Management Application of Statistical Population Reconstruction to Wild Game Populations

Abstract

Historically, management agencies in the United States have monitored most game populations through an ad hoc approach which combines indices, harvest data, hunter surveyed data, and occasional demographic evaluation. However, changing management priorities and increased scrutiny require more informative and defensible means of monitoring harvested populations. Statistical population reconstruction (SPR) is a flexible modeling system which simultaneously analyzes age-at-harvest data, hunter effort data and any additional demographic data which are available, producing estimates of abundance, natural survival and harvest rate, as well as their associated variances. An SPR based monitoring framework provides comprehensive analysis of commonly collected data and represents a statistically rigorous and defensible alternative to the currently popular approaches. However, applications of SPR have previously been limited to small scale, highly monitored populations, primarily due to a lack of formal evaluation of data requirements and guidance for management application. In this dissertation I provide the guidance necessary for broad scale application of SPR modeling to monitor harvested species. I rigorously evaluate the relative utility of auxiliary data sources as well as minimum harvest and hunter effort data requirements for SPR models, providing necessary guidance for resource managers seeking to apply SPR. I present a historic population reconstruction based on SPR parameter estimates as an illustrative example of the management application potential of SPR output. I comprehensively evaluate models to project reconstructed abundance into the future in order to further increase the management utility of statistical population reconstruction. I provide a detailed explanation of model structure and assumptions, allowing resource managers to critically evaluate SPR models. Finally, I offer guidance on the customization of SPR models necessary to adequately model the harvest regimes and data collection methodologies which are unique to each harvested population, thus increasing the number of populations which can be modeled. This dissertation will facilitate the broad scale management application of SPR, thus increasing the rigor and efficiency with which harvested game populations are monitored.