Sandhill crane population monitoring, modeling, and harvest decision making

摘要

Managing wild animal populations can be challenging. This dissertation research was motivated by information needs outlined by sandhill crane ( Grus canadensis) researchers and managers. Questions and concerns arose about the current management strategy for the Rocky Mountain and Mid-Continent Populations, including whether there is potential for transient dynamics to dominant population change, what the limitations of the current population monitoring programs mean for making harvest decisions, and whether the current harvest decision process is appropriate to meeting long-term population objectives.;In Chapter 1, the objective was to organize the current scientific and management literature of sandhill cranes, as no such effort had been done since the mid-1990's. Since this dissertation is directly aimed at understanding RMP sandhill cranes, it was important to be able to frame research findings in the broader context of sandhill crane conservation and management. One significant finding is that, despite a considerable effort from researchers and management agencies dedicated to the conservation of sandhill cranes, there are surprisingly few studies that estimate population-level parameters. This made it challenging to construct realistic population models.;In Chapter 2, I investigate the relationship between the annual proportion of juveniles in the RMP of sandhill cranes and hypotheses related to variability in climate factors. Specifically, I was interested in simultaneously evaluating hypotheses about how short- and long-term drought, as well as weather patterns, affect the production of juvenile sandhill cranes while also developing a predictive model. I discuss predictive modeling and discuss how generalizable predictive models can be useful and perhaps needed by conservation and management decision makers. A major finding was that both short- and long-term drought, but mostly long-term drought, may severely limit the ability of sandhill cranes to produce young because of breeding habitat limitations. In addition, I found the number of freezing days and snowpack during the summer months can also negatively affect juvenile recruitment, while spring snow-pack has a positive effect.;In Chapter 3, I investigate the potential for sandhill crane populations to be governed by non-equilibrium dynamics. More specifically, I investigated transient potential, population growth under non-equilibrium dynamics, transient and asymptotic sensitivity to vital rate perturbations, and whether observed harvest could impact crane stage structure and lead to population inertia. However, the largest transient effects will dominate population trajectories of these species when stage structures are highly biased towards the younger and non-reproducing individuals, a situation that may be rare in established populations of long-lived animals.;In Chapter 4, I focus on assessing the past and current population monitoring data of the RMP. This data are used directly to make annual harvest decisions, but it is unknown whether the variation observed in the fall counts and population index could be indicative of population dynamics. By integrating population-level data, including juvenile recruitment, harvest, and survival estimates into a stochastic stage-based population model, I was able to assess whether annual changes in the observed counts and fall population index were biologically realistic. I found spring migratory and fall staging area counts of the RMP to exhibit annual change that was biologically unrealistic with spring counts being more variable that fall counts. In contrast, I found the three-year moving average fall population index to exhibit biologically reasonable annual change. In addition, I investigated the usefulness of smoothing counts using a hierarchical Bayesian time series (HBTS) modeling approach and compare that to the current approach of using a moving three-year average. .;In Chapter 5, I focus directly on the harvest decision process by using a simulation approach to compare the current RMP sandhill crane harvest decision framework and an adaptive resource management (ARM) framework. I consider nine different scenarios with varying levels of monitoring, modeling, and decision uncertainty. The underlying population dynamics follow a complex stochastic stage-structured, density-dependent population model. The two primary objectives of this study were to evaluate the robustness of the current crane harvest decision framework and to evaluate whether an ARM framework with relatively simplistic population models and uncertainty about the population size could decrease risks associated with the RMP framework.;Ultimately, the long-term management of sandhill cranes will require integrating harvest management, where decisions occur on a relatively short time-scale, with the long-term decisions associated with habitat management of key breeding, migratory, and wintering areas. Important considerations for future conservation strategies includes ensuring that migratory sandhill crane populations can find suitable breeding wetlands and adequate resources on migration and at wintering grounds. Migratory populations rely on agricultural production to meet energetic requirements, which means that shifting agricultural production may have a significant impact on crane behavior and population dynamics. (Abstract shortened by UMI.).