A general method of quantitatively assessing genetic similarity among a set of cultivars of a given crop is proposed, and its application to dry beans in the United States is demonstrated. The method is based upon the multi-variate technique of Principal Components Analysis. Using this method it was possible to calculate a ‘distance’ metric between any two cultivars in the set and to show that such distances were highly inversely correlated with genetic relationship estimated from a knowledge of breeding ancestry.
On the basis of distances among cultivars within given production regions (states in the US in this case) and knowledge of the acres of each cultivar grown in the region, an average weighted distance metric appropriate to each was calculated. Each derived distance metric serves as an index of ‘genetic homogeneity’ for the crop in that region. Arguments are presented for relating the degree of vulnerability to a disease epidemic to the distance index. Indexes are calculated for nine of the major bean producing states in the US from which it is concluded that, from the standpoint of genetic vulnerability, Colorado is most vulnerable and California least vulnerable to a region-wide epidemic affecting the bean crop. It is suggested that the method demonstrated here is of nearly universal applicability, and particularly meaningful with respect to self-fertilizing crops.