DM TEVERSON¹, JD TAYLOR², CS MUSHI³, SZ BISANDA⁴ and JM LENNE¹

¹Natural Resources Institute, Chatham Maritime, Kent ME4 4TB, UK; ²Horticulture Research International, Wellesbourne, Warwick CV35 9EF, UK; ³Selian Agriculture Research Institute, Arusha, Tanzania; ⁴Ministry of Agriculture Research and Training Institute, Mbeya, Tanzania

Background and objectives
Most studies of biodiversity start with the assumption that biodiversity is good, and is maximized in populations of wild plants and primitive domestic crops. Such populations often show considerable variation in both morphological characters and their response to biotic and abiotic stress. Crop plants lacking the uniformity associated with modern cultivars are usually referred to as landraces. Landraces appear to provide sustainable yields under a range of biotic and abiotic stresses, and are typically employed in low-input subsistence agricultural systems.

Wild relatives and landraces of crop species have long been recognised as important sources of useful genes for breeding improved crop cuitivars. Concern over the apparent loss of wild plants and landraces has been the driving force behind the conservation of genetic resources. However, comprehensive studies of useful characteristics in such populations are often lacking, and the justification for random collecting and preserving germplasm in ex situ collections is not always clear. The aim of the present study was to dissect and quantify biodiversity in traditional Tanzanian bean (Phaseolus vulgaris) landraces in respect of resistance to the major biotic stress: plant disease.

P. vulgaris is subject to a wide range of diseases caused by bacterial, fungal and viral pathogens. For this study we identified five priority pathogens: angular leaf spot (Phaeoisariopsis griseola), anthracnose (Colletotrichum lindemuthianum), common bacterial blight (Xanthomonas campestris pv. phaseoli), halo blight (Pseudomonas syringae pv. phaseolicola), and bean common mosaic virus (BCMV), all of which are widely distributed throughout the bean-growing areas of Africa and worldwide. A common feature of these pathogens is the existence of races/strains, which show pathogenic variation in respect of the host species and cultivars they attack. The existence of races (pathogenic variants) is a major limiting factor in breeding for resistance.

Since African bean landraces survive despite the 'massed ranks' of pathogens and all their variants, and without pesticide inputs, the prime objective of this study was to examine landrace diversity for disease resistance and to address the questions: do landraces represent unique collections of resistance genes (lacking in most modern cultivars) and if so, is a greater effort to conserve landraces necessary and how might this be achieved?

Results and conclusions
From a 6-year study of disease resistance in components and subcomponents of landrace mixtures from the southern highlands of Tanzania, there is clear evidence of the value of landraces both as an ongoing method of disease control and as a source of important and possibly unique resistance genes to a range of pathogens of international importance. Currently landrace mixtures constitute the predominant bean production method in this region, although under some threat from the increasing use of pure varieties. Local knowledge of the characteristics of individual landraces is an integral part of the deployment and maintenance of landraces. In effect, the local farmers are already practising a form of in situ conservation. The findings of the present study provide the scientific raison d'etre for the existence of landraces. This knowledge, in combination with existing local knowledge, may provide the impetus for active in situ conservation as the ideal long-term solution.