Modelling the origins of legume domestication and cultivation

Ladizinsky's (1987) mathematical model of the domestication of lentils and other Near Eastern legumes is invalid. Ladizinsky believed that the model predicts that high rates of seed harvesting would lead to fixation, in wild populations, of genes conferring lack of seed dormancy (a domesticated trait). In fact, however, the model gives rapid fixation of alleles for non-dormancy under all circumstances; gathering by humans has no effect on allelic frequencies. It appears that in these species cultivation must have preceded the morphological changes that distinguish domesticated from wild plants. The addition of more realistic assumptions to the model does not alter this conclusion. I suggest several scenarios that could explain Near Eastern legume domestication: the most plausible of these postulates that cultivation of cereals led to scheduling conflicts which necessitated the abandonment of harvesting of wild legumes, and hence the initiation of their cultivation. Mathematical modelling may be able to contribute to our understanding of agricultural origins, but it must be carried out with greater rigor and closer attention to the theoretical literature.