Rapid scavenging of avian carcasses in the desert Southwest: Implications for mortality surveys not associated with infrastructure

Abstract Avian populations have experienced widespread declines with mortality attributable to a range of factors. Carcass surveys have been employed to document and quantify avian mortality in both infrastructure‐related and non‐infrastructure‐related scenarios. To improve the accuracy of mortality estimates obtained from carcass surveys, correction factors can be used to control for sources of bias such as the removal of carcasses by scavengers. However, given differences in how carcasses appear on the landscape between infrastructure‐related mortality and non‐infrastructure‐related mortality events it is possible that estimates of carcass persistence from infrastructure‐focused studies may not be relevant for all mortality monitoring. We conducted a study of carcass persistence on White Sands Missile Range in the Tularosa Basin of New Mexico, USA, simulating mortality unrelated to infrastructure. We placed and monitored 214 avian carcasses over 2 seasons, determined how long each carcass persisted prior to being scavenged, and examined the influence of habitat, temporal, and scavenger guild variables using Weibull survival models. We found that 211 of the 214 placed carcasses were scavenged and median persistence time was 1.53 days. Our analysis showed that the null model was the most competitive, and the 85% confidence intervals for all predictor variables overlapped zero, indicating none were informative for carcass persistence times. Overall, our findings suggested that persistence times in a non‐infrastructure‐related mortality scenario are similar to those associated with infrastructure, underscoring the need to account for rapid scavenger removal when designing mortality monitoring programs.