| The variation contained in cultivated plants and their wild ancestors provides the raw genetic material required to evolve in response to changing environmental conditions, pathogens, and human selection pressures. In crop species, contemporary patterns of genetic variation reflect historical processes associated with the evolutionary process of domestication such as selection, genetic bottlenecks, dispersal, and gene flow with wild ancestors. This study documents the origin and domestication history of a Mesoamerican fruit tree, jocote (Spondias purpurea L., Anacardiaceae). Jocotes are cultivated for their edible fruits, which are eaten fresh, stewed, or used to make jams and beverages. They are propagated vegetatively and grown in living fence rows, backyard gardens, and orchards throughout Mexico, Central America, and portions of South America and the Caribbean. Wild populations of S. purpurea are found in the dry forests of Mesoamerica.;The origins of the S. purpurea lineage and the relationship of S. purpurea and two sympatric congeners were investigated using DNA sequences from the nucleus and the chloroplast. Alleles recovered in S. purpurea trees form a unique lineage, although there is limited evidence for lineage sorting or hybridization among Spondias species in a restricted region of southern Mesoamerica. Phylogeographic analyses of chloroplast alleles suggest that cultivated S. purpurea trees were derived from wild populations in at least two distinct regions within Mesoamerica, western Central Mexico and Central America. The amount and structure of genetic diversity in wild populations, and cultivated populations (including living fences, backyard trees, and orchards) was quantified using amplified fragment length polymorphism (AFLP) data. Genetic diversity estimates for cultivated populations are significantly different from the levels of variation detected in wild populations. Wilcoxon two-group tests reveal that while genetic diversity in orchard populations differs significantly from the variation found in wild populations, backyard populations differ with only marginal significance, and levels of variation in living fence populations are statistically indistinguishable from the diversity of wild populations. Finally, genetic variation is structured differently in wild populations, living fences, backyards and orchards, which most likely reflects relative selection pressures and rates of vegetative propagation. |
