Maize Diversity and Population Structure Related to Ethno-Linguistic Variation

There is a link between crop diversity and culture variation. However, there are few studies that focus on this topic at the local scale. This research took place in Southern Mexico, an area of high cultural and biological richness. It seeks to understand the relationship among ethnic and environmental variation and maize landraces distribution and population structure. Analyzing an up to date database of 18,355 georeferenced maize observations, I found that areas of high race richness and diversity occur in northwest, south central, and southern Mexico in the states of Oaxaca, Jalisco, Michoacan and Chiapas. Higher diversity is located in central west, northwest, central northwest of Mexico. There is no evidence of a decline in race richness over time. Our results show that maize race richness is strongly associated to environmental heterogeneity, mainly located in mountain ranges.;Chatino and Mixtec are both languages of the Otomanguean family. The family is possibly related to the people whom domesticated maize. I worked in four villages in the Sierra Sur, two Mixtec and two Chatino, one village of each group was located at low elevation (400 masl) and the other at medium elevation (1200 masl). I collected maize samples, interviewed farmers, planted common gardens and did molecular analysis. We found that maize race distribution is not independent of ethnicity but it is independent of elevation in those villages. Although we do not find population structure based on morphological traits, we find genetic differentiation by SSR markers among villages and between ethnic areas. As with maize races, genetic structure within the populations collected by us is due to ethnicity instead of variation in elevation. Larger values of Fst (0.18) were calculated compared to other maize studies in Mexico.;Research in agriculture usually assumes that yield is a primary selection trait for farmers practicing traditional agriculture, leading to the hypothesis that the cultivars or landraces found in farmers' fields are the highest yielding ones available to them. Reciprocal common gardens were planted in each village to test if local maize landraces were higher yielding in their respective villages, a finding that would suggest that they are selected because they are better adapted than landraces from the other villages. We also tested resistance to fungal disease (ear rot caused by Fusarium) that is cited by farmers in the region as a major problem in their maize production. The results showed that maize planted in the village of origin do not always yield better. There are significant interactions between common garden site, fertilizer use and seed source. We found that landraces from a particular village yield better in most of the sites. Regarding ear rot, there is some evidence that landraces are less susceptible when grown away from the village of origin. These results suggest that factors other than local adaptation may determine the distribution of maize landraces in the region. Environmental factors could be micro-niches defined by soil quality. And social factors could be seed management and exchange and the existence of ideotypes for each ethno-linguistic group.;Looking at seed management and seed flow, the findings show a similar seed management and amount of seed from external sources in villages in both ethno-linguistic areas, but seed exchange is virtually zero among villages of different ethnicity. A metapopulation model considering a single source of migrating seeds and no structure in seed flow estimates a Fst (0.07) smaller that an actual Fst estimated by SSR (0.18) implying that the actual seed exchange contributes to genetic structure. Our research provides insights about the importance of social origin in structuring maize population and diversity at local scale. As the results showed several factors are working together, but more research is needed to test for the presence of maize ideotypes linked to ethnicity that shape maize seed selection and acquisition. (Abstract shortened by UMI.).