Quantitative trait, genetic, environmental, and geographical distances among populations of the C4 grass Trachypogon plumosus in Neotropical savannas

Venezuelan savannas are exposed to land-use changes and biological invasions which compromise their persistence and function. The native C-4 grass Trachypogon plumosus is the most important component of the savannas under diverse combinations of climate and soils, suggesting substantial interpopulation variation. We examined quantitative traits and isozyme variation of nine populations of this grass and related these estimates to geographical and environmental features of sampled locations. Isozyme diversity estimates were based on 10 polymorphic enzyme systems whereas 21 quantitative traits, from field and controlled growth conditions, were evaluated. Distance matrices for quantitative traits, isozyme, geographical and environmental data were subjected to clustering analysis. Correspondence between quantitative trait distance and genetic distance, and their association to geographical and environmental distances were analysed with Mantel tests. All quantitative traits differed significantly among populations. The average Q(ST) calculated for eight quantitative traits measured in the greenhouse was 0.157. Isozyme diversity differed significantly among populations. About 28% of total isozyme variation occurred among populations. Significant positive associations were detected between environmental, quantitative field traits, and geographical distance as well as between the later and genetic distances. Genetic distances did not correspond significantly with quantitative traits nor did environmental distances. Ecologically meaningful associations were detected between field quantitative traits, environmental, and geographical data using cluster analysis. Our results support the hypothesis that processes of the neutral type are mainly responsible for the variation patterns observed in T. plumosus populations in Venezuelan savannas. Variation observed for quantitative traits among populations seems to be due to the effect of environmental conditions on phenotypically plastic traits, and not the result of directional selection favouring different phenotypes in different populations.