The Wildlife Habitat Response Model : A Tool for Estimating Terrestrial Wildlife Habitat Responses to Fuel Treatments

This paper describes the Wildlife Habitat Response Model (WHRM), a web-based computer tool for evaluating potential effects of fuel reduction projects on terrestrial wildlife habitats in dry coniferous forests of the western United States. WHRM uses species-habitat associations to predict how fuel treatments may affect species habitat suitability. Users input the amount of change in forest floor components, down wood, standing dead wood, understory vegetation, and live trees based on fuel treatment objectives, desired future conditions, or predictions from forest stand development models. Using data gleaned from the scientific literature, WHRM identifies the habitat elements that are important for species in terms of reproduction, food acquisition, and shelter from predators and environmental extremes. Managers can then enter variations of proposed fuel management activities into the model to determine how alternatives may influence these habitat elements. The output of WHRM is strictly qualitative (for example, will management activity result in a positive, negative, or have no affect on habitat conditions for a species). The model output and background information can help managers compare alternatives and develop the environmental consequences components of NEPA-type assessments for fuel treatment activities. DRAFT WHRM DOCUMENTATION (CURRENTLY IN REVIEW) PLEASE DO NOT CITE OR REFERENCE, SUBJECT TO CHANGE 8/22/2005 2 Introduction Managers face a difficult task in predicting the effects of fuel treatments on wildlife within the dry interior forests of the western United States. Few empirical studies are available from which to draw inferences and thus there is uncertainty and some concern about how sensitive wildlife species will respond to mechanical thinning, prescribed burning and their alternatives (Pilliod and others, in press). Anecdotal observations suggest that wildlife mortality during forest thinning and prescribed burning operations is minimal and inconsequential to populations of terrestrial species (Folk and Bales 1982; Komarek 1969). Populations are more likely to respond to the rapid changes and sometimes prolonged recovery in forest structure and composition that result from fuel management activities. When empirical studies on the effects of fuel treatments on wildlife habitats are unavailable, some predictions may still be possible by first identifying the habitat requirements of a species and then estimating how fuel management activities will alter the habitat elements that are important to a species’ survival and reproduction. The Wildlife Habitat Response Model (WHRM) is a webbased predictive computer tool that was developed to meet this need. WHRM is based on species-habitat relationships and organized similarly to an envirogram. Species-habitat relationships are descriptive mechanisms used by wildlife biologists and managers for linking species with the habitats in which they are associated. Typically organized in matrices and databases, species-habitat relationships can be general (for example, species-biome) or specific (for example, species-habitat element) and can have qualifiers on habitat usage such as resident, seasonal, occasional, breeding, foraging, and other descriptive terms. An envirogram is a graphic representation of causal relationships linking indirect to direct causes of species responses to the environment (Andrewartha and Birch 1984). Each causal pathway is listed under of one of four categories: mates, resources, predators, and mal-entities. These four categories represent the general habitat requirements of a species and can be described as those features or habitat elements that are needed for an animal to (1) find a mate, reproduce, and successfully rear offspring (for example, breeding sites, birthing areas, and nest sites), (2) acquire the nutrition necessary for survival and reproduction (for example, foraging habitat, forage, and habitat for prey which influences prey availability), (3) escape predation, and (4) seek shelter from environmental hazards (for example, daytime or nighttime temperatures, extreme weather events, seasonal climate fluctuations, and unpredictable disturbances such as drought, fire, or flooding). The goal of this paper is to provide a user’s guide to the Wildlife Habitat Response Model. In the following sections we describe the concept, inputs, outputs, utility, limitations, and assumptions of the model. In chapter XI of this document, we provide a case-study of how WHRM can be used for fuel treatment planning. Description The goal of our modeling effort was to produce a straightforward tool to help fuel planners and NEPA specialists (that is, non-biologists) to conceptualize and qualitatively predict how fuel reduction treatments might affect the habitats of terrestrial wildlife species living in dry, coniferous forest ecosystems of the western U.S. Hence, WHRM is intended for planning purposes and general assessments and is not intended to replace detailed population-level assessments of threatened, endangered, or sensitive species. DRAFT WHRM DOCUMENTATION (CURRENTLY IN REVIEW) PLEASE DO NOT CITE OR REFERENCE, SUBJECT TO CHANGE 8/22/2005 3 WHRM was designed to predict potential habitat suitability relative to pre-treatment conditions and can be used to compare the potential effects of different treatments on a species’ habitat. Habitat suitability within a given area is based on the availability of habitat elements required for successful reproduction, food acquisition, predator avoidance, and shelter from environmental hazards and stresses. In essence, WHRM packages a wildlife-habitat database into a user-friendly computer-based tool. WHRM first identifies the habitat requirements of a chosen species and then allows users to investigate how proposed fuel management activities and alternatives may influence the critical habitat needs of a species. The predicted response of a species to proposed habitat changes is based on species-habitat associations reported in the scientific literature for ponderosa pine (Pinus ponderosa) and dry-type Douglas-fir (Pseudotsuga menseizii), lodgepole pine (Pinus contorta) and mixed conifer forests in the western U.S. (see map and definition, pages xx-xx, this document). Therefore, WHRM predictions are specific to these forest types and may not apply to a species across its range. Furthermore, WHRM predictions are most appropriate within the stand being modeled. Predictions have no temporal scale, but users can produce time-specific predictions based on the values selected for changes or recovery in each habitat element. Predictions from forest stand development models, such as the Forest Vegetation Simulator with the Fire and Fuels Extension (FVS-FFE) can aid in predicting how various habitat elements may change over time. Species-habitat associations used in this model were generated from an extensive search of over 450 peer-reviewed published articles on wildlife and wildlife habitats. In most cases, we used original data from papers published in scientific journals. For each paper, we recorded the location of the study, habitats investigated, whether the paper was a disturbance paper (in other words, describing habitat associations in a disturbed or recovering environment), species studied, the response variable used (for example, occurrence, abundance), a description of the habitat elements with which a species was significantly correlated (probability less than 0.05 or 0.10, depending on the usage in the paper), and the direction of the correlation. Habitat elements were then placed into standard categories (see table 1). If the habitat categories described in a paper and those used in WHRM did not match exactly, we used a liberal, inclusive approach. For example, if a paper reported snags in the size class 12 to 24 inches diameter-at-breastheight (d.b.h.), we used size classes 10 to 19 inches and 20 to 29 inches d.b.h. for WHRM. If size classes were not provided in a paper, we used a general category (for instance, snags (size not specified)). Based on information reported in a paper, we placed each habitat association described for a species into one of four categories: reproduction, non-consumptive foraging habitat, forage or prey habitat, and shelter from predators or environmental extremes. Some habitat associations fell into two or more categories depending on the life history of the animal. If not specified in a paper, we categorized remaining habitat associations based on field guides, species accounts, and other general references. Data from papers outside the target area were not included in WHRM unless no other information was available. Under such circumstances, we first included papers from similar forest types (for example, moist, coastal Douglas-fir) within the region and then other forest types (for example, spruce). DRAFT WHRM DOCUMENTATION (CURRENTLY IN REVIEW) PLEASE DO NOT CITE OR REFERENCE, SUBJECT TO CHANGE 8/22/2005 4 Table 1. List of habitat elements important for wildlife. Definitions of habitat elements adapted from O’Neil and others (2001). Units are not displayed because WHRM uses percent change from pre-treatment conditions and thus any units can be used. Forest Component Habitat Elements Definition Bare Mineral Soil Exposure The inorganic soil layer beneath the humus. This HE is usually expressed as percent exposure in a given area representative of the stand. Although not generally considered a wildlife HE, bare mineral soil represents a lack of duff and litter cover (1-(duff+litter cover)) and therefore is indicative of poor habitat quality for some species and increased erosion potential. Duff Cover The matted layer of organic debris beneath the litter layer. Decomposition is more advanced than in litter layer; intergrades with uppermost humus layer of soil. This HE is usually expressed as percent cover of a given area representative of the stand. Duff depth can also be important for wildlife habitat, but is not included. Grass Cover The amount of ground cover composed of grasses. This HE is usually expressed as percent cover of a given area representative of the stand. Forb/Herbaceous Cover The cover of understory, non-woody vegetation layer beneath the shrub layer that includes forbs, mosses, and ferns. This HE is usually expressed as percent cover of a given area representative of the stand. Shrub Cover A measure of shrub density and usually visually estimated as a vertical projection of shrub crown diameter onto the ground. Shrubs are further subdivided into short shrubs (0-18” tall) and tall shrubs (>18” tall). Tall shrubs are usually considered ladder fuels whereas short shrubs are not. If shrub height is not specified in a study, it is listed under "shrub cover (all size classes)". Shrub species and number of shrub canopy layers are important for wildlife habitat, but are not specified. Forest Floor Understory Vegetation DRAFT WHRM DOCUMENTATION (CURRENTLY IN REVIEW) PLEASE DO NOT CITE OR REFERENCE, SUBJECT TO CHANGE 8/22/2005 5 Forest Component Habitat Elements Definition Litter Cover The upper layer of loose, organic (primarily vegetative) debris on the forest floor. Decomposition may have begun, but components still recognizable. This HE is usually expressed as percent cover of a given area representative of the stand. WHRM does not specify litter depth, but this can influence