Results:
Across the surveyed plot locations in Jamaica and Hispaniola the combined probabilities of predatory bird species (essentially the expected species richness of avian predators along a hypothetical 1km transect) ranged from 2.36 to 8.00 (Figure 1). Predatory bird presence in the Dominican Republic had a wider range (2.36 - 8.0) than Jamaica, which showed less variation in predation pressure (3.67 - 6.71). On average, however, predator presence was statistically indistinguishable between islands (5.41 for Jamaica and 5.46 for the Dominican Republic; p-value = 0.83).
Across all plots, abundances of lizards were ~21% higher on average in Jamaica, with mean abundance in 15m radius (706 m2) plots of 65.39, compared to 53.95 in the Dominican Republic (p-value<0.001). Both islands were similar when comparing species richness of local communities. The average species richness on both Jamaica and the Dominican Republic was slightly less than 4 species (3.55 and 3.63, respectively, p-value = 0.83).
We first assessed the relationship between climate and the presence of both anoles and bird predators individually. Species richness of anoles was positively correlated with mean annual temperature (Spearman correlation coefficient = 0.70, p-value<0.001, figure 4b) and negatively with precipitation (Spearman correlation coefficient = -0.35, p-value<0.001). Similar patterns existed for anole abundance: a positive relationship with temperature (Spearman correlation coefficient = 0.60, p-value,0.001) and a negative relationship with precipitation (Spearman correlation coefficient = -0.29, p-value<0.01). In contrast, predatory bird occurrence was correlated only with mean annual temperature (Spearman correlation coefficient = 0.64, effect of temperature on abundance = 0.12, p-value<0.001, r2 = 0.50; Figure 2a) while its correlation with mean annual precipitation trended negative, but was non-significant (Spearman correlation coefficient = -0.14, p-value = 0.18, r2 = 0.01). While predation pressure correlated with climate temperature, there was still substantial residual variation in the relationship, allowing for the potential to detect independent effects of predation and climate on lizard communities.