Propagation distance thresholds of woody Cerrado species in a pasture matrix

Abstract

Question: Is there a decrease in the richness, structural variables and composition similarity of woody Cerrado species with habitat-matrix distance, potentially influenced by soil compaction and propagule dispersal strategies in the pasture matrix? Location: We studied three landscapes of a pasture matrix in areas of the Cerrado biome in southwestern São Paulo, Brazil. Methods:We sampled 110 plots of 2 m × 2 m in the matrix, stratified in ranges perpendicular to the edge of the habitat, and 10 plots in the forest with no edge effect (60 m from the edge). We undertook soil sampling and collected information about the plants (taxonomy, height, circumference and establishment form by germination or resprouting). We defined habitat distance as the predictor variable for the following response variables: abundance, richness, establishment form, mean height, mean diameter at soil height and dispersal syndrome of the individuals sampled; in addition to the porosity and bulk density of the sampled soil. Results:We found significant differences between habitat and matrix for abundance, richness, mean height and germination number, but non-significant differences for resprouting number and mean diameter at soil height. There was no relationship regarding matrix-habitat distances for structural variables, and habitat species composition was similar only for the first 10 m distance from the habitat edge. Conclusion: Pasture did not prevent plant establishment but might act as an environmental filter. Pasture matrices modify the composition and diversity of the plant community compared with the Cerrado community. Pastures are dominated by resprouting, autochory and grassland species. Regenerating plants from the propagule bank of the matrix showed the potential for natural regeneration of Cerrado areas in pasture as a restoration strategy.