Submitted to: Meeting Proceedings
Publication Type: Proceedings
Publication Acceptance Date: January 26, 2010
Publication Date: July 9, 2010
Citation: Anderson, D.M. 2010. Geospatial methods and data analysis for assessing distribution of grazing livestock. Fourth Grazing Livestock Nutrition Conference Proceedings, July 9-10, 2010, Estes Park, Colorado. p. 57-90. Interpretive Summary: Managing free-ranging animals is not a trivial task. It involves understanding a complex and multifaceted interface composed of both biotic and abiotic components. Managing this interface presents many challenges to scientists charged with the goal of accurately understanding and subsequently managing the spatial distribution of free-ranging animals in an ecologically and economically efficient manner. To accomplish this, data must not only be collected using the proper tools but subsequently summarized and analyzed in a manner that interpretation will lead to valid and practically useful information for stewards of both public and private management entities. This review paper provides a comprehensive but not exhaustive review of the literature on the subject of methods and data analysis for assessing free-ranging livestock distribution. Between 1926 and 2009, at least 68 factors were identified that affected animal distribution. Furthermore, there are many mathematical approaches that have been applied to these data to interpret the factors and their interactions. Many of the original tools and techniques had their origin in the wildlife community. Today the global position system (GPS) offers researchers in both the wildlife and livestock production communities many advantages to address geospatial questions with precision and accuracy never before possible. Yet with this technology come new challenges not the least of which is how to manage and analyze these extremely large volumes of data. The only practical way to address the challenges associated with geospatial data analysis is to assemble teams of qualified persons who actively desire to work together. At a minimum these disciplines should include: ethology, range science, electronics, geospatial statistics and modeling. By using this approach there will not be a "one size fits all” approach to analyzing free-ranging animal data; rather an “optimum" approach for each unique research situation should result and the "most familiar" approach will be left to those scientists who choose not to work as part of a team.
Technical Abstract: Free-ranging livestock research must begin with a well conceived problem statement and employ appropriate data acquisition tools and analytical techniques to accomplish the research objective. These requirements are especially critical in addressing animal distribution. Tools and statistics used to describe the plant-livestock interface must consider that foraging is highly place dependent, foraging is not a random process and livestock foraging mainly occurs in groups. Since animal movement implies consecutive locations, each the result of a host of interactions between the animal and its surroundings, classical agronomic statistics may not provide the optimum analytical tools to address spatially related phenomena. By far the greatest amount of research on quantitative tools to understand the use of space by animals has come from the wildlife community and more recently from ecologists exploring the distribution of both animal and plant species over large geographic areas. However, interest in geospatial data analysis methods is also found among animal-range scientists, primarily because new electronic technologies make it possible to accumulate data from free-ranging animals in seconds or less. Therefore, the rate at which data can now be collected produces inordinate amounts of information that must be correctly analyzed. Consecutive animal locations on a landscape are not statistically independent; therefore, autocorrelation is no longer a theoretical but real concern and must be addressed in animal distribution analyses. This paper presents an overview of research that affects landscape use together with possible geospatial tools that can be used when evaluating free-ranging animal distributions. Geospatial methodology is not trivial nor is it a mature field, but rather is evolving; therefore, the range-animal scientist should work as a team member with other disciplines including but not limited to statisticians, modelers, computer scientists, cognitive scientists, physicists, sociologists, and possibly geographers when attempting to understand animal distribution. Literature referenced in this paper is comprehensive in breath but not exhaustive in detail and thus serves as a beginning point for those wanting to know what has already been researched concerning livestock distribution on landscapes.