Using DNA to test the utility of pellet-group counts as an index of deer counts

Abstract

Despite widespread use of fecal pellet-group counts as an index of ungulate density, techniques used to convert pellet-group numbers to ungulate numbers rarely are based on counts of known individuals, seldom evaluated across spatial and temporal scales, and precision is infrequently quantified. Using DNA from fecal pellets to identify individual deer, we evaluated the relationship between pellet-group count and count of Sitka black-tailed deer (Odocoileus hemionus sitkensis) during a 3-year study (2006–2008) in 3 watersheds in southeast Alaska, USA. We surveyed 141,054 m2 of transect, counted 10,569 pellet groups, and identified 737 unique deer. We used a multilevel mixed-effects generalized linear model to analyze expected deer count as a function of pellet-group count. Pellet-group count was a significant predictor of DNA-based index of deer count, but that relationship varied by transect, watershed, and year, indicating that extrapolation of a single linear relationship across space and time was not possible. More importantly, most of the variation in our models was residual and unexplained. Assuming that our DNA-based results were a more accurate and precise metric of true deer count, we do not support the use of pellet-group count to index deer count in southeast Alaska unless confounding factors are accounted for at fine spatial (e.g., habitat patch) scales. Because of the difficulty in routinely evaluating the influence of confounding variables in remote and unmanaged landscapes, we suggest that wildlife programs in these environments consider alternatives, such as DNA-based methods, for monitoring trends in ungulate populations. © 2013 The Wildlife Society.