PS 40-18 - Remaining native savanna and landscape heterogeneity predict the abundance of a global invader

Background/Question/Methods

Commodity-driven deforestation continues to replace Neotropical native forests and savannas creating opportunities for non-native species; however, few studies have investigated the predictors of invader abundance. We investigated the relative importance of land cover and spatial and temporal landscape heterogeneity metrics in predicting the abundance of a global invader, the European hare (Lepus europaeus Pallas, 1778; Mammalia, Lagomorpha), in Neotropical landscapes. We sampled 55 landscapes from southeastern Brazil with camera-traps and transect surveys. This region is a transition zone between the Cerrado (Brazilian Savanna) and the Atlantic Forest. Landscapes were randomly selected and widely distributed ( > 6 km apart) along a gradient of native vegetation cover and landscape heterogeneity. We estimated European hare abundance using an occupancy model (Royle and Nichols model) that accounts for species imperfect detection induced primarily by variation in local hare abundance. This hierarchical model estimates the species’ abundance from replicated surveys (i.e., detections/ nondetections) on sampled landscapes. We modeled the mean species abundance, λ, using all possible additive combinations of up to four uncorrelated predictors (|r| < 0.7 and VIF < 10). We used AICc-based model selection and cumulative AICc (w₊) weights to determine the most important predictors (w₊ > 0.50) of European hare abundance.

Results/Conclusions

Savanna cover extent (w₊ = 0.79) and landscape compositional heterogeneity, quantified by the Shannon’s Diversity Index (w₊ = 0.57), were key predictors of European hare abundance. Hare abundance was positively associated with compositional heterogeneity, suggesting heterogeneous agricultural landscapes might support local abundance “hot spots” of this invader. We also found hare abundance to be negatively associated with the savanna cover, suggesting this native open canopy habitat may hinder this invader from reaching large local populations. We predict that hare distribution will expand northwards given the ongoing replacement of Cerrado remnants by intensive commodity row crops in northern Brazil; however, hare populations may remain low due to the highly homogeneous row crop landscapes. This reasoning suggests that the impact of the invasive hare on biodiversity in intensively managed agricultural landscapes might be correspondingly low, with an expansion rate likely varying spatially given the diversity of Brazilian land cover regimes. Nevertheless, little is known about the direct and indirect impact of European hare on native Neotropical biodiversity and future studies should investigate hare interaction with native species in different agricultural regions to further clarify this invasion process.