Performing Department
Ecosystem Science & Management
Non Technical Summary
Wildlife populations are increasingly affected by changes to climate dynamics and land-use. We need a complete understanding of how landscape-level changes effect animal movement and space-use, and how these processes contribute to survival and reproduction, in order to make robust management decisions. Therefore, we need to:1) develop statistical methods to connect ecological processes,2) apply these methods to wildlife populations to answer relevant management questions, and3) use these methods, along with additional statistical tools, to determine why species and sub-populations respond differently.This project focuses on achieving these goals for wildlife both in Pennsylvania and across North America. Work will focus on a range of systems:potential systems and questions include determining the effect of forest management practices and hunting pressure on ungulates (white-tailed deer and elk), assessing the contribution of movement processes to the genetic composition of waterfowl populations, evaluating the connection between habitat-use and mortality in waterfowl populations and how this can contribute to population trends, improving our understanding of bat species' habitat use and population trends, and determining temporal changes in stream fish habitat use across multiple species.
Animal Health Component
20%
Research Effort Categories
Basic
40%
Applied
20%
Developmental
40%
Goals / Objectives
Objective 1: Develop and evaluate methods for modeling animal movement, space-use, or population dynamics and work towards mechanistically connecting these processes. This will include both models that use a single data source and rigorously integrating multiple data types into a single model to improve estimation and prediction of population processes. Robust methods are required to make accurate inference on drivers of these processes and account for individual variation in responses.Objective 2: Determine spatial and temporal effects of the landscape and climate on animal movement, space-use, and population dynamics. This will include incorporating multi-scale predictors and responses of animal populations that may only be observable by integrating multiple data types. Work will focus on the effects of habitat management, invasive species, disease, climate change, and other factors with the goal of informing on the ground management and conservation.Objective 3: Investigate sources of variation across species to spatio-temporal effects of landscape and climate on animal populations. Management actions can often have unintended or unanticipated effects (e.g., predator release) on other species.
Project Methods
Objective 1Analytical methods for modeling animal movement and integrated data methods will be developed in a hierarchical Bayesian framework. We will use simulations and case studies to evaluate the ability of our models to explain and predict ecological processes. We will also evaluate the information content of different types of data and how the necessary data may vary with the system in question. We will also investigate spatio-temporal dependence in ecological processes that may be a result of unmodeled sources of variation.Objective 2We will use the models developed or explored in Objective 1 to determine how changing landscape and climate dynamics contribute to population dynamics of wildlife, while accounting for measurement error. Models will be fit using custom samplers or pre-existing packages developed for fitting Bayesian models. The methods generated in Objective 1 will be flexible and can be used for a number of different systems of interest. For example, we can determine if white-tailed deer movement and space-use varies as a function of temporal and spatial predictors of habitat and hunter effort and how different movement behavior confers survival benefits.Objective 3We will incorporate multi-species dynamics into our models using ideas developed in the community ecology literature. In addition, we will use methods from the statistical literature, such as multiple imputation and two-stage models, to make inference on multi-species models when a single model is computationally infeasible. Comparing environmental responses across species will highlight species that can benefit from the same management actions, as well as predict species for which actions may be detrimental. In addition, we will use results from Objective 2 to determine natural history drivers for different types of responses.