| Historically, the processes linked to economic change, such as population growth and agricultural modernization, have been a threat for crop diversity. These negative effects are changes in both genetic diversity (i.e. genetic erosion) and the nature of crop evolution. In the case of Mexico, agricultural modernization has had limited negative effects on maize diversity despite significant institutional efforts to promote modern technologies that include improved varieties. By contrast, market development and integration may have a greater potential to induce changes in the farmers' management of maize diversity. Integration with world markets, also known as globalization, is a cornerstone of economic changes involving the inhabitants of rural villages mainly through their participation in the labor markets at different geographic ambits.;The aim of this study is the analysis of the relationship between farmers' integration with the off-farm labor market in Puebla Valley, Mexico and maize diversity. The research results are presented in three chapters. The first chapter analyzes maize diversity using ethnographic data and the concepts of variety and race, as well as the management of seed lots. The second chapter analyzes the pattern of the maize phenotypic variation based on experimental measurements of both phenological and morphological traits. The last chapter addresses the relationship between maize diversity and households' labor allocation outside of agriculture. Maize diversity is measured in two ways: (1) phenotypic variation within population, here expressed as the average Mahalanobis distance, and (2) by the number of varieties per household. The analysis presented here finds no relationship between diversity and farmers' involvement in non-agricultural activities, contrary to what was hypothesized. Instead, maize phenotypic diversity is negatively related to the household's wealth status since the more diverse maize populations are managed by the poorest households.;These poor households are headed by young and poorly educated men and their families who live in small houses, own the smallest landholding and lesser number of fields where they plant varieties whose seed keep for a short time. This rapid seed turnover is not related to the availability of family labor, since poor households have the biggest workforce and rely less on hired labor. Hence, labor allocation to seed selection and maize production is not a constraint for these households. In contrast, the better off households plant more varieties, have more land and keep the basic seed lot for a long time, but they manage less diverse maize population.;The greater diversity of maize managed by poor households derives from the way they manage seed lots. Other practices related to seed management, such as seed selection, infusion, as well as natural cross-pollination, are common to all households regardless of their wealth status. The differences between the poorest and wealthiest households regarding seed lots and their consequences for phenotypic diversity deserve further research.;This research challenges the generalized assumptions that the number of varieties planted per household is directly related to increased diversity at a village or at a regional level, and that seed lots belonging to the same variety are similar in their contribution to overall diversity. Determining spatial diversity by the relative frequency and distribution of each variety is erroneous in light of the findings of this study.;The maize populations of the study area share many similarities regardless of their varietal or racial affiliation. A relatively narrow range characterizes differences among varieties in terms of seed lots or land area, resulting in a defined structure. Thus, the concepts of race and variety have limited usefulness for assessing maize diversity because there is not a defined type of plant associated with each racial group or variety. For instance, when accessions are grouped by their phenotypic similarity, each group consists of a mixture of racial groups and varieties. Previous assessment of maize diversity based on the concept of race using ecological indices indicates that there is a considerable variation across geographic locations and scales, such as villages, environments (valley versus foothills) and micro-regions. The analysis of the phenotypic diversity presented here demonstrates the limitations of this type of analysis to get a reliable picture on the current status of maize diversity. |
