Quantitative classification of coffee agroecosystems spanning a range of production intensities in central Veracruz, Mexico

Coffee production has attracted considerable attention globally, due to its economic, social, and ecological importance. The capacity of coffee farms to conserve the biodiversity and environmental benefits offered by adjacent forest ecosystems varies greatly in relation to varying cultivation strategies. However descriptions of these strategies are scarce and largely qualitative in nature, thus hindering comparisons between studies. A rigorous quantitative classification of this agroecosystem was undertaken, in order to address these concerns. For this purpose, a multivariate analysis was applied, in order to analyze the changes in the biophysical structure and management of 18 coffee plantations and three fragments of montane cloud forest, spanning a wide variety of cultivation intensities in central Veracruz, Mexico. This analysis identified five main classes of vegetation structure, ranging from sun exposed to rustic coffee plantations, with the mean height of shade trees, vertical vegetation diversity, tree richness and abundance and coffee plant density, representing the most important structural descriptors, referring to the farms studied. Analysis of the frequency and type of management practices employed (fertilization, weed and pest control) yielded three groups of farms ordered along a gradient, ranging from conventional to alternative practices. Together, these analyses yield a robust quantitative classification system for coffee farms in central Veracruz, which differs in several important ways from accepted qualitative classification schemes. As vegetation structure and management practices did not co-vary in this analysis, future studies should include standardized measurements of both dimensions used to describe coffee farms, in order to improve understanding of how intensification affects conservation potential and help to identify more sustainable production strategies. (C) 2009 Published by Elsevier B.V.