Jump to main content.

Goal - Sound Science - Modeling

Terrestrial Habitats - Effects, Modeling and Extrapolation

EPA’s ecological research program seeks, through scientific leadership, to increase understanding in order to assess, improve, and restore the integrity and sustainability of ecosystems over time. Specifically, research in this area – Processes and Modeling Research – will develop models to understand, predict, and assess the response of ecosystems to multiple stressors at multiple spatial and temporal scales.


By 2008, ORD is committed to develop a new generation of environmental modeling tools to protect ecosystems at the local, watershed, and regional scales. These models will support decision makers in their efforts to make better ecologically sustainable choices.

To address these objectives, NHEERL has developed an Implementation Plan for research to address wildlife population endpoints through terrestrial habitat quantity, quality and distribution and as affected by multiple stressors across many temporal and spatial scales. The plan calls for WED scientists to take the lead in terrestrial habitat and wildlife population modeling while collaborating with the other Ecological Research Divisions to address the overall Agency problems.

Under these plans, this research project will respond to Program Office needs in three specific problem areas. First, the Scientific Advisory Panel for the Federal Insecticide, fungicide, and Rodenticide Act (FIFRA) specifically recommended that the Office of Pesticide Programs conduct probabilistic assessments of risks to ecosystems associated with pesticide use. Second, the Office of Prevention, Pesticides, and Toxic Substances needs efficient methods, including models, to review, register, regulate thousands of chemicals in a timely fashion. Finally, the Office of Water has a need for improved methodology for probabilistic assessment of the impact of habitat alteration on aquatic-dependant terrestrial wildlife. Thus, while the research is conducted under a Sound Science Goal, it specifically supports activities under the Aquatic Habitat goal as well as the Safe Communities/Pesticide Effects goal.


There are common threads to the Agency problems that we have identified: all three can be addressed by developing models that relate stressor exposure to effects on wildlife populations through effects on the structure and function of plan communities and ecosystems, and all three require the ability to extrapolate effects in biological scale, space, and time. We have identified three over-arching questions to guide our research. First, do changes in habitat quantity, quality, and distribution explain quantitative changes in wildlife populations? Second, what are the characteristics of habitat that are susceptible to stressors, resulting from changes in diversity, foodweb structure, and ecosystem function? Third, what is the likelihood that stressor exposure will affect non-target animal and plan species over variable spatial and temporal scales?

To address these key questions, and the Agency problems, we propose the following specific science questions:

1. How do wildlife populations respond to anthropogenic stress?

2. How do we utilize data collected at one biological level (e.g., the individual) to protect at a different level (e.g., populations, landscapes, or regions)? Specifically how can response to stressors in individuals be extrapolated to populations, communities/assemblages, ecosystems, and regions; and how can uncertainty in the response of individuals to stressors be quantified and propagated through spatial and temporal extrapolations?

3. What are the mechanisms by which stressors affect critical habitats?

4. What characteristics of structure and function of habitat affect populations of aquatic-dependent terrestrial wildlife?

5. How do interactions among stressors, or between stressors and the natural environment, influence the above questions?




Project Leader: Nathan Schumaker, Phone - 541-754-4658, E-Mail - schumaker.nathan@epa.gov

Principal Investigators:

Research Support:

The Terrestrial Habitat Project will increase EPA’s capabilities to apply the best available science to wildlife and terrestrial risk assessments. The project is developing spatially explicit models for use in evaluating the risks to wildlife populations resulting from changes in landscape structure and habitat quality. Wildlife are highly valued by society, but the viability of wildlife populations depends upon critical habitat attributes that are modified by natural forces and human activity. Current wildlife risk assessments do not adequately resolve the long-term population-level consequences of human activity. Furthermore, such assessments usually are driven by a single issue and so do not address the multiple stressors that wildlife actually encounter. This research will develop an improved risk assessment methodology that links the direct effects of pesticide and herbicide use, and other stressors, to the indirect effects associated with habitat alteration.

This project is developing models that will predict changes in landscape structure, and resulting effects on wildlife, resulting from both natural and anthroprogenic activities. This will be accomplished through the use of a number of linked simulation models. The principal models to be used in this research are listed in the table 1.

Table 1

Model Name

Model Function


Estimate future trends in primary production resulting from changes in

climate, nutrient inputs, fire regimes, and other abiotic variables.


Estimate future forest growth rates resulting from changes in primary



Estimate future forest composition and structure as a function of changes in primary production.


Estimate changes in wildlife population viability resulting from both

direct (e.g. pesticides) and indirect stressors (e.g. habitat quality).

GEM (Global Ecosystem Model), TREGRO, AND ZELIG are existing models that will be modified and/or parameterized for this study. FORCLIM and PATCH are EPA models that are being tailored specifically to this project. Collectively, this suite of models will be used to conduct spatially-explicit, place-based, risk analyses that integrate trends in environmental conditions, land use, pesticide and herbicide use, and other activities. The risk analysis methodologies developed in this project will all have a wildlife population endpoint. Our goal is to develop a robust set of spatially-explicit techniques that can be applied in a variety of locations and circumstances. While we will conduct a series of case studies to illustrate the methodologies we develop, the ultimate product of this research will be technologies that can be used by our EPA clients.

The modeling framework we will construct is illustrated in Figure 1. This analysis will integrate three classes of stessors: (a) changing environmental conditions, (b) habitat alteration, and (c) direct effects on wildlife caused by pesticide use and other human activities. Changing environmental conditions within forested landscapes will be captured by the GEM and TREGRO models, which will subsequently drive FORCLIM and ZELIG. This coupling of models will make it possible to link changes in air quality, temperature, nutrient inputs, etc. to changes in habitat quality, quantity, and pattern. Other modeling tools can later be added to project future habitat conditions in non-forested landscapes such as grasslands or desert scrub. Projections of landscape change generated this way will be combined with estimates of habitat degradation and loss to estimate future habitat conditions. These habitat-change scenarios will be coupled to other stressors (e.g. pesticide use) within the PATCH model, which will provide estimates of wildlife population viability into the future (Figure 2). The spatially-explicit nature of this approach will make it transferable to a broad range of landscapes, stressors, and species (Figure 3).



If there is an Area footer, put it here.

Local Navigation

Jump to main content.