Although the concept of ‘sustainable development’ has been hailed for its reconciliatory poten- tial, it has failed in practice to resolve enduring conflicts. Exploring the evolution of the concept—from its 19th-century antecedents through Brundtland to contemporary ‘Panglossian’ interpretations—the paper argues that difficulties of implementation are not transient but have deep roots. No conception of sustainable development can be adopted without making fundamental ethical and political choices, but the debate about such choices is nevertheless of great importance. It is argued that we should abandon our search for a singular, consensual definition of sustainability, but try as best we can to make progress in the absence of consensus.

UK agriculture is undergoing significant change with reduced subsidies for food production, increasing consumer demands for food safety and traceability, and environmental concerns including climate and demographic change. The International Treaty on Plant Genetic Resources for Food and Agriculture adopted by the United Nations Food and Agriculture Organisation supports the use of genetic resources for research and breeding. Mining genetic resources for useful genetic variation is perceived as a major benefit of genebanks. However, utilization by breeders may be constrained by poor characterization of genetic resources, a widening gap between improved and unimproved material, and the disruption of well- adapted genotypes during introgression. Breeders working with grasses and forage legumes for sustainable agriculture are fortunate in the wealth of genetic variation available both within the primary species of interest and among related species. New DNA technologies allow more targeted approaches to the use of these genetic resources. Possibilities for gene transfer between related species using conventional techniques expand the available gene pools while potential use of genetic transformation extend these even further.

Plant genetic resources are conserved so that they can be used to improve crop plant pro- duction and in other ways. However, it is often asserted that use of ex situ conserved germplasm is inadequate and that genetic diversity maintained in genebanks is underutilized. In part, this reflects an incomplete recognition of what constitutes use of plant genetic resources, and of the many different ways in which material from genebanks contributes to improved agricultural production. Based on recent information from surveys of distribution of germplasm from genebanks, and from surveys of users, we suggest that the evidence indicates that there is substantial use of ex situ conserved materials for a wide range of different uses. We suggest that barriers to use of ex situ conserved germplasm may often result from a lack in numbers of users, and from limitations in capacity to effectively utilize the genetic diversity present in genebanks to reduce genetic vulnerability and increase sustainability in modern production systems.

Farmers in local communities have a considerable amount of information on their genetic resources both at the individual and community level. The farmers' knowledge is interweaved into their daily practices as a reserve accumulated over the generations. The Farmer Field Fora approach (FFF) has been conceptualized in order to activate farmers to identify the existing resources both in the crops they cultivate and in their knowledge, to recognize its value and to start using it more consciously. The FFF approach links the use of existing diversity to the possibility of having new varieties and to improving productivity. In weekly meetings farmers observe changes in the field between varieties, debate on them and draw conclusions. In every session a theme on plant development is debated among farmers with the aid of a facilitator. Close contact with researchers allows involvement with the breeding practices and enables the aims and conduct used in research to be understood. This empowers farmers to participate in a more active way in research, to apply more precise ways of observing and managing their genetic resources, and to link their work to conservation activities at the national and local level. The results of the FFF trial, which was carried out in 10 villages in the region of San in Mali in 2001, show an increased interest among farmers to observe varieties. The appreciation criteria used is multiple including yield, earliness, strong tillers, high number of tillers, and resistance for diseases and Striga. It seems obvious that even though the goal of the approach is on-farm conservation of plant genetic resources, it must be closely linked to farmers' interest. The variation of the results between villages demonstrates variability of climatic condition in this region and a high genotype—environment interaction. Further modifications of the approach are required in order to improve equitable participation of different stakeholders and improve the institutionalization of on-farm conservation.

Genealogical data have been used very widely to construct indices with which to examine the contribution of plant breeding programmes to the maintenance and enhancement of genetic resources. In this paper we use such indices to examine changes in the genetic diversity of the winter wheat crop in England and Wales between 1923 and 1995. We find that, except for one period characterized by the dominance of imported varieties, the genetic diversity of the winter wheat crop has been remarkably stable. This agrees with many studies of plant breeding programmes elsewhere. However, underlying the stability of the winter wheat crop is accelerating varietal turnover without any significant diversification of the genetic resources used. Moreover, the changes we observe are more directly attributable to changes in the varietal shares of the area under winter wheat than to the genealogical relationship between the varieties sown. We argue, therefore, that while genealogical indices reflect how well plant breeders have retained and exploited the resources with which they started, these indices suffer from a critical limitation. They do not reflect the proportion of the available range of genetic resources which has been effectively utilized in the breeding programme: complex crosses of a given set of varieties can yield high indices, and yet disguise the loss (or non-utilization) of a large proportion of the available genetic diversity.

Oilseed rape is a very undeveloped crop with regard to efficiency of production and the agronomic practice used to maximize its potential. The genetic potential to modify oilseed rape is limited by the narrow genetic base found within the breeding gene pool, resulting in limited novel variation available for exploitation. Novel variation is, however, present in wild diploid ancestors of oilseed rape and has been made available by developing synthetic Brassica napus. This is illustrated through the use of this material to develop an understanding of pod shattering which is one of the most agronomically important characteristics of the crop. Through a variety of approaches it is shown how progress has been made to understand this trait and how this understanding is being used to improve the crop such that efficiency of production will be enhanced.

The agri-climatic adaptation of an alfalfa variety in North America is categorized by its Fall Dormancy Rating (FDR1 = dormant to FDR9 = non-dormant). Presently, only relatively dormant, winter-hardy varieties (FDR1–4) are grown for seed and herbage at the northern latitudes of western Canada. However, there is considerably greater demand for seed of varieties that have FDR ≥ 4. A study was conducted in the Peace River region of north-western Canada to determine the relationship between FDR and seed production of alfalfa with FDR ≥ 4. Trials were established in each of two consecutive years with four varieties representing each of six FDR categories, FDR4–9. Growth characteristics were determined for two successive years of seed crops for each year of establishment. Seed yield (as a percentage of that for FDR4) was 84, 52, 40, 29 and 39% for FDR5 to FDR9, respectively, whereas total biomass at seed maturity was 89,73, 73, 57 and 57% for FDR5 to FDR9, respectively. Seed yield was more sensitive to changes in FDR than total biomass. Short rotations of specific alfalfa varieties with FDR ≥ 4 could be an option for seed growers in the northerly latitudes of western Canada, provided greater financial compensation is given for higher FDR varieties, and provided genetic drift can be retained within acceptable limits.

The perception by arable farmers in the UK of research organizations and companies acting as sources of information relating to plant genetic resource utilization is represented using brand strength plots. In development situations there exist considerable difficulties in achieving technology transfer. The method of brand analysis applied in this paper could be extended to characterize options in the selection of vectors for technology transfer in development programmes. A survey of potential vectors such as television, radio, extension services, univer- sities, government agencies and commercial companies would reveal their comparative brand strength and hence utility for achieving development objectives.