All of the research conducted in our lab follows the same general approach. We use theory from population ecology, metapopulation ecology, and landscape ecology to develop hypotheses about the effects of management actions on animal population dynamics. We evaluate these hypotheses by first formalizing them as statistical models, and then fitting these models to data from carefully designed field studies and experiments. This process often requires the development of new statistical models, which is one component of our lab’s program.
The following is a list of some of the major themes we work on, with links to specific projects:
A central challenge faced in applied ecological research is to understand how populations will respond to future environmental change, including change resulting from management actions. Population viability analysis provides a framework for addressing this challenge, but traditional PVA methods don’t adequately account for spatial and temporal variation in demographic rates, and they typically ignored parameter uncertainty. Bayesian methods and spatial models overcome these issues, and we use these tools to predict future outcomes under various environmental change scenarios. When embedded in a formal decision analysis context, this approach can be used to identify optimal conservation actions.
Conservation and Agriculture
Agriculture is the primary land use throughout much of the world, and the demands for agricultural products continue to increase with concomitant increases in population size and standards of living. In addition, traditional agricultural practices have given way to intensified production systems that often remove all native vegetation. Meeting the demands for agricultural products while conserving native biodiversity is therefore an extremely challenging problem, but we see opportunities in coupling high-yield methods with explicit conservation offsets and market-based incentive programs. Our lab is studying such systems so that they can be implemented to benefit farmers and biodiversity.
Virtually all natural disturbance processes have been disrupted by human activities. Dams prevent flooding, fires are suppressed, and the effects of wind storms have been minimized by the even age structure of many forests. The problem with suppressing natural disturbance regimes is that a tremendous number of species evolved to utilize the habitat created by these processes. Many of these disturbance-dependent species are now declining, and since restoring natural disturbance is often impossible in built landscapes, active management options must be considered. Our lab studies ways in which habitat for these species can be created while considering the needs of species that do not depend on disturbance.
Trailing edge populations
All species distributions shift in time, but the rate of change is increasing as a result of rapid shifts in land use and global climate change. By definition, populations at the trailing edge of a shifting distribution face pressures to move or adapt to environmental change, yet these populations may be genetically unique and hence in need of conservation attention. This is especially true in the southern Appalachian Mountains where many species have trailing edge populations. Our lab is interested in determining how unique these populations are, and what conservation strategies can be used to mitigate the effects of rapid environmental change faced by these species.
Integrated population models
A mechanistic understanding of the factors affecting population dynamics requires knowledge of demographic parameters. A problem that faces virtually all applied ecologists is that estimating these parameters requires arduous capture-recapture studies that can only be implemented over relatively small regions and over short time periods. However, we often wish to make inferences over larger spatial and temporal domains. Our lab is developing novel methods for combining capture-recapture data with “cheaper” data such as count data or occupancy data to effectively scale up population models.
Spatial predator-prey dynamics
Management of many species is hardly possible without considering the target species’ predator(s) or prey. As well-developed as the theory of predator-prey dynamics is, methods to actually apply the theory are limited due to the computational challenges faced when modeling coupled space-time processes. These obstacles can now be overcome to some extent, and we are developing and applying novel methods to use predator-prey theory to inform management actions.
Hierarchical models are the ideal tool for drawing inferences about conditionally related ecological processes. They can also be used to account for the various sources of observation error (e.g. imperfect detection) that are associated with most ecological datasets. Our lab routinely applies and extends these tools to evaluate hypotheses and predict the effects of management actions on population dynamics. Much of the model development work is conducted in collaboration with Andy Royle, Marc Kery, and Jim Nichols.