases, either the application or the grant of protection was subsequently withdrawn. Thus in five years of experience with the FAO-CGIAR Agreements, one could say that the level of apparent ‘abuse’ was approximately 0.00001% per annum (Fowler et al., forthcoming). The low rate of ‘abuse’ does not, of course, excuse those situations in which it has taken place, but it adds some perspective to the most celebrated cases, all of which involved mistaken grants of protection that did not meet the applicable law and were therefore withdrawn or overturned.

23. Virtually all have been associated with plant breeders’ rights. Unequal Exchange? Recent Transfers of Agricultural Resources 199

Patents

Traditional patents may now also be used in the United States, Japan, and Australia to protect plant varieties, and in these three countries plus much of Europe and some developing countries to protect genes and other components. ‘The plant containing the gene naturally is not novel for the purposes of the patent law’, and thus not covered by patent provisions according to most observers (Barton, 1998). According to the US Patent and Trademark Office, ‘a patent on a gene covers the isolated and purified gene, but does not cover the gene as it occurs in nature’ (USPTO, 2001). Most experts conclude, therefore, that only a ‘biotechnological’ use of a patented gene would be under the control of the patent owner; in other words, if the gene occurs naturally or is obtained through conventional methods, the patent may not block its use, or the use of the sample that contained it. Thus, it is generally thought that patents do not affect breeding work undertaken with plants, such as genebank samples of farmer varieties, that naturally contain the patented gene (Bragdon, 2000; C. Correa, pers. comm. 2000). In addition, patent laws are national in character, meaning that a gene patent granted in the United States is neither applicable nor enforceable in Tanzania, for example, or any other country, unless patent protection is also available, sought, and granted there (Bragdon, 2000). Nevertheless, the use of patents is problematic and troubling, as it has the potential, in certain jurisdictions, for restricting some uses of materials held ‘in trust’ for the international community by the CGIAR.

IPRs and the effect on the public domain

The high level of requests for and distributions of samples – more than 100,000 per annum – would, in itself, seem to present a prima facie case that the materials have largely remained in the public domain and continue to be available for use after they are acquired. Presumably, few would request materials they could not use because of IPR restrictions. Thus, as with widespread notions about the direction of germplasm flows, there is actually less than meets the eye to popular press accounts that IPR laws are being used to any substantial extent, or effectively, to privatise and practically remove materials from the ‘commons’. While this is the current state of affairs, one cannot be sure that the situation will remain so. IPRs are evolving quickly in response to both technological development and political pressures. If the public domain and the public interest are to be protected, the international community, civil society organisations, and organisations such as the CGIAR, will need to remain vigilant, and certain responsibilities towards the material and its defence will need to be clarified. In view of the broad distribution of genetic resources over the past 10,000 years and consequently the likelihood of multiple sources for the same material, one might expect more difficulties in (i) identifying and defining what material is in the public domain, (ii) ascertaining who has the responsibility for ‘protecting’ this public domain, and (iii) determining how and on what grounds this might be done. Thus far, all cases of alleged abuse have been settled without recourse to the courts. However, CGIAR centres are unlikely to commit themselves to costly legal procedures to protect the 500,000 samples they hold ‘in trust’ on a sample-by-sample basis no matter how much they may value 200 Cary Fowler, Melinda Smale and Samy Gaiji keeping the material in the public domain.24 This raises the question of what responsibilities the international community has for material held in trust for it.

The future of genetic resources in the public domain

Two avenues to alleviating the political tension between IPRs and ‘public’ germplasm have been considered recently. The first would involve strengthening the language in Material Transfer Agreements, making access to genetic materials conditional upon agreeing not to exercise any IPRs that would abridge future and further research and use. This approach would allow ‘public’ materials to be used to create IPR-protected products, but would not allow them to be ‘locked up’. Those accessing the materials in the future would find only one limitation: they would not be able to use them to create something identical to that made by the previous user who had obtained a patent on it. As long as plant-related IPR laws exist, it is inevitable that genes found in the CGIAR’s ‘in trust’ collections will be contained within IPR-protected products. So ubiquitous are some genes in these collections, that it would probably not be possible to construct a plant – or even a human being – without resort to them. The granting of access, however, need not imply permission to exercise IPRs in such a way as to restrict further use (Fowler et al., forthcoming). Access can come with conditions established contractually and/or through international agreement. The second avenue is under negotiation in the FAO Commission on Genetic Resources. A proposal, backed by the European Union, Norway, and the G-77 countries, calls for the payment of royalties into a fund for the implementation of agreed plans and programmes whenever use of the accessed ‘in trust’ materials

…results in a product that is a plant genetic resource covered by any form of intellectual property right or commercial protection…that restricts utilization of the product for research and plant breeding…. (FAO, 2000)

The above text is opposed, thus far, by the US, Canada, Australia and New Zealand, but is still under debate. The proposal implicitly legitimises the use of IPRs with genetic material covered by the eventual agreement at FAO, but establishes a benefit-sharing mechanism to compensate the public for any IPR-related restrictions on future use of the material.

Summary and conclusion

The data on germplasm transfers presented and analysed in this article cover a period of over 20 years. It was a period of intense plant collecting in developing countries, one in which transfers from developing countries were probably at a never-to-be-repeated high. Nevertheless, transfers to developing countries were considerably larger than

24. It would appear that FAO has been satisfied with CGIAR performance in honouring the terms of the agreements, since it has never complained that the CGIAR has acted improperly in distributing samples or enforcing the agreements. Relations between the two institutions concerning this matter are perceived to be excellent. Unequal Exchange? Recent Transfers of Agricultural Resources 201 receipts from them. It seems evident that we are now in a period of reverse transfers; materials (both farmer varieties and improved materials) are flowing in large quantities into the historic centres of diversity from genebanks and breeding programmes. Regarding germplasm flows, the findings of this study can be summarised as follows:

(i) Developing countries have been and continue to be net recipients of a large amount of germplasm samples from the genebanks and breeding programmes of IARCs. First, they receive substantially more germplasm samples from CGIAR genebanks than they furnish,25 a situation likely to become even more pronounced in the future. Secondly, they receive significantly more germplasm samples from the genebanks than do developed countries. Thirdly, distributions of germplasm samples from CGIAR genebanks and crop programmes to private companies appear to be minor for most of the crops studied. Finally, a massive amount of improved germplasm is flowing in nursery shipments from centre breeding programmes to developing countries. For the crops considered here, the provision of this material is also skewed in favour of developing countries compared with developed countries. (ii) Material provided to CGIAR centres is not ‘lost’ by the supplier. Source countries can usually continue to access the material from their own collections, from farmers’ fields, or from the CGIAR centre, all of which have policies ensuring ‘open’ access. (iii) Restoration of germplasm following accidents, natural disasters, war and civil strife, has helped a large number of countries recover material provided earlier to the centres. No precise data are available regarding the number of materials provided in these circumstances

Individual countries, whether developed or developing, can expect to be net recipients in all categories of crops, and virtually every particular crop, under a multilateral system of facilitated access. This is already the case today. If the current inflow:outflow ratio of germplasm samples remains as it is, but access is subject to restrictions or ‘bilateralised’, virtually all countries will be required to obtain substantially more from others than they themselves have to offer. If such transfers become monetary-based, all developing countries stand to lose. Decreased access to and use of PGRFA will add considerably to transaction-associated financial losses, through lower agricultural productivity. Incidents of restrictions being imposed through IPRs on the use of materials distributed by the CGIAR are statistically rare. In most cases, they have been associated with the improper and transient grant of patents or plant breeders’ rights for varieties, a situation which argues against the likelihood that they will pose major difficulties in the future. The sheer size of current flows also indicates that the materials are still seen as useful and capable of being used without legal hindrance. However, patents on components of varieties – genes, alleles, cells, etc. – are potentially more problematic.

25. As already stated, comparisons measuring ‘diversity’ are not possible, and in any case would not be particularly relevant to the principal thesis of this article. Nevertheless, as argued earlier, most individual developing countries may also be net recipients of ‘diversity’. 202 Cary Fowler, Melinda Smale and Samy Gaiji

Currently applicable in a limited number of countries, these laws, if expanded more widely and interpreted as restricting the use of naturally-occurring genes in naturally- occurring samples, could pose a significant threat to the practical use of CGIAR and other ‘public’ materials. Under current interpretations, however, such patents do not appear to be posing frequent or major obstacles to the availability or use of samples held by the CGIAR. On-going negotiations at the FAO, as mentioned earlier, will address questions of IPRs in relation to the access to and use of genetic resources. They will determine whether future transfers of genetic materials will be handled much as they have been during the period covered by this study, through a regime of ‘facilitated’ access, or bilaterally on a case-by-case basis. Negotiating positions are likely to be influenced by perceptions of the balance of benefits and losses with the contemporary unrestricted flows of germplasm. Based on the data provided in this article, it can be argued that continued facilitated access in today’s world is a ‘win-win’ situation for all, and one that is particularly important for developing countries.

References

Auerbach, C. (1980) ‘The Relation of Legal Systems to Social Change’, Wisconsin Law Review. Barton, J. (1998) ‘The Impact of Contemporary Patent Law on Plant Biotechnology Research’, in Intellectual Property Rights III. Global Genetic Resources: Access and Property Rights. Madison, WI: Crop Science Society of America/American Society of . Berg, T., Bjørnstad, A., Fowler, C. and Skrøppa, T. (1991) Technology Options and the Gene Struggle. Noragric Occasional Paper Series. Aas: Agricultural University of Norway. Bragdon, S. (2000) ‘What is Patentable? Gene Patents, Plant-related Innovation and the Functioning of the Multilateral System of Exchange of Plant Genetic Resources’. Unpublished manuscript. Rome: IPGRI. Brush, S. (ed.) (1999) Genes in the Field: On- Conservation of Crop Diversity. Boca Raton: Lewis. Córdova, H. and Pandey, S. (2000) ‘Developing Core Germplasm and Integrating Interdisciplinary Approaches for the Improvement of Maize’. Unpublished project document. El Batan, Mexico: CIMMYT. Crosby, A. (1972) The : Biological and Cultural Consequences of 1492. Westport, CT: Greenwood Press. Crosby, A. (1986) Ecological Imperialism: The Biological Expansion of Europe, 900- 1900. Cambridge: Cambridge University Press. De Candolle, A. (1886) Origin of Cultivated Plants. New York: Hafner (1967 reprint). De Souza Silva, J. (1989) ‘Science and the Changing Nature of the Struggle Over Plant Genetic Resources: From Plant Hunters to Plant Crafters’. Ph.D dissertation, University of Kentucky. Evenson, R. E. and Gollin, D. (1997) ‘Genetic Resources, International Organizations, and Improvement in Rice Varieties’, Economic Development and Cultural Change 45 (3): 471-500. Unequal Exchange? Recent Transfers of Agricultural Resources 203

FAO (1998a) The State of the World Plant Genetic Resources for Food and Agriculture. Rome: FAO. FAO (1998b) Restoring Farmers’ Seed Systems in Disaster Situations: Proceedings of the International Workshop on Developing Institutional Agreements and Capacity to Assist Farmers in Disaster Situations to Restore Agricultural Systems and Seed Security Activities. Rome: FAO. FAO (1999) ‘Composite Draft Text of the International Undertaking on Plant Genetic Resources, incorporating the Chairman’s Elements’. CGRFA-8/99/13 Annex. Rome: FAO. FAO (2000) ‘Composite Draft Text of the International Undertaking on Plant Genetic Resources. Fourth Inter-Sessional Meeting of the Contact Group’, CGRFA/CG- 4//00/2, Neuchatel, Switzerland, 12-17 November. Rome: FAO. FAO/CGIAR (1994) The Agreement Between [name of Center] and the Food and Agriculture Organization of the United Nations (FAO) Placing Collections of Plant Germplasm Under the Auspices of FAO. Rome: FAO. Ford-Lloyd, B. and Jackson, M. (1986) Plant Genetic Resources: An Introduction to their Conservation and Use. London: Edward Arnold. Fowler, C. (1994) Unnatural Selection: Technology, Politics and Plant Evolution. Yverdon: Gordon and Breach Science Publishers. Fowler, C. (2000) ‘The Plant Patent Act of 1930: A Sociological History of its Creation’, Journal of the Patent and Trademark Office Society 82 (10). Fowler, C. and Mooney, P. (1990) Shattering: Food, Politics and the Loss of Genetic Diversity. Tucson, AZ: University of Arizona Press. Fowler, C., Hawtin, G., Iwanaga, M. and Engels, J. (forthcoming) ‘Germplasm and Related Information and the Question of Derivatives in the FAO-CGIAR Agreements’, in Issues in Plant Genetic Resources. Rome: IPGRI. Guarino, L., Rao, V.R. and Reid, R. (1995) Collecting Plant Genetic Diversity. Wallingford, Oxon: CABI/IPGRI. Harlan, H.V. and Martini, M.L. (1936) ‘Problems and Results in Barley Breeding’, in Yearbook of Agriculture, 1936. Washington, DC: US Department of Agriculture. Harlan, J.R. (1975) Crops and Man. Madison, WI: American Society of Agronomy/Crop Science Society of America. Hawtin, G. (2000) ‘Statement to the Chairman’s Contact Group: FAO Commission on Genetic Resources for Food and Agriculture’. Unpublished. Neuchatel, Switzerland. Heisey, P.W., Lantican, M. and Dubin, H.J.. (1999) ‘Assessing the Benefits of International Wheat Breeding Research: An Overview of the Global Wheat Impacts Study’, Part 2 of the 1998-9 World Wheat Facts and Trends. Mexico, DF: CIMMYT. Hoisington, D., Khairallah, M., Reeves, T., Ribaut, J-M., Skovmand, B., Taba, S. and Warburton, M. (1999) ‘Plant Genetic Resources: What Can They Contribute Toward Increased Crop Productivity?’ Paper presented at the NAS colloquium ‘Plants and Population: Is There Time?’ and published in Proceedings of the National Academy of Sciences, Vol. 96. International Fund for Agricultural Research (IFAR) (1994) Agriculture in [Name of Country]: The Role of International Agricultural Research Centres. Arlington, VA: IFAR. 204 Cary Fowler, Melinda Smale and Samy Gaiji

Jackson, M., Loersot, G., Rao, S., Jones, M., Guimaraes, E. and Ng, N. (1997) ‘Rice’, in in Trust. Cambridge: Cambridge University Press. Mooney, P. (1983) ‘The Law of the Seed: Another Development and Plant Genetic Resources’, Development Dialogue (1-2): 7-172. Morris, M. L. and López-Pereira, M. A. (2000) 1999 Impacts of Maize Breeding Research in Latin America, 1966-1997. Mexico, DF: CIMMYT. Organisation for African Unity / Scientific, Technical and Research Commission (1998) Declaration and Draft Model Law by the OAU / STRC Task Force on Community Rights and Access to Biological Resources. Addis Ababa: OAU. Palacios, X. F. (1998) Contribution to the Estimation of Countries’ Interdependence in the Area of Plant Genetic Resources. Background Study Paper No. 7, Rev. 1. Rome: FAO Commission on Genetic Resources for Food and Agriculture. Richards, P. and Ruivenkamp, R. (1997) Seeds and Survival: Crop Genetic Resources in War and Reconstruction in Africa. Rome: IPGRI. Shiva, V. (1990) ‘The Seed and the Spinning Wheel: Biotechnology, Development and Biodiversity Conservation’, in Proceedings of the International Conference on Conservation of Genetic Resources for Sustainable Development. Røros, Norway. Shiva, V. (1997) Biopiracy: The Plunder of Nature and Knowledge. Cambridge, MA: South End Press. Shiva, V. (1999) Stolen : The Hijacking of the Global Food Supply. Cambridge, MA: South End Press. Simmonds, N.W. (1976) Evolution of Crop Plants. London: Longman. Smale, M, Reynolds, M., Warburton, M., Skovmand, B., Trethowan, R., Singh, R., Ortiz-Monasterio, I., Crossa, J., Khairallah, M. and Almanza, I. (forthcoming) Dimensions of Diversity in CIMMYT Bread Wheat from 1965 to 2000. Mexico, DF: CIMMYT. Sperling, L. (ed.) War and Crop Diversity. Agricultural Research and Extension Network Paper No. 75. London: Overseas Development Institute. System-wide Genetic Resources Programme (SGRP) of the Consultative Group on International Agricultural Research (1996) Report of the Internally Commissioned External Review of the CGIAR Genebank Operations. Rome: IPGRI. ten Kate, K. and Laird, S. (1999) The Commercial Use of Biodiversity: Access to Genetic Resources and Benefit-Sharing. London: Earthscan. UPOV (1991) International Convention for the Protection of New Varieties of Plants of December 2, 1961, as Revised at Geneva on November 10, 1972, and on March 19, 1991. Geneva: UPOV. United States Patent and Trademark Office (2001) ‘Utility Examination Guidelines’, Federal Register 66 (4) 5 January. Vavilov, N.I. (1926) ‘Studies on the Origin of Cultivated Plants’, Bulletin of Applied Botany 16 (2). Visser, B., Eaton, D., Louwaars, N. and Engels, J. (2000) Transaction Costs of Germplasm Exchange Under Bilateral Agreements. Rome: FAO/Global Forum on Agricultural Research. Zeven, A.C., and Zhukovsky, P.M. (1975) Dictionary of Cultivated Plants and Their Centres of Diversity. Wageningen: Pudoc.