Ecological Modeling and Global Change

We use population models to 1) understand the mechanisms behind population declines in native biota and 2) test potential risk management options for declining populations. A focus of this research theme is the application of stochastic population models to study the effects of global change on risks of population decline. While habitat loss is the single most dominant threat to global biodiversity it also promotes myriad other threats that exacerbate declines in biota: logging of old growth forests creates access routes for poachers and disease, land conversion eliminates native pollinators, urbanization introduces new predators to native fauna, and habitat loss and fragmentation encourage invasive species which in turn alter natural disturbance regimes. Climate change is expected to interact with these threats.

While we work on both plants and animals, our current research examines the impacts of global change on plants in ecosystems where fire plays an important ecological and evolutionary role. In these systems, habitat loss and fragmentation indirectly alter the fire regime and promote invasive species, predators and disease. We link empirical data, population models, species distribution models, climate data, urban growth models and fire hazard functions to investigate the impacts of climate change, habitat loss and fragmentation, altered fire regime and invasive species on a range of plant functional types in southern California. We aim to identify the species traits that make plants vulnerable to these threats in isolation and combination and in doing so, develop a conceptual framework that will make it easier to predict the species and communities most vulnerable to global change.

We are extending this work to investigate potential management and climate change adaptation strategies to reduce risk of species decline and extinction.

Characterization and Treatment of Uncertainty    

A major focus of our work is analyzing the extent to which uncertainty impacts our knowledge and understanding of ecological systems. The fields of engineering, computer science and human and ecological risk assessment have developed methods for uncertainty analysis that have immediate utility in ecology and conservation biology. We investigate and apply a variety of techniques for dealing with the different sorts of uncertainty inherent in the systems ecologists study and the data they collect.

We have compared a range of methods including Monte Carlo analysis, probability and dependency bounds convolutions, interval analysis, fuzzy sets and information gap decision theory and applied them to a variety of problems in conservation biology and ecological risk assessment to demonstrate how treatments of uncertainty can be implemented, how conclusions can change when uncertainty is considered and how the results of uncertainty analyses can be used in decision making.

Applications include the characterization and treatment of uncertainty in: estimates of global mass extinction rates (Regan et al. Am Nat 2001); IUCN categories and criteria (Regan et al. Biol Cons 2000; Regan et al. Cons Biol 2000); soil screening levels for chemical contamination to wildlife (Regan et al. ET&C 2002); a food-web model of contaminant exposure (Regan et al. HERA 2002); ranking of management options for endangered species (Regan et al. Ecol Appl 2005); and marine reserve spacing rules (Halpern et al. Ecol Letts 2006).

Ongoing work in this area focuses on characterization of uncertainty in across a range of models used to study global change impacts to biodiversity; guidance on uncertainty analysis in the IUCN Red List criteria; determining the length of time series data necessary to capture population trends under variability and uncertainty; and model selection and validation with error-prone data.

Applied Decision Theory

While numerous quantitative methods are currently employed to assess the risks of threat to biodiversity in the face of uncertainty, current research in conservation biology falls short of fully addressing the trade-offs between
scientific investigation and the social, economic and management issues of multi-use environments. For instance, formal reserve design algorithms have become important practical tools for conservation planning, yet they usually rely on only a subset of biological and social values and those values are often not elicited from all the relevant stakeholders in an equitable way. One of the over-riding issues in stakeholder groups is aggregation of values
across group members. Ad hoc consensus and negotiation are the usual course of action; however, there are serious impediments to this due to group member motivation and status.

Our research in this theme focuses on formal decision making tools to avoid some of the pitfalls of behavioral decision making. We have assessed the types of social values that motivate protection of biodiversity and evaluated the types of values that are most useful for decision making. We are currently reviewing the types of social aggregation methods available for environmental decision making and the types of problems most amenable to them. We have also reviewed games within social choice theory to characterize strategies for decision making under conflict in conservation biology. This work holds promise for groups operating under conflict or with opposing agendas, which is often the case in conservation management. It is also a departure from the way conservation biologists usually think about decision making in their field.

We are currently applying stochastic dynamic programming, a decision framework for optimizing decisions under risk and change, to design cost-effective fire management strategies that maintain biodiversity while minimizing conflict with land-use activities.