1. A definition of genetic erosion
The level and structure of genetic diversity in plant species � whether wild or cultivated � is shaped by the five evolutionary forces of mutation, recombination, migration, genetic drift and selection (natural and artificial). Apart from mutation, these are in turn affected by the interaction of the plant with humans and its environment (biotic and physical) and by the reproductive biology of the species, through the intermediacy of the differential survival and isolation of individuals and populations.
Genetic diversity is always changing, but the Report on the State of the World�s Plant Genetic Resources (FAO, 1996), summarizing country reports, suggests that "recent losses of diversity have been large, and that the process of �erosion� continues." It points out that while loss of genes is of particular concern, loss of gene complexes and unique combinations of genes (as in different landraces) can also have important consequences. Genetic erosion may thus be defined as a permanent reduction in richness or evenness of common localized alleles or the loss of combination of alleles over time in a defined area. This definition recognizes that diversity has two distinct components in (i) the number of different entities and (ii) their relative frequencies. It also suggests that it is specifically loss of locally adapted alleles that is most significant. Genetic erosion will be detrimental to the short-term viability of individuals and populations, the evolutionary potential of populations and species, and the direct use of genetic resources (Brown et al., 1997). Recent genetic erosion and/or the risk of imminent genetic erosion are key factors in determining the priority given to different areas for conservation interventions � whether ex situ, in situ or a combination of both.
2. Indicators of genetic erosion
Because of its potential impact, the Global Plan of Action includes a priority activity on "developing a monitoring and early warning system for loss of plant genetic resources for food and agriculture." Brown et al. (1997) provide a useful list of features or indicators that could be measured singly or in combinations on individuals and populations of a given species in a defined area as part of a systematic effort to monitor changes in genetic diversity in the species.
- Number of sub-specific entities. Formal taxonomic categories such as sub-species and also entities such as ecotypes, chromosome races and landraces groups "are a useful first approaximations of genetic diversity within a species."
- Population size, numbers and isolation. Small populations are at relatively greater risk of loss of alleles, increased inbreeding and extinction due to stochastic events. The number and isolation of populations in an area will reflect both the overall genetic diversity in the area and how this is structured.
- Environmental amplitude. The number of distinct habitats or environments in which the species is found in a study area (for example based on existing ecological and climatic classifications) reflects highly adaptive variation.
- Genetic diversity at marker loci. In the past few years, advances in molecular biology have resulted in the development of a number of powerful new techniques that have found important applications as diagnostic tools for investigating genetic variation in plants and animals. These tools, the most commonly used of which are RFLP (Restriction Fragment Length Polymorphism) and RAPD (Randomly Amplified Polymorphic DNA) analysis, complement longer-established techniques such as isozyme analysis and will be supplemented in future by other techniques now being developed, such as AFLP (Amplified Fragment Length Polymorphism) analysis and various microsatellite approaches. They provide a wide range of molecular approaches for the study of important biological topics in the field of conservation and sustainable use of plant and animal genetic resources, including the amount and structure of genetic diversity within species, and how this changes in time.
- Quantitative genetic variation. Additive genetic variance of metric characters within populations reflects variation at multiple loci and is a measure of the "ability of a lineage to � adapt to changing � conditions."
- Inter-population genetic structure. Markers and quantitative measures could be used to gauge not only the diversity within populations, but also the level of genetic differentiation among populations, which is an important component of overall genetic diversity in an area.
- Amount and pattern of mating. This is the primary determinant of partitioning of genetic variation, with changes affecting individual fitness (e.g. inbreeding depression) and the viability of population. Mating events can respond rapidly to population changes. The different parameters that could be monitored include outcrossing rate, fixation index, fecundity and progeny fitness, pollinator abundance, sex ratios, and density of reproductive individuals.
3. Estimating past genetic erosion
The characteristics listed above can be used not only to provide a baseline against which to assess future genetic erosion in a particular area, but also, in specific circumstances, to estimate the genetic erosion that has occurred in the past in an area. The State of the World report states that "because no one can say exactly how much diversity once existed �, no one can say exactly how much has been lost historically." However, there are circumstances in which it may be possible to overcome this problem. Two approaches may be recognized in the study of past genetic erosion, i.e. temporal and spatial comparisons.
Temporal comparison. In genetic erosion studies involving temporal comparisons, the same population(s) are studied at different times. This may be done by re-sampling, by historical comparison or through indigenous knowledge (IK) surveys.
- Re-sampling involves the direct comparison of samples collected at different times. The samples could be germplasm accessions conserved ex situ or dried plant material stored in a herbarium. The basis of the comparison could be local knowledge, conventional morphological characterization, agronomic evaluation, or molecular markers. Problems associated with this approach include the availability and quality of passport data (necessary fo exact site identification), the comparability of collecting strategies at different sampling times and the possibility of genetic erosion having occurred ex situ. Re-sampling studies have been carried out for wild potato species in Arizona and New Mexico (del Rio et al., 1997) and various crops in Albania and southern Italy (Hammer et al., 1996). All of the features listed in the previous section could be documented in such studies; in the two examples quoted above molecular markers and sub-specific entities (landraces) were used respectively.
- By historical comparison is meant here comparison of observations of the current situation with published observations on the situation at different time(s) in the past, where actual specimens from earlier times are not available. Because specimens are not available, not all the features mentioned in the list in the previous section will be relevant for such studies (e.g., molecular markers), but it may be possible to compare the numbers of sub-specific entities such as landraces and population numbers and sizes, for example. Sources of the historical information could be experts, the formal literature (both professional agricultural, botanical, ethnographic, linguistic, development, etc. and "amateur," e.g. explorers and travelers), gray literature (e.g. reports of extension departments, field books of collectors) and remote sensing. An interesting example is provided by Varisco (1985), who compared his observations of sorghum landraces in Yemen with medieval accounts. Many quoted examples of genetic erosion involving numbers of varieties at different times in the past, for example those mentioned in many of the country reports and summarized in the State of the World report, are derived from this kind of study.
- The oral history, knowledge and experiences of local people are potentially rich sources of information on genetic erosion at the variety level. IK surveys of genetic erosion consist of asking local people about what genetic diversity they may have lost. However, IK is unevenly distributed within communities, often protected, fragile, error-prone in transmission, and location-specific, which means that methodological sophistication is necessary for its documentation and interpretation. Some of the techniques of Participatory Rural Appraisal (e.g., time lines, trend lines and historical transects), combined with ethnobotanical methods (e.g., for the elicitation of folk taxonomies) can be adapted to reveal changes in time in the crop varieties being used in a village or the abundance of specific wild species.
Spatial comparison. In contrast to the kinds of temporal comparisons described above, genetic erosion studies based on spatial comparisons involve different population(s) being investigated at the same time for any of the characteristics listed in the previous section. In essence, in these studies space is taken as a surrogate for time. Two main types of spatial comparisons may be recognized, i.e. intensive and extensive.
- In intensive spatial comparison studies, a small number of nearby populations are compared which differ in a key factor but are otherwise similar. Examples of such "controlled comparisons", typically of a disturbed versus an undisturbed situation, include on-going IPGRI-sponsored studies of the effects of fragmentation and selective logging on genetic diversity of forestry species in Costa Rica, Vietnam and Brazil and Brush�s (1992) work on cultivated potatoes in two Peruvian valleys of contrasting levels of modernization and commercialization.
- In contrast, extensive studies are typically country- or region-wide and involve comparing many areas in terms of several factors at once. Geographic Information System (GIS) technology is a useful tool in such studies, as exemplified by on-going work by IPGRI and the International Potato Centre (CIP) on cultivated potatoes in Peru. In this study, isozyme data from the CIP ex situ potato collection are being used to estimate levels of genetic variation in each 20-20 km square in a grid covering the whole potato-growing area of the country. Data on the extent of potato cultivation and the environment in each square are used as the independent variables in a multiple regression model, with genetic diversity as the dependent variable. Negative deviations from the model are then mapped to identify possible area of genetic erosion, and correlations sought with a number of socio-economic variables (such as accessibility and population growth) to identify possible causative factors for any lower-than-expected level of genetic diversity.
4. Assessing the risk of future genetic erosion
Agenda 21 (chapter 15) states that "the current decline in biodiversity is largely the result of human activity and represents a serious threat to human development." Various attempts have been made to list the threats faced by plant diversity, both wild and cultivated, at the local, national and global level. These include WRI, IUCN and UNEP (1992), WCMC (1992), Gomez-Campo et al. (1992), UNEP (1993) and Dahl and Nabhan (1992). The country reports which provided the raw data for the State of the World�s Plant Genetic Resources also listed a number of putative cause of genetic erosion. These included the introduction of new varieties of crops and civil strife for cultivated species and deforestation, overgrazing and urbanization for wild species. Such lists "can be used as an evaluation tool for any local community wishing to impede genetic erosion" (Dahl and Nabhan, 1992) as well as to assess the danger of erosion taking place in the future.
Monitoring various putative causative factors is clearly one possible approach to assessing the risk of future genetic erosion within a genepool in a given area. However, the relationship between such factors and genetic erosion may not be straight-forward � it may well be non-linear, site-specific and involve complex interactions among factors. As Brush (1995) points out, there are also forces which counteract those tending to cause genetic erosion. This approach therefore needs to be complemented with direct measurement of genetic diversity if causative links are to be established between specific factors and genetic erosion.
Once the past association between genetic erosion and different causative and countervailing factors(s) has been investigated in temporal and/or spatial comparisons, a predictive model may be constructed based on the assumption that the association will continue into the future. Thus, temporal comparisons at a number of different sites and/or a spatial comparison study could suggest that a particular factor might be responsible for genetic erosion in a particular genepool. The current presence or strength of that factor could then be mapped over the region in question to identify those areas where it might be expected to cause further significant genetic erosion in the future.
5. Monitoring genetic erosion by national programmes
On the basis of the foregoing, a framework can perhaps be suggested for the organization of monitoring systems for genetic erosion by national programmes. Such a system might consist of the following components:
- Studies of past genetic erosion in priority genepools, i.e. spatial and temporal comparisons, using a number of the indicators listed in Section 2.
- Identification of major factor(s) contributing to genetic erosion in the target genepool, according to the above studies.
- Mapping of the strength of the identified putative causative factor(s) or some other closely correlated feature over the mandate region
- Fieldwork in areas identified at high risk of genetic erosion to gather baseline genetic diversity data (again using a number of indicators), verify level of risk and plan possible conservation interventions in partnership with affected local communities.
- Setting up a network of community-based correspondents for continuous monitoring of genetic diversity and genetic erosion risk.
References
Brown, A., A. Young, J. Burdon, L. Christides, G. Clarke, D. Coates and W. Sherwin (1997) Genetic indicators for state of the environment reporting. Australia: State of the Environment Technical Paper Series (Environmental Indicators), Department of Environment, Sport and Territories, Canberra.
Brush, S.B. (1992) Ethnoecology, biodiversioty and modernization in Andean potato agriculture. J. Ethnobiology 12:161-185.
Brush, S.B. (1995). In situ conservation of land races in centers of crop diversity. Crop Science, 35, 346-354.
Dahl, K. and G.P. Nabhan (1992) Conservation of Plant Genetic Resources. Grassroots Efforts in North America. ACTS Press, Nairobi.
Hammer, K., H. Knupffer, L. Xhuveli and P. Perrino (1996) Estimating genetic erosion in landraces � two case studies. Genetic resources and Crop Evolution 43:329-336.
Gomez-Campo, C. and collaborators (1992) Libro Rojo de Especies Vegetales Amenazadas de España Peninsular e Islas Balneares. Ministerio de Agricultura y Alimentacion, Madrid.
Rio, A.H. del, J.B. Bamberg, Z. Huaman, A. Salas and S.E. Vega (1997) Assessing changes in the genetic diversioty of potato gene banks. 2. In situ vs ex situ. Theoretical and Applied Genetics 95:199-204.
UNEP (1993) Guidelines for Country Studies on Biological Diversity. UNEP, Nairobi.
Varisco, D.M. (1985) The production of sorghum (dhurah) in highland Yemen. Arab Studies 7:53-88.
WCMC (1992) Global Diversity: Status of the Earth's Living Resources. Chapman and Hall, London.
WRI, IUCN and UNEP (1992) Global Biodiversity Strategy. WRI, Washington DC.
