Biotecnologías Agrícolas
Agricultural Biotechnologies in crops, forestry, livestock, fisheries and agro-industry  Biotech-banner

Conference 13 - Background

The role of biotechnology for the characterisation and conservation of crop, forest, animal and fishery genetic resources in developing countries

1. Introduction

On the occasion of World Food Day 2004, the United Nations Secretary-General Kofi Annan urged "individuals and institutions alike to give greater attention to biodiversity as a key theme in our efforts to fight the twin scourges of hunger and poverty and achieve the Millennium Development Goals". He also noted that the unprecedented loss of biodiversity over the past century was a major cause for alarm, where "many freshwater fish species, which can provide crucial dietary diversity to the poorest households, have become extinct, and many of the world's most important marine fisheries have been decimated. Food supplies have also been made more vulnerable by our reliance on a very small number of species: just 30 crop species dominate food production and 90 per cent of our animal food supply comes from just 14 mammal and bird species -- species which themselves rely on biodiversity for their productivity and survival. There has been a substantial reduction in crop genetic diversity in the field and many livestock breeds are threatened with extinction". On the same occasion, FAO's Director-General Jacques Diouf also underlined that although forests are among the world's most important repositories of biological diversity, the world forest cover is decreasing at an alarming rate.

It is in this context of declining agricultural biodiversity, that the FAO Biotechnology Forum will host this e-mail conference to consider the role that biotechnology can play in the characterisation and conservation of crop, forest, animal and fishery genetic resources in developing countries. Biotechnology is a broad collection of tools and these tools can be applied for a range of different purposes (e.g. genetic improvement of populations; disease diagnosis and vaccine development; improvement of feeds). The focus in this conference will be on biotechnology tools, such as molecular markers or cryopreservation and reproductive technologies, that can be used directly for the characterisation and/or conservation of genetic resources for food and agriculture. Genetic modification and GMOs will not be considered here.

On 5-7 March 2005, as part of the preparations for this e-mail conference, an international workshop was held in Turin, Italy, entitled "The role of biotechnology for the characterisation and conservation of crop, forestry, animal and fishery genetic resources", organised by the FAO Working Group on Biotechnology, the Fondazione per le Biotecnologie, the ECONOGENE project and the Società Italiana di Genetica Agraria. Proceedings of the workshop include 20 papers covering applications of molecular markers, cryopreservation and reproductive technologies and can be consulted by anyone looking for more detailed technical information in this area. The workshop was split into Session I on the status of the world's agricultural biodiversity; Session II on the use of biotechnology for conservation of genetic resources and Session IV on genetic characterisation of populations and its use in conservation decision-making. Session III presented results from ECONOGENE, a European Union funded project combining a molecular analysis of biodiversity, socio-economics and geostatistics to address the conservation of sheep and goat genetic resources and rural development in marginal agrosystems in Europe.

This Background Document aims to provide information that participants in the e-mail conference will find useful for the debate. Firstly, a brief overview of genetic resources for food and agriculture is provided (Section 2), followed by more specific information regarding the current status of the genetic resources in the different food and agricultural sectors (Section 3). A brief description of the relevant biotechnologies is then given (Section 4), followed by a discussion of some key issues and some questions that might be addressed in the e-mail conference (Section 5). References to articles mentioned in the document are listed in Section 6.

2. A Brief Background to Genetic Resources for Food and Agriculture

As described in Diamond (2002), domestication of crops and livestock, where species selected from the wild adapt to a special habitat created for them by humans, was a momentous development with major consequences for human societies, since "it provides most of our food today, it was prerequisite to the rise of civilization, and it transformed global demography". Domestication seems to have arisen independently on at least five different occasions in different parts of the world where, for example, hunter-gatherers in the Fertile Crescent (an area in Asia, including parts of Iran, Iraq, Israel, Jordan, Lebanon, Syria and Turkey), began domesticating barley, peas, wheat, cows, goats, pigs and sheep about 10,000 years ago. Sometimes, the same species (e.g. cow, sheep) was domesticated independently in different places. The process of domestication has continued ever since, e.g. the rabbit and strawberry were domesticated in the middle ages.

Domestication had major impacts on human societies, making the transition from hunger-gathering to a settled farming existence possible. It also led to major changes in the species that were domesticated. Since their domestication, domestic crops and livestock have been used in almost all environments of the world and many have been used in particular environments for very long periods of time e.g. Soay sheep on Hirta, off West Scotland, are thought to be the survivors of the sheep first brought to Britain by the first farmers 5,000 years ago. Each environment has a certain set of stressors that impacts the production environment. If we consider livestock, these might include climatic features (heat/cold, amount of rain, humidity); feeding conditions (quality/quantity vegetation available, access to water) or diseases (parasitic infections etc.) and, over time, populations adapted to the environment and its specific stressors. Farmers have also exercised a major influence on the domesticated species by preferentially selecting for specific morphological or production traits.

Unlike crops and livestock, domestication is generally a very recent phenomenon in the fishery and forestry sectors. Only a small number of forest tree species have been, to some degree, domesticated, and such action has been taken only over the past half century. Such tree species, used in plantation forestry, are usually no more than 1-2 generations removed from their "wild relatives" (see e.g. Sigaud, 2005). For fish, apart from a few species, such as common carp domesticated about 2,000 years ago, aquaculture (fish farming) is a relatively new development (Bartley, 2005) and the majority of fish consumed today still come from wild populations, although it is predicted that by 2030 consumption from the ever-growing aquaculture industry will exceed capture fisheries.

The agricultural genetic resources to be considered in this e-mail conference, therefore, include a wide range of species, from those domesticated 10,000 years ago to those domesticated in recent times and those which have not been domesticated but are nevertheless of importance for food and agriculture for humankind.

As will be described in more detail in Section 3, many of these genetic resources are endangered. The need to conserve these resources is now widely accepted, generally justified for one or more of a variety of reasons, including their importance as an insurance against future changes in market needs and production conditions; as a source of material for scientific research and future germplasm development; and for cultural and historical reasons, because they are part of our heritage, passed down from previous generations (e.g. Ruane, 2000). The issue has received considerable attention at the international policy-making level. For example, it is addressed under the Convention on Biological Diversity, a legally-binding agreement with 188 Parties, through its agricultural biodiversity work programme, and under the International Treaty on Plant Genetic Resources for Food and Agriculture, a legally-binding treaty with 69 Parties that entered into force in June 2004.

Two major strategies for conservation can be distinguished: in-situ conservation, where a population is maintained in its natural or agricultural habitat, or ex-situ conservation, where it is maintained outside of this habitat. Ex-situ conservation can involve living individuals (e.g. animals in zoos, fish in aquaria, crops or trees in botanical gardens and arboreta) or tissues or genetic material from individuals (such as seeds, pollen, sperm, embryos, DNA). Ex-situ conservation is important for crops, where an estimated 6 million plant, mainly crop, accessions (i.e. samples of a variety collected at a specific location and time) are stored in national, regional and international collections (FAO, 1998a, Chapter 3). Of the 6 million accessions, over 80% are in national collections and about 90% are stored in seed genebanks which, depending on the seed storage facilities, may allow short-, medium- or long-term storage. A genebank is a storage facility where germplasm of plant or animal origin is stored in forms such as seeds, pollen, semen or embryos; or in in vitro culture (germplasm kept as sterile plant tissues or plantlets on nutrient gels); in cryogenic storage; or in a field genebank i.e. as plants growing in the field. For the livestock, forestry and fishery sectors, in-situ conservation is currently more important than ex-situ conservation.

Characterisation of genetic resources goes hand in hand with their conservation as it is fundamental both for understanding what is being conserved and for choosing the genetic resources that should be conserved. Characterisation enables us to identify the key features (both the strengths and weaknesses) of the available genetic resources, and this knowledge can also be used to develop breeding programmes for sustainable use of the genetic resource to harness and/or disseminate the positive attributes identified in the population. Characterisation can also play an important role regarding issues of access to, and benefit sharing from, agricultural genetic resources, as well the promotion and control of "bioprospecting" (i.e. biodiversity prospecting, the search for commercially valuable compounds, substances or genetic material in nature). If a country has characterised its genetic resources, it should be appropriately positioned to develop and implement conservation strategies for targeted species, and to promote and oversee any potential commercial or other enterprises arising from the knowledge generated; to protect the resources (or any of their products) from inappropriate bioprospecting, including that carried out by foreign companies or countries; and to negotiate appropriate compensation for use of these resources by third parties, be they national or foreign.

The kinds of features that could be characterised for the different agricultural populations include their morphological and physiological features; their phenotypic and economic performance in different environments; their interaction with, and relation to, the environment; their value as social or cultural objects; their role in scientific research; their diversity at the gene and DNA sequence level; their population size and degree of endangerment.

3. Status of Genetic Resources in the Different Food and Agricultural Sectors

In this section, we attempt to give an overview of the current situation regarding genetic resources in each of the four sectors.

3.1 Crop genetic resources

Compared to the large numbers of wild plant species available, only very few were actually domesticated (Diamond, 2002). Although this number is small, just a sub-set of them again are of special importance globally. For example, considering plant-derived energy intake worldwide, just three crops (wheat, rice and maize) provide over 50% of the intake and 30 crops provide 95% (FAO, 1998a, Chapter 1). Splitting the world into 17 sub-regions, there are only 12 major crops which supply more than 5% of the plant-derived energy intake in one or more sub-regions i.e. wheat, rice, maize, millet, sorghum, potato, sugar cane, soybean, sweet potato, cassava, the common bean and related species (Phaseolus), banana/plantain. It should, however, be kept in mind that although some specific crops may not have any global/regional significance they may, nevertheless be important at the individual country level (e.g. tef in Ethiopia).

Although, as seen, the number of major crops species is quite small, the diversity within each species can be very substantial. Within a species, cultivated crop varieties (cultivars) can be broadly categorised as farmer's varieties (also known as landraces or traditional varieties i.e. populations that are the product of breeding or selection carried out by farmers, either deliberately or not, continuously over many generations) and modern varieties (also known as high-yielding or high-response varieties, that are the products from professional breeders working in publicly-funded research institutes or private companies). Farmer's varieties tend to maintain high levels of genetic diversity and to be the focus of conservation efforts (FAO, 1998a).

The first comprehensive worldwide report on the status and use of crop genetic resources was published by FAO in 1998. This State of the World's Plant Genetic Resources for Food and Agriculture (SoW-PGR) was developed through a participatory country-driven process and the primary data sources for the publication were detailed Country Reports from a total of 154 countries. In Annex 2 of the report (FAO, 1998a), the state of diversity within each of the 12 major crops is documented. It shows, for example, that the global ex-situ collection of millet contains roughly 90,000 accessions, of which 2% are wild relatives, 33% are landraces and old cultivars, 5% are advanced cultivars and breeding lines while 60% are unknown and mixed. Although incomplete data makes it difficult to give a definitive picture of the kind of material stored in the world's ex-situ collections, it is estimated that roughly half the accessions for which the type of material is known are advanced cultivars and breeding lines, while a third are landraces and about 10% are wild species (FAO, 1998a, Chapter 3).

In their Country Reports, nearly all countries said that genetic erosion (i.e. the loss of genetic diversity, generally as a result of social, economic and agricultural changes) was taking place and was a serious problem. The main cause they cited was the replacement of local varieties or landraces by improved and/or exotic varieties and species and numerous examples were provided (e.g. displacement of traditional barley and durum wheat varieties by introduced varieties in Ethiopia). Other important causes they cited included overexploitation of plant genetic resources, including overgrazing and reduced fallow periods in shifting cultivation; deforestation and land clearance; population pressure and urbanisation; environmental degradation (e.g. desertification, flooding) and changes in agricultural systems (FAO, 1998a, Chapter 1). Preparations are underway for the 2nd report on the SoW-PGR, which should be finalised in 2008.

3.2 Forest genetic resources

Most forest tree species are characterised by inherently high levels of variation and extensive natural ranges. They are long-lived, frequently outcrossing organisms, with generally large distribution ranges. A high level of intra-specific genetic variation is needed to ensure present-day and future adaptability to changing environmental conditions. It is also needed to maintain options and potential for improvement to meet changing end-use requirements. Forests provide a wide range of goods and services such as timber, fibre, fuelwood, food, fodder, gum, resins, medicines, pharmaceutical products and environmental stabilisation. Similar goods and services are often provided by a wide range of genera and tree species. Despite the availability of a large number of forest tree species, less than 500 have been systematically tested for their present-day utility for human beings and less than 40 species are included in intensive selection and breeding programmes (FAO, 2000a).

Estimates for the year 2000 indicate that the world's forest cover was about 3.9 billion hectares, of which 47% was in the tropics, 33% in the boreal zone (i.e. the northernmost forest zone), 11% in temperate areas and 9% in the subtropics (FAO, 2001). About 95% of the forest cover was in natural forest and 5% in forest plantations (defined as forest stands established by planting or/and seeding in the process of afforestation or reforestation). Of the roughly 190 million hectares of forest plantation, 48% had an industrial end-use (e.g. pulpwood for paper); 26% a non-industrial purpose (e.g. fuelwood, soil and water conservation); and for 26% it was not specified.

Most of the forest genetic resources worldwide are therefore found in natural, largely unmanaged forests which, however, are frequently intensively used for the provision of both wood and non-wood products, for local, national and international use. Conversion of forests to other land uses represents the greatest threat to forests and their diversity. According to FAO (2001), "increasing pressure from human populations and aspirations for higher standards of living, without due concern to the sustainability of the resources underpinning such developments, heighten these concerns. While some land use changes are inevitable, it is important that such changes be planned and managed to address complementary goals. Concerns for biological and genetic conservation should be major components of land use planning and forest management strategies". Unlike the situation in crops, however, no global overview of the status of forest tree genetic resources has yet been carried out, although the idea of developing a first global assessment of forest genetic resources is currently being discussed (Sigaud, 2005).

3.3 Animal genetic resources

Of the 50,000 known bird and mammalian species, about 30 have been used extensively for agricultural purposes, with fewer than 14 accounting for over 90% of global livestock production (FAO, 1999). In the millennia following their domestication, animal species were influenced by both natural and artificial selection in a wide range of geographic and climatic areas and production systems, leading to the development of large numbers of genetically diverse domestic animal breeds.

Many of these breeds are currently endangered. In developing countries, the introduction of exotic breeds and their spread through indiscriminate crossbreeding is a major reason for loss of local, indigenous breeds. Other factors include changes in breeders' preferences because of short-term socio-economic influences; degradation of the ecosystem in which the breeds were developed; natural disasters (e.g. drought and diseases); wars and other forms of political unrest and instability (FAO, 1998b).

In 1993, FAO published the first World Watch List for Domestic Animal Diversity, providing a basic global overview of the state of the genetic resources of seven mammalian species (buffalo, cattle, donkey, goat, horse, pig and sheep). The 3rd edition, expanded to cover 30 mammalian and bird species, showed e.g. that of the over 4,000 breeds with population data, roughly 30% could be classified at a high risk of loss and that chickens and cattle have large numbers of breeds at risk while the horse and goose have the highest percentages of breeds at risk of loss (FAO, 2000b). In addition, apart from the wild boar and wild red jungle fowl (ancestors of the domestic pig and chicken respectively), the putative wild ancestors of the major livestock species are in danger of extinction or are already extinct (Hanotte and Jianlin, 2005).

In a similar process to that described for the crop sector, FAO is preparing the first State of the World's Animal Genetic Resources, that is expected to be completed by 2006. So far, 151 countries have accepted to submit country reports and the objective of the country and global assessments is to provide a comprehensive analysis of the status and trends of the world's farm animal biodiversity and of their underlying causes, as well as of local knowledge regarding its management (Cardellino, 2005).

3.4 Fishery genetic resources

Like forestry, and unlike crops and livestock, wild populations currently play a greater role than domesticated populations in the fishery sector. When considering the status of the world's fishery genetic resources, both the aquatic species that are fished (capture fishing) as well as those that are farmed (aquaculture) can be included. Bartley (2005) gives an overview of the status of aquatic species that are fished or farmed worldwide.

For capture fisheries in 2002, FAO Members reported that 974 taxa (i.e. taxonomic units such as a family, genus or species) of fin-fish, 143 taxa of crustacea, 114 taxa of molluscs, 26 taxa of plants and 73 taxa of miscellaneous animals (such as sea urchins, sea cucumbers and marine mammals) were taken from the world's capture fisheries. These numbers are, however, likely to be an underestimate as reporting is often incomplete. Although over 1000 taxa are represented, about 10 species make up about one third of capture fisheries production. Unlike the crop and animal sectors, there is no systematic effort to describe the state of the world's fishery genetic resources below the species level. Nevertheless, description of the fish genetic stocks at the within-species level is important for fisheries management and is an active research area (Bartley, 2005; Primmer, 2005).

For aquaculture in 2002, FAO Members reported that, apart from a small number of aquatic plants and animals, a total of 153 species of fish, 60 species of molluscs and 44 species of crustaceans were farmed. The relative novelty of aquaculture can be shown by the fact that its contribution to global supplies of fish, crustaceans and molluscs has grown from just 4% in 1970 to 30% in 2002 (Bartley, 2005).

Aquatic science, in general, lags behind terrestrial sciences in identifying species, understanding ecosystem relationships and assessing potential uses for genetic resources. Communities of life on the ocean floor are poorly understood and, every year, some aquatic species become extinct before they have even been identified. Fish genetic resources worldwide are currently threatened by overfishing, but also by habitat destruction (e.g. dam building) and pollution from human activities (Greer and Harvey, 2004).

Every 2 years, FAO publishes the State of World Fisheries and Aquaculture which, inter alia, provides a global overview of capture fisheries (marine and inland) and aquaculture. Regarding exploitation of marine fish stocks (estimated production was 84.5 million tonnes in 2002), the most recent edition indicates that there has been a consistent downward trend since 1974 in the proportion of stocks offering potential for expansion, coupled with an increase in the proportion of overexploited and depleted stocks, rising from about 10% in the mid-1970s to close to 25% in the early 2000s. The percentage of stocks exploited at, or beyond, their maximum sustainable levels varies widely among fishing regions. The report says that the information available continues to confirm that, despite local differences, the global potential for marine capture fisheries has been reached, and more rigorous plans are needed to rebuild depleted stocks and prevent the decline of those being exploited at, or close to, their maximum potential (FAO, 2004a).

For inland fish stocks (estimated production was 8.7 million tonnes in 2002), unlike the major marine fish stocks, they are less well defined and occur over much smaller geographical areas, such as individual lakes, rice fields or rivers, or over vast areas such as transboundary watersheds that are often situated in areas that are difficult to access. These factors make it costly to monitor the exploitation and status of fish stocks and, in fact, very few countries can afford to do so. As a result, the majority of countries report only a small fraction of their catch of inland fisheries by species, further compounding the problem of accurate assessment. There are, however, indications that these resources are undervalued and threatened by habitat alteration, degradation and unsustainable fishing activities (FAO, 2004a).

4. Overview of Relevant Biotechnologies

As seen in Section 3, a wide range of genetic resources are currently being used by humankind for food and agricultural purposes and many of them are in danger of being lost, due to reasons such as over-exploitation (fish), replacement of local with international germplasm (crops/livestock) or habitat change or destruction (crop/fish/forest/livestock). Here, we will give a brief overview of molecular markers and cryopreservation, in vitro culture and reproductive technologies and how they can be used for the characterisation and/or conservation of these genetic resources.

4.1 Molecular markers

The importance of characterising genetic resources was described in Section 2 as well as the wide range of features that can be characterised in each population, such as morphology, phenotypic performance, degree of endangerment etc. Whereas phenotypes (e.g. yield, growth rate) or morphological traits (coat colour, seed shape) are influenced by both genetic and environmental factors, use of molecular markers reveals differences at the DNA level that are not influenced by the environment (De Vicente et al., 2005). Molecular markers usually do not have any biological effect and are normally assumed to represent neutral loci unaffected by selection. They are identifiable DNA sequences, found at specific chromosomal locations on the genome and transmitted by the standard laws of inheritance from one generation to the next. A number of types of molecular markers exist, such as RFLPs, RAPDs, AFLPs, microsatellites and SNPs, which can differ in a variety of ways, such as the amount of time, money and labour needed or the amount of genetic variation found at each marker in a given population. Microsatellites, for example, are simple DNA sequences, usually 2 or 3 DNA bases long, repeated a variable number of times in tandem. For more background information on molecular markers, see Section 2.1 of the Background Document to Conference 10 of this Forum on molecular marker assisted selection.

Molecular markers are used in a variety of approaches to characterise and conserve genetic resources, and here we will briefly describe the main ones. Some of the approaches are applied in each of the crop, forestry, animal and fishery sectors (e.g. estimating the genetic relationships between populations within a species) while others are more sector-specific (e.g. identifying duplicate accessions in crop genebanks or monitoring effective population sizes in capture fish populations).

4.1.1 Estimating the genetic relationships between populations within a species

This is one of the major uses of molecular markers in agricultural research. Molecular markers are typed in populations of the same species that are e.g. located in different places and, by comparing the frequencies and presence/absence of marker loci in the different populations, inferences can be made about how genetically related they are to each other. The more differences in markers found between the populations, the greater the genetic distance between them and the populations are inferred to be less genetically related to each other than to populations that have more similar marker loci frequencies. Taking a simple example for illustrative purposes (in practice many markers would be used simultaneously), if individuals in populations A, B and C are typed for a molecular marker M (with two alternative forms or alleles called M1 and M2) and all individuals typed in populations A and B have allele M1 while those in population C have M2, this suggests that A and B are genetically more similar to each other than to population C. In using molecular markers to estimate genetic distances between populations, it is assumed that they represent neutral loci (i.e. do not reflect genetic differences between populations at genes that are or have been under natural or artificial (farmer) selection) and that the use of a relatively small number of independently segregating marker loci will be a good predictor of the overall genomic diversity of a population (Hanotte and Jianlin, 2005).

Typically, a genetic diversity study involve sampling biological tissue (e.g. blood, leaves, fish scales) from a certain number of individuals from a number of populations of interest and typing them in the laboratory for a number of molecular marker loci. A statistical analysis of the marker data is then carried out to estimate the genetic relationships between the populations. For example, Lanteri and Barcaccia (2005) sampled 30 individuals each from 7 populations of wild cardoon identified in Sardinia and Sicily and each individual was genotyped for 32 microsatellite markers as well as 7 AFLP primer combinations. They used a statistical analysis method called principal coordinate analysis to show that the Sardinian and Sicilian populations were clearly differentiated, representing two distinct gene pools.

Such studies have been routinely applied in each of the agricultural sectors. For example, in developing a management strategy for commercially exploited fish, molecular markers are used to characterise the within-species genetic structure of the populations being harvested, to determine the units between which limited gene flow occurs i.e. to identify units which should not be overfished (Primmer, 2005). A recent questionnaire-based survey of animal diversity studies carried out during the last 10 years provided information on a total of 86 projects, involving 13 livestock species from 93 countries (FAO, 2004b). The majority of projects were for ruminants, especially cattle. Blood was the most frequently used biological sample and microsatellites were the most commonly used molecular markers.

As pointed out by Lenstra et al. (2005), these kinds of studies are not only relevant for the purposes of characterisation of genetic resources, they can also provide useful insights into the history of domestication. Evidence from molecular markers has been much used to elucidate whether individual crop or livestock species were domesticated just once at a single site or independently several times in different parts of the world (Diamond, 2002).

4.1.2 Establishing and managing genebanks

As described in Section 2, genebanks are an important tool for conservation of crop genetic resources, although currently of more limited importance for livestock, forestry or fish. Molecular marker information can be used in a number of genebank-related activities, such as sampling of material, management of the collections and stimulating use of germplasm stored therein (Lanteri and Barcaccia, 2005).

When sampling germplasm to create a genebank, information from the use of molecular markers to assess the within and between population genetic diversity (as described in 4.1.1) can be used in conjunction with data on other characteristics of the populations to assist decision-making regarding which material should be sampled (e.g. Simianer, 2005). The amount of germplasm to be stored may be limited, due e.g. to financial reasons, so it would be important to sample material from genetically diverse individuals.

Once a genebank has been established, molecular markers can also be used to assist in a number of genebank management activities. As a first example, regeneration of seeds or other reproductive plant material in storage is an important task in a genebank, and the regeneration requirements (how frequently accessions need to be regenerated) depend on factors such as the species concerned, the storage conditions and the quality of the individual accessions (FAO, 1998a, Chapter 3). Markers can be used to check whether changes in alleles or in allele frequencies are taking place over time as a result of the adopted conservation strategies e.g. of frequent regenerations (De Vicente et al., 2005). Secondly, while safety duplication of unique accessions acts as an insurance against possible accidents, unintentional duplication or overduplication of accessions is, however, wasteful (FAO, 1998a) and markers can be used to detect them (De Vicente et al., 2005). Thirdly, to reduce costs and/or encourage greater and more efficient use by farmers and breeders, core collections (i.e. a subset of accessions selected to contain the maximum available variation in a small number of accessions) have been established for many crop species. Again, molecular markers can assist in this task.

Finally, it is argued (e.g. Ferreira, 2005) that molecular markers offer opportunities to increase the use of crop genetic resources from genebanks and they will therefore stimulate increased use of genebank material for research purposes, thereby playing an important role in the sustainable conservation of these genetic resources. Since a major justification for the existence of germplasm collections is use of the conserved accessions, De Vicente et al. (2005) underline the importance of genetically characterising them so that valuable genes can be identified for use in development of new plant varieties by conventional and advanced breeding methodologies.

4.1.3 Gene flow from domesticated populations to wild relatives

Wild relatives of domesticated species are of conservation interest as they represent one of the components of agricultural biodiversity. They have also proven useful for genetic improvement of a range of cultivated varieties (FAO, 1998a, Chapter 1). As described in the Background Document to Conference 7 of this Forum (on GMOs and gene flow), crossing of domesticated populations with their wild relatives has been well documented in some crop, forest tree, animal and fish species. For example, in crops, gene flow has been observed between rice and perennial rice, between maize and teosinte and between sugar beet and wild beet, while in animals there is evidence for crossing of domestic cattle with wild North American bison and of domestic pigs with European wild boars. In forestry, gene flow from intensively bred forest trees to natural populations of the same or closely related species ("wild relatives") has been a cause for concern in many species, e.g. the European black poplar (Populus nigra). For fish, crossing of escaped farmed Atlantic salmon with wild Atlantic salmon is also a much-discussed problem.

Gene flow from a domesticated population will cause changes in the genetic diversity of wild populations. In cases where diversity is systematically reduced over time by such geneflow, it may potentially lead to genetic extinction of the wild population in its original genetic state (see e.g. Papa, 2005). Molecular markers can assist in conservation of the wild relatives through their use to distinguish hybrids from non-hybrid wild relatives and to monitor gene flow from domesticated populations to their wild relatives.

4.1.4 Estimating and monitoring the effective population size

When developing conservation strategies for wild species, accurate estimates of the effective population size (Ne) are important because Ne predicts a number of parameters such as the rate of inbreeding (Primmer, 2005), and increases in inbreeding (due to mating of genetically related individuals) can lead to the reduced abilities of individuals to survive and reproduce (a phenomenon called inbreeding depression). Ne can be estimated using pedigree information and/or knowledge about the population breeding structure (male/female mating ratio, family sizes etc.). However, for wild populations, especially of aquatic species, it can be difficult to get this information. Molecular markers can be used to estimate Ne, something which has been done, for example, in several commercial fish populations (Primmer, 2005). Similarly, molecular markers can also be used as a tool to monitor fish population sizes over time and to detect any major recent declines in Ne.

4.2 Cryopreservation, in vitro culture and reproductive technologies

As described in Section 2, there are two major strategies for conservation, in-situ and ex-situ, the latter involving e.g. storage of genetic material in genebanks or keeping live animals or plants in zoos or botanical gardens respectively. Biotechnology, through cryopreservation (freezing) technologies and in vitro culture, provides additional tools for ex-situ conservation.

Cryopreservation involves the preservation of germplasm in a dormant state by storage at ultra-low temperatures, usually in liquid nitrogen (-196C) and it can be used to preserve biological material (e.g. seeds, sperm, embryos) of crop, livestock, forest or fish populations indefinitely in genebanks. It allows large quantities of genetic material to be stored in a limited amount of space. If the living populations later become extinct, they can in many instances be regenerated using the frozen genetic material. In addition, if the living populations are declining in numbers and develop problems due to inbreeding, these problems can be alleviated by introducing unrelated genetic material from the genebank. In animals, collection of germplasm for cryopreservation and production of live offspring from cryopreserved genetic material depends on a number of associated reproductive technologies, such as embryo transfer. In vitro culture is a plant biotechnology that involves application of slow growth procedures to germplasm accessions kept as sterile plant tissues or plantlets on nutrient gels.

Here, we will provide a brief overview of the current status of these biotechnologies, focusing on cryopreservation as it is applicable in all four sectors. Information provided on cryopreservation in livestock/fish and crops/forest trees is based primarily on Hiemstra et al. (2005) and Panis and Lambardi (2005), and the references therein, respectively.

4.2.1 Livestock and fish

Before considering the main types of genetic material that have been cryopreserved, a brief overview of relevant terminology might be useful. In sexual reproduction, through fertilisation there is a fusion of 2 haploid gametes to form the diploid zygote, a cell which begins dividing and develops to become an embryo (note, haploid and diploid cells have one or two sets of chromosomes respectively). The female gamete is called the ovum or egg, and is derived from a cell called an oocyte that has undergone two meiotic divisions. The male gamete is called the sperm (contained in fluid called semen).

For cryopreservation purposes, semen is currently the biological material of greatest importance, although embryos and, to a lesser degree, oocytes and somatic cells, have also received considerable research attention. Success in freezing these materials varies considerably from species to species, which is not surprising given the wide diversity in reproductive systems (e.g. fertilisation is usually external in fish and internal in birds, where the female stores the sperm before fertilisation). Collection of biological materials for the genebank and production of live offspring following thawing of formerly cryopreserved material is made possible through a variety of reproductive technologies, such as artificial insemination (AI) using semen, multiple ovulation and embryo transfer using embryos etc.

Freezing of semen and subsequent successful AI with thawed semen was first achieved over 50 years ago and semen of most livestock species can now be frozen adequately. There are large differences between species in insemination techniques and pregnancy rates using fresh or frozen semen, where AI with frozen semen is most successful in cattle. For domesticated birds, although AI is less widely used than in mammals, reasonable success rates (albeit at times highly variable) with frozen semen have been reported for the major bird species. For fish, sperm cryopreservation has been tested in over 200 fish species with external fertilisation and "the present state of the art for many species of fish seems to be adequate for the purpose of gene banking" (Hiemstra et al., 2005).

For embryos, cryopreservation in cattle is now a routine procedure. It has also been reported for other mammalian species, such as horse, pig, rabbit and sheep (ERFP, 2003). Embryos to be cryopreserved can be collected using non-surgical (cattle and horses) or surgical methods (other mammalian species). Inducing multiple ovulation using hormones can increase efficiency of the process. Alternatively, embryos can be collected through in vitro maturation and in vitro fertilisation of oocytes (a more technically demanding approach). Production of live offspring following cryopreservation involves non-surgical (only in cattle and horses and, to a lesser extent, pigs) or surgical embryo transfer to recipient animals. For bird and fish species, however, cryopreservation of embryos has not been successful, mainly because of the large size, the high lipid content and the polar organisation of the ova and early embryos.

For mammalian oocytes, live offspring have been reported in cattle and horse from embryos produced using cryopreserved oocytes, although the efficiency and reliability of using oocytes for generating offspring is still much lower today than with cryopreserved embryos. Oocytes for cryopreservation in the genebank can be collected from ovaries of dead animals or by ovum pick-up (recovery of oocytes using an ultrasound probe) on live animals. Production of live offspring following thawing of cryopreserved oocytes involves in vitro maturation and fertilisation and eventual embryo transfer to recipient animals. Cryopreservation of bird and fish oocytes has not yet been successful, mainly for the same reasons given earlier for embryos.

For somatic cells (i.e. cells not involved in sexual reproduction), a number of cell types (e.g. skin fibroblasts, mammary cells) can now be cryopreserved. Collection of material for the genebank is easy and cheap. Production of live offspring (which, apart from mitochondrial DNA, are genetic clones of the individual which donated the somatic cells) is, however, complicated and has a low rate of success. It involves culturing the thawed cells and then transferring their nuclei (i.e. the organelle which contains the genetic material) to, or fusion of the somatic cells with, enucleated (i.e. without a nucleus) oocytes from another individual. The resulting embryos are then cultured and eventually transferred to recipient animals. In mammals, live offspring have been obtained from embryos generated from somatic cells in a number of mammalian species (cattle, goats, horse, pigs, rabbits and sheep) and in fish (zebrafish), but not yet in poultry. The low success rates, as well as the many developmental problems encountered, are current major limitations to application of the technology.

4.2.2 Crops and forest trees

As described in Section 2, about 90% of the 6 million plant accessions in genebanks, mainly crops, are stored in seed genebanks. However, storage of seeds is not an option for crops or trees that do not produce seed (e.g. bananas) or that produce recalcitrant or non-orthodox seed (i.e. seed that does not survive under cold storage and/or the drying conditions used in conventional ex situ conservation, e.g. mango, coffee, oak and several tropical forest tree species). Likewise, for plant species that are propagated vegetatively to preserve the unique genomic constitution of cultivars, such as fruit and several timber and ornamental trees. It also presents practical problems in most forest tree species in regard to the necessary, periodic regeneration of seedlots, as tree species usually have a long vegetative period prior to producing flowers and seed (lasting from several years to several decades), and as they require large areas to carry out such an operation, due to outcrossing and their size.

Genetic resources of these species are normally conserved either in vivo, i.e. living collections, or by in vitro culture, where slow growth procedures allow the plant material to be held for 1-15 years under tissue culture conditions with periodic regeneration (subculturing), depending on the species. Normally, growth is limited using low temperatures often in combination with low light intensity or even darkness. Temperatures in the range of 0-5C are employed for cold tolerant species and 15-20C for tropical species. Growth can also be limited by modifying the culture medium and reduction of oxygen levels available to the cultures (Rao, 2004). The former option (in vivo) is the cheapest and thus the most widely adopted, but exposes the material to the harms of the external environment, e.g. contamination with pathogens or parasites. In vitro culture, on the other hand, ensures that the collections are kept free from pest and diseases and protected from adverse conditions. Maintenance of in vitro collections is, however, labour-intensive and there is still a risk of losing accessions due to contamination, human error or somaclonal variation, i.e. mutations that occur spontaneously in tissue culture, with a frequency that increases with repeated subculturing (Panis and Lambardi, 2005).

In all these situations, as well as for long-term storage of seeds from orthodox species, cryopreservation offers an alternative tool or complementary strategy for ex-situ conservation. Following plant cell, tissue or organ storage at low temperatures, plants can then be regenerated and there seems to be no clear evidence so far of morphological, cytological or genetic alterations arising as a result of cryopreservation (Panis and Lambardi, 2005). They, however, conclude that, despite the fact that cryogenic procedures are now being developed for an increasing number of recalcitrant seeds and a wide range of tissues and organs, the routine utilisation of cryopreservation for the preservation of plant biodiversity is still limited. They note that the main drawback for a wider application of plant cryopreservation is the unavailability of efficient cryopreservation protocols for many plant species.

They report, nevertheless, that for various herbaceous (i.e. non-woody plants), hardwood (i.e. broadleaf, deciduous trees) and softwood species (i.e. coniferous trees), cryopreservation of a wide range of tissues and organs such as cell suspensions, embryogenic cultures, pollen, meristematic tissues and seeds has been achieved. Cell suspensions (cells in culture in moving or shaking liquid medium) have been cryopreserved from banana and grape while cryopreservation of embryogenic cultures (i.e. callus or suspension cultures with potential to differentiate somatic embryos) is an advanced technology in conifers and is already being successfully applied to numerous commercial species, including pines. Methods exist for the cryopreservation of pollen from many crops, but its application is still rather limited and restricted to a few research centres. Meristematic tissues (also called meristems, i.e. plant tissues, such as shoot tips, in which the cells are capable of active division and differentiation into specialised tissues such as shoots and roots) are the most common explants (i.e. a portion of a plant aseptically excised and prepared for culture in a nutrient medium) for the cryopreservation of vegetatively propagated species, such as fruit trees