Most of the genetic diversity of North American alfalfa cultivars has been accounted for by nine germplasm sources listed in descending order of winter hardiness as follows: Medicago sativa ssp. falcata, Ladak, M. sativa ssp. xvaria, Turkistan, Flemish, Chilean, Peruvian, Indian and African. In most instances, the breeder assigns a fall dormancy score and the relative proportions of each of the nine germplasm source for each cultivar at registration. The fall dormancy score (1 = dormant to 9 = non-dormant), determined by measuring plant height in October after harvest in early September, is used to indicate cultivar adaptation for different regions. This study examines the relationship between germplasm composition and plant height, the equivalence of fall dormancy. The signs on the partial regression coefficients of a multiple regression analysis of plant height on the proportional content of the nine sources of germplasm showed that the fall dormancy fell essentially into two classes, namely, a dormant category, comprising cultivars containing a large contribution of Falcata and Ladak, and a non-dormant category, in which Indian and African germplasm predominate. This does not necessarily preclude the influence of any of the other germplasm sources on fall dormancy, since they represent a rich source of diversity. However, nine distinct classes were not recognisable, perhaps because of the lack of an exact equivalence between fall dormancy class and plant height of the fall regrowth. Since these observations have not been derived in a common nursery, the latitude and latitude × cultivar effects have been disregarded. These limitations should be recognized when using the currently assigned fall dormancy ratings to predict cultivar adaptation. Key words: Alfalfa, fall dormancy, sources of germplasm
Spatial pattern and genetic diversity estimates are linked in stochastic models of population differentiation
