Spatio-temporal dynamics of a tree-killing beetle and its predator

15 February 2016

Weed, Aaron; Ayres, M. P.; Liebhold, Andrew; Billings, Ronald

Resolving linkages between local-scale processes and regional-scale patterns in abundance of interacting species is important for understanding long-term population stability across spatial scales. Landscape patterning in consumer population dynamics may be largely the result of interactions between consumers and their predators, or driven by spatial variation in basal resources. Empirical testing of these alternatives has been limited by the lack of suitable data. In this study, we analyzed an extensive network of spatially replicated time series to characterize the local and regional processes affecting spatio-temporal dynamics of a tree-killing bark beetle (Dendroctonus frontalis or SPB) and its key predator (Thanasimus dubius) across the southeastern United States. We first used a mechanistic model to evaluate factors affecting the stability of 95 predator-prey time series and then conducted spatial analyses to evaluate scale dependence in the factors affecting the geographical patterning of this system. Across the region, population fluctuations of both species were correlated in space beyond 400 km but there was notable spatial variation in the deterministic and stochastic processes influencing forest-scale (local) fluctuations. Time series analyses indicated that local dynamics of SPB and T. dubius are not cyclical. Instead, the abundance of T. dubius responded almost instantaneously to changes in SPB abundance. Spatial variation in long-term forest-scale abundance of both species was linked most strongly to the abundance of pine habitat indicating a stronger role for resource availability in SPB population dynamics than top-down effects. Our results are consistent with other studies indicating that animal populations tend to be synchronized in space via spatially correlated processes such as weather; yet local dynamics tend to be linked to smaller-scale host patterns. Our study provides a rare empirical assessment of how local processes scale up to produce landscape patterns that influence forest ecology and forest management.