Resource information
CONTEXT : Understanding the variability of vegetation distribution and its determinants is a central issue for addressing the effects of edges on ecological processes. Recent studies have revealed inconsistencies in the patterns of responses to edge effects that raise important questions about their determinants. We investigated the edge effect response patterns by adapting a recently proposed continuous approach to the case of small forest fragments in southwestern France. METHODS : We surveyed forest understory vegetation (composition, species richness, and percent cover) and abiotic variables (soil temperature, moisture, pH, and canopy openness) along 28 transects across hard forest edges. We tested five statistical models to describe the response pattern of each variable (1) over all transects and (2) per transect. We then compared the response patterns as a function of the attributes of the edge (orientation, topography, and adjacent land cover) and forest patch size. RESULTS : Over all transects, a general decreasing trend was observed for all variables as the distance from the edge increased. In the individual transects, we evidenced a large variability in the response patterns that was not related to edge attributes or to patch size. CONCLUSION : It is difficult to assess the depth of edge influence in highly fragmented forests and to identify the determinants of edge effects. We recommend that care should be taken with studies using pool of transects, and that further studies should be carried out including situations with neutral patterns, in order to gain a broader understanding of edge effects on vegetation.
