Extended Data Fig. 6: Differences in α, β and γ diversity between fragmented and continuous landscapes using a meta-analytical approach.

This orchard plot shows the effect size (log-response ratio, LRR) of the overall difference between continuous and fragmented landscapes across studies (n = 37). The values of α, β and γ diversity were computed using (a) all possible plot pairs or (b) only the nearest plot pairs (controlling for distance decay effects) in both continuous and fragmented landscapes. We also calculated α, β and γ diversity using the observed species richness without controlling for commonness or sampling effort, and with individual-based rarefaction giving greater relative weight to rare species (rarefied species richness; q = 0) and individual-based rarefaction giving greater relative weight to abundant species (effective number of species given Simpson diversity; q = 2). Solid points represent the LRR comparing α, β and γ diversity between continuous and fragmented landscapes, and the error bars represent 95% confidence intervals. Positive effect sizes indicate that continuous landscapes have higher diversity than fragmented landscapes, while negative effect sizes would indicate that fragmented landscapes have higher diversity. Transparent points indicate effect sizes from individual sites, and their sizes are proportional to the precision (inverse of the square root of the variance) of the individual effect size estimates.