An approach to the optimal allocation of conservation funds to minimize loss of genetic diversity between livestock breeds

About 30% of all farm animal breeds worldwide are at risk of extinction. To prevent this irreversible erosion of genetic diversity, the limited funds available for conservation need to be allocated in the most efficient way. Applying the diversity concept of Weitzman [Quart. J. Econ. CVII (1992) 363; Quart. J. Econ. CVIII (1993) 157] this paper presents a framework for the allocation of a given budget among a set of breeds such that the expected amount of between-breed diversity conserved is maximized. As a novel methodological contribution, a functional relationship between conservation funds spent in one population and the conservation effect in terms of reduced extinction probability is suggested. Based on arguments from population genetics, three different functions are derived, which may reflect the range of possible functions in typical conservation situations. The methodology is illustrated with an example of 23 African zebu and zenga cattle breeds. The results indicate that conservation funds should be spent on only three to nine of the 23 breeds, depending on the model used. Highest priority is given to breeds, for which the 'conservation potential', that is, the product of extinction probability and marginal diversity is maximum, and these are not necessarily the most endangered breeds. The methodology can be extended to the maximization of total utility, which incorporates diversity, as well as other direct use, and special value, characteristics. However, a number of essential input parameters such as extinction probabilities and economic values are lacking and realistic models for developing cost-efficient conservation strategies have to be derived. Given these lacking bits of information become available, the methodology suggested provides a breakthrough towards applicability of diversity-based approaches for decision taking in conservation programs. (C) 2003 Elsevier Science B.V. All rights reserved.