Species turnover drives β-diversity patterns across multiple spatial scales of plant-galling interactions in mountaintop grasslands

This study describes differences in species richness and composition of the assemblages of galling insects and their host plants at different spatial scales. Sampling was conducted along altitudinal gradients composed of campos rupestres and campos de altitude of two mountain complexes in southeastern Brazil: Espinhaco Range and Mantiqueira Range. The following hypotheses were tested: i) local and regional richness of host plants and galling insects are positively correlated; ii) beta diversity is the most important component of regional diversity of host plants and galling insects; and iii) Turnover is the main mechanism driving beta diversity of both host plants and galling insects. Local richness of galling insects and host plants increased with increasing regional richness of species, suggesting a pattern of unsaturated communities. The additive partition of regional richness (gamma) into local and beta components shows that local richnesses (alpha) of species of galling insects and host plants are low relative to regional richness; the beta (beta) component incorporates most of the regional richness. The multi-scale analysis of additive partitioning showed similar patterns for galling insects and host plants with the local component (alpha) incorporated a small part of regional richness. Beta diversity of galling insects and host plants were mainly the result of turnover, with little contribution from nesting. Although the species composition of galling insects and host plant species varied among sample sites, mountains and even mountain ranges, local richness remained relatively low. In this way, the addition of local habitats with different landscapes substantially affects regional richness. Each mountain contributes fundamentally to the composition of regional diversity of galling insects and host plants, and so the design of future conservation strategies should incorporate multiple scales.