Widespread winners and narrow-ranged losers: Land use homogenizes biodiversity in local assemblages worldwide
Fig 2
Modelled effects of human pressures on RCAR.
Effects are shown as a percentage difference from the value in minimally used primary vegetation, and were taken from the best-fitting models among those having no interactions between explanatory variables (results including interactions are shown in S3 Fig). Human pressures considered were as follows: land-use type and intensity, proximity to roads, and human population density. Error bars show 95% confidence intervals. Land-use type was classified as: primary vegetation, mature secondary vegetation, intermediate secondary vegetation, young secondary vegetation, plantation forest, cropland, pasture, and urban. We considered 3 alternative approaches for estimating species’ range size: range occupancy based on GBIF records (opaque points in A and B), range extent based on GBIF records with outliers removed (translucent points in A and B), and, for vertebrates only, range extent based on expert-drawn range maps (opaque points in C and D). For comparison, we also show estimates based on range occupancy from GBIF records for vertebrates (translucent points in C and D). We considered community-average range sizes both weighted (RCAR; A and C) and unweighted (RAR; B and D) by species’ abundance. For the models of all taxonomic groups (A and B), each land-use class was subdivided into 3 levels of human intensity of use—minimal, light, and intense (see S4 Table)—as indicated by different plotting symbols (light and intense intensity levels were combined for the secondary-vegetation classes). Effects of proximity to roads, vegetation removal, and human population density are shown here at the lowest, median, and highest values in the modelled dataset. Proximity to roads as shown here is the reverse of the distance to nearest road measure that was fitted in the models. Minimally used primary vegetation furthest from a road, and at the lowest level of vegetation removal and human population density, was used as the baseline. The site-level data underlying the models shown here are freely available (DOI: 10.6084/m9.figshare.7262732). GBIF, Global Biodiversity Information Facility; H, highest values in the dataset for each continuous effect; HPD, human population density; ISV, intermediate secondary vegetation; L, lowest values in the dataset for each continuous effect; M, median values in the dataset for each continuous effect; MSV, mature secondary vegetation; Plantation, plantation forest; PR, proximity to roads; Primary, primary vegetation; RCAR, Relative Community Average Range Size; YSV, young secondary vegetation.