Contact rates in ecology : using proximity loggers to explore disease transmission on Wyoming's elk feedgrounds

Abstract

Infectious diseases are an important consideration in the management of wildlife populations, and contact rate is a key parameter for understanding the epidemiology of such diseases. In the first section of this thesis, I review current issues and challenges that researchers face when designing animal contact studies and analyzing contact data. I examine how characteristics of methods for collecting contact data affect inferences that can be drawn about contact structures; describe applications of social network analysis of contact data to disease ecology and animal behavior, focusing on sampling issues and dynamic networks; suggest how new technologies can be used to answer important questions about variation in individual contact rates within populations; and propose a new statistical approach for analyzing contact data in a linear modeling framework. In the second section, I describe an experimental field study that used proximity loggers (a new technology for measuring contact rates) to understand transmission of Brucella abortus on elk feedgrounds in Wyoming. Brucellosis is a bacterial disease that causes abortions in elk and is transmitted by contact with infectious aborted fetuses. Supplemental feeding of elk on winter feedgrounds is believed to exacerbate B. abortus transmission by aggregating elk at high densities, increasing their chance of contacting infectious fetuses. I evaluated the effectiveness of a proposed low-density feeding strategy by comparing elk-fetus contact rates (as measured by proximity collars and video cameras) during high-density and low-density feeding treatments that provided the same total amount of food at different densities. Low-density feeding led to >50 percent reductions in the total number of contacts and the number of individuals contacting a fetus. Elk contacted fetuses and random control points equally, suggesting that elk were not attracted to fetuses but encountered them incidentally while feeding. The relationship between contact rate and disease prevalence is non-linear and simple disease models suggest that low-density feeding may result in dramatic reductions in brucellosis prevalence, though this depends on the amount of transmission that occurs on and off feedgrounds.