Distance from forest and community properties for all 19 sites.

<p>Regression lines are only fit to oil palm sites. Estimated richness is Poisson-compound gamma estimated (Y=-0.002x + 16.11, R²=0.66, F=18.87, df=10, P<0.01). Abundance is the mean hourly capture rate (Y=-0.001x + 8.38, Rsq=0.74, F=27.05, df=10, P<0.001). Similarity to forest is reported using the Cao method (Y=-0.0001x + 1.27, R²=0.81, F=43.08, df=10, P<0.0001). Mean phylogenetic distance is the observed scores for each site (Y=-0.001x + 23, R²=0.56, F=9.97, df=10, P<0.05). Geographic range represents the abundance-weighted mean number of biogeographic regions in which species occur (Y=0.0002x + 4.76, R²=0.50, F=10.06, df=10, P<0.01). Tongue length is the abundance-weighted mean across species (Y=0.0003x + 9.73, R²=0.14, F=1.63, df=10, P=0.23). Body mass is the abundance-weighted mean across species (Y=5.73E-4x + 77.44, R²=0.002, F=0.02, df=10, P=0.89).