Inferring the nature of anthropogenic threats from abundance time series records

Background/Question/Methods

Planning and managing the recovery of at-risk populations and species requires correct identification of threatening processes. In most cases, threatening processes are inferred based on expert opinion and qualitative information about life history, extent of occurrence, and recent or anticipated environmental change. However, long-term abundance time series records are becoming available for a large number of wildlife populations worldwide and these data provide rich opportunities for investigating the population and species-level effects of anthropogenic environmental change. Improved tools and methods for inferring ecological processes from observed abundance patterns are urgently needed to facilitate and enhance critical assessment of future population trends and species extinctions. To this end, we are developing and testing a flexible framework for drawing inferences about the nature of anthropogenic threats and for inferring causes of species endangerment based on abundance time series records as well as ancillary sources of quantitative and qualitative information.

Results/Conclusions

Preliminary results suggest that our method can reliably distinguish between habitat loss and overharvest scenarios. For simulated populations subjected to habitat loss (linear decline in carrying capacity), our algorithm successfully discriminated the correct underlying model from a candidate model set including several harvest and null (no threat) scenarios for 74 of 80 replicate time series (n = 80, 5 candidate models). However, false positive rates have been high (25% or higher in some cases) for some other simulated threat scenarios, suggesting that model averaging techniques may be necessary to infer mechanisms of decline and to project future population trends while accounting for structural uncertainty. Currently we are evaluating the performance of model-averaged population risk metrics for inferring population-level risk in the face of uncertain threats. Overall, we anticipate that this project will represent an important advance for conservation biology, where there is a critical need for improved methods for identifying the fundamental threats to species. Our framework may ultimately assist conservation organizations such as the International Union for Conservation of Nature (IUCN) in documenting threatening processes and evaluating extinction risk.