Detecting potential declines in a threatened dung beetle with N- mixtured models: Eurysternus impressicollis Castelnau 1840 in north-western Venezuela

AbstractImplementation of conservation planning and management strategies for threatened dung beetle species have been hampered by lack of reliable information about temporal and spatial patterns in abundance. Distinguishing “real” patterns responding to ecological processes, from “artifacts” created by sampling limitations, is not a simple task. Pitfall trap data from NeoMaps standardized surveys (2006 and 2009) were combined with environmental variables from remote sensors, and used N-mixture models to evaluate how observed changes in detection and abundance ofEurysternus impressicollis, a Vulnerable dung beetle species in northwestern Venezuela, could be attributable to differences in sampling effort or changing conditions in habitat across time. Minor temporal changes in detectability had a positive relationship with sampling effort, but a negative relationship with evapotranspiration. Sampling effort was enough to detect the species where it was present, so lack of detection in other transects are likely to represent true absences. Temporal change in abundance was explained by vegetation condition and temperature so, the low abundance recorded in 2009 were not a sampling artifact, but were likely to reflect multi-year fluctuations in environmental conditions. Also, the model allowed to explain landscape variation in abundance, being more abundant in localities with denser, more humid forest. This approach could be widely applied to pitfall data from other dung beetle monitoring programs across the world to inform conservation status assessment and guide future studies on the distribution and ecology of dung beetle threatened species.