Challenges of using behavior to monitor anthropogenic impacts on wildlife: a case study on illegal killing of African elephants


Animal Conservation DOI: 10.1111/acv.12309

Date Published


S. Z. Goldenberg1,2,3, I. Douglas-Hamilton3,4, D. Daballen3 & G. Wittemyer1,2,3 1 Department of Fish, Wildlife, and Conservation Biology, Colorado State University, Fort Collins, CO, USA 2 Graduate Degree Program in Ecology, Colorado State University, Fort Collins, CO, USA 3 Save the Elephants, Nairobi, Kenya 4 Department of Zoology, University of Oxford, Oxford, UK
Monitoring anthropogenic impacts on wildlife can be challenging, particularly when human activities affecting wildlife are cryptic. Using anti-predator behaviors as proxies for perceived pressure is appealing because of the relative ease with which they can be recorded and the presumed relationship between the threat of interest and a predator stimulus. However, behaviors are plastic and affected by factors unrelated to human activity. Consequently, it is critical to assess the relationship between behavioral indicators and their context before interpretation. In this study we used a combination of behavior, movement and demography from a threatened population of African elephants in northern Kenya to determine whether reaction to research vehicles was indicative of poaching pressure. We used mixed effects models predicting reaction of elephants to observer vehicle approaches in which we treated individuals as random effects and included ecological, anthropogenic, spatial, social and demographic predictor variables. Contrary to our hypothesis, recorded levels of reactive behavior did not increase with poaching levels in either a population-level dataset or a data subset of individuals whose spatial behavior was precisely known via radio-tracking. Rather, primary productivity positively predicted reactive behavior in both datasets. This relationship was heightened by the presence of musth males in the radio-collar dataset. Reactivity was not related to the time since entering the protected areas, but increased among groups that spent less time in the protected areas. Inter-individual differences were apparent, suggesting the importance of inherent differences (e.g. personality) across groups. In our study, elephants plagued by a severe human threat did not react defensively to humans in another context, suggesting nuanced discrimination of threats. Our study demonstrates the caution that should be taken in designing studies that use behavioral indices to represent threat and contributes to a growing body of literature employing behavioral indicators to monitor wildlife populations of conservation concern.