GIS in Wildlife Management and Research

By: Patrick Kolar

Annotated Bibliography

For: Geo 565

Winter Term-2007

Oregon State University

The application of GIS in wildlife studies has proven to be extremely valuable. Landscape features are instrumental in determining many aspects in the lives of wildlife including their distribution and abundance. By viewing the relationships between wildlife and their habitats from a spatial and/or temporal scale, wildlife managers and researchers are able to analyze patterns and make predictive statistical models. Radio and satellite telemetry have proven to be useful tools that can be used with GIS. By plotting a point (or group of points) from telemetry as a discrete object on a map, this physical location can then be compared and analyzed in relation to other points, lines, and areas, or even to a continuous field, such as topography, land ownership, etc. This ability to collect data on an animal without physically viewing it or disturbing it and then making predictions about its behavior, habitat requirements, or locations of other animals can be very useful in studies of threatened or endangered species. These animals are typically at low population levels, are difficult to locate, and may be sensitive to disturbance. The following annotated bibliography for GEO 565 gives a few examples of how GIS can be used in the management of wildlife. Many of these articles indicate the importance of GIS use in wildlife habitat management and analysis of animal movement across landscapes, and the use of models in predicting past, current, and future wildlife needs; all personal areas of interest for me.

1-Aubry, Keith B. (2006). Geographic distribution and broad-scale habitat relations of the wolverine in the contiguous United States. Journal of Wildlife Management. In press.

With possible listing for protection under the Endangered Species Act in the near future, an understanding of not only current threats, but also historical population declines and range loss of wolverines in North America is vital information. In order to understand the current and historical distribution of an elusive forest carnivore such as wolverines, the authors compiled records from a wide array of sources since European settlement and evaluated their authenticity. By using each record as a data point in GIS with spatial and temporal precision, the authors were able to investigate broad-scale habitat relationships for three different time periods, “current” (1995-2005), “recent” (1961-1994), and “historical” (prior to 1961) based on major changes in wolverine distribution. Spatial data approximating past environmental conditions, consisting of Kuchler’s potential natural vegetation types, Holdridge’s climatic life zones, and EASE snow cover maps during the critical spring denning time period, were compared with the data points from past observations. Although wolverines seemed to be associated with a few climatic and habitat attributes, spring snow cover was the only layer to account for historical distribution patterns. Using this information, the authors were able to not only evaluate the amount of range loss to wolverines to date, but also evaluate those areas suitable for reintroductions.

2-Burns, Catherine E., K. M. Johnston, O.J. Schmitz (2003). Global climate change and mammalian species diversity in U.S. national parks. Ecology: Sept. 30 2003. Vol.100, no. 20.

With great increases in greenhouse gases expected to continue into the future, global climate change threatens to alter the current distribution of many organisms through habitat alternation via shifts in vegetation communities. National parks in the U.S., usually regarded as the sanctuaries for many species, will also suffer from these changes. This paper uses models predicting these changes in climate under a doubling of atmospheric carbon dioxide to predict vegetation shifts and likelihood of species distribution in U.S. parks. Current climate was first used to predict present mammalian species distributions using the Faunmap database and vegetation communities using VEMAP as a calibration using logistic regression in GIS. Next, a new map was created depicting future climate change and the new vegetation and mammal distributions. After overlaying the current and forecasted distributions in GIS, the number of species lost and gained was calculated by using a query for an estimate of turnover in mammal species and species richness for the parks. This future estimate will enable park service biologists to plan ahead for the loss in diversity or reshuffling of mammalian communities.

3-Dykstra Cheryl R., F.B. Daniel, J.L. Hays, M.M. Simon (2001). Correlation of Red-shouldered Hawk Abundance and Microhabitat Characteristics in Southern Ohio. The Condor: Vol. 103, No. 3 pp. 652-656.

Previous studies have suggested a wide variety of habitat characteristics to determine the difference in abundance of red-shouldered and red-tailed hawks. Because red-shouldered hawks seem to prefer mature forests near riparian areas/wetlands with surface water while red-tailed hawks prefer open areas, younger or logged forests, it is possible that a correlation exists between the two. This study used playback recordings of each hawk species during breeding season to determine presence or absence along stream corridors and estimate abundance. Landcover types were analyzed in the Spatial Analysis Extension of ArcView GIS and classified using a LANDSAT Multi-regression Land Cover data grid. By using the abundance and landcover data, statistical analyses showed that red-shouldered hawk response to the recordings was only strongly correlated to the number of small ponds within the steam corridor and no other landcover type. Although red-tailed hawks, a habitat generalist species, were not correlated with any landcover index, the two species were inversely correlated. This study provides a reliable method to classify red-shouldered hawk abundance along stream reaches based on an index of the number of ponds and also suggests that they are negatively affected by landscape changes that would benefit red-tailed hawks.

4-Gough, Mary C., and Stephen P. Rushton (2000). The application of GIS-modeling to mustelid landscape ecology. Mammal Rev: Vol. 30, No. 3& 4, 197-216.

For many members of the Family Mustelidae (the weasel family), an encompassing landscape ecology model can be very difficult to develop due to the highly dynamic nature of their different prey and the variation of spatial and temporal scales in which these animals operate. This paper reviews many different approaches for the modeling of mustelids-habitat relationships. Using surrogate characteristics in the environment important to the resource availability for mustelids, these GIS-based models analyze landscape and other features that are less variable. These models allow the data from continuous habitat areas such as land cover attributes, vegetation, and topography to be overlaid with specific point locations such as den sites, home range estimations, or foraging areas important to the species viability. By bringing landscape information together with these specific requirements, GIS allows extrapolation to larger-scale analysis that is difficult to obtain with the limited investment of purely field-based studies. Although the use of GIS-based modeling in mustelid species is relatively new, it creates an opportunity for a wide variety of management and research uses and has already been widely used to better understand the mustelid-landscape interactions.

5-Hoving, Christopher L., D.J. Harrison, W.B. Krohn, R.A. Joseph, M. O’Brien (2005). Broad-scale predictors of Canada lynx occurrence in eastern North America. Journal of Wildlife Management: Vol. 69, No. 2 pp. 739-751.

Although a number of studies have been conducted on Canada lynx in western North America, a threatened species in the lower 48 states, very little is known about them in the east. The authors in this study used spatial models in GIS in order to quantify the relationships between lynx distribution and habitat data in order to evaluate potential limiting factors in the environment. Lynx occurrences were documented in the northeast states from state and federal records including observations, trapping data, tracks, and road-kills, that were then converted into a GIS point coverage using Arc/Info 7.2.1. Other points randomly chosen by the Movement extension in ArcView 3.1 were used as a comparison for areas where lynx were not documented. Using these two sets of data points, models were created for factors considered important to lynx occurrence including snowfall, road density, bobcat density and land use/land cover. These habitat and climate variables were each modeled using a variety of GIS and statistical functions. The best of these models predicted lynx occurrence related to snowfall and deciduous forest cover very accurately. With the potential for global climate change to alter snowfall and forestry practices to change forest structure, this information can prove valuable for managers trying to restore lynx in the eastern U.S. The evaluation of potential habitat from this analysis also shows that most lies on private land straddling the border with Canada. This will further emphasize the importance of cooperation between many local and international landowners and agencies.

6-Lambert, Catherine M.S., R.B. Wielgus, H.S. Robinson, D.D. Katnik, H.S. Cruickshank, et al. (2006). Cougar Population Dynamics and Viability in the Pacific Northwest. Journal of Wildlife Management: Vol. 70, No. 1 pp. 246–254.

With the recent rise in cougar/human conflicts near urban areas, it has been hypothesized that the populations of these cats are increasing and expanding into human settlements. This widely accepted view has prompted many managers to consider higher harvest rates and more extreme measures of removal for these animals. This study analyzed data from radio-tagged cougars in the study area using the animal movement extension in ArcView GIS to estimate the minimum mean annual density of all cougars. This method, which has been used successfully in other carnivore studies, creates a 95% adaptive-kernel composite home range of all adult females using the data from all animals captured. Although the field data was not able to meet all the assumptions of this method, the density estimates do provide relative minimum values that can be used to track density among years. Data from survival and fecundity rates was estimated for the stable age distribution, or age we would expect if demographic rates stayed constant. The data from this age structure was used to suggest that the population of cougars in this area is actually declining, and that efforts should be made to reduce exploitation of the cougars, especially adult females.

7-Olson Gail S., E.M. Glenn, E.D. Forsman, J.A. Reid, et. al. (2004). Modeling demographic performance of northern spotted owls relative to forest habitat in Oregon. Journal of Wildlife Management: Vol. 68, No. 4 pp. 1039-1053.

The authors of this study were interested in developing a predictive statistical model for northern spotted owl habitat based on past demographic data in Oregon. While previous models used presence/absence data, this model focused on two primary measures of demography: annual survival rates and annual productivity. After using mark-recapture methods to first build models for survival and productivity, the authors gathered habitat data for correlation. Forest types were categorized using a digitized map of forest cover along with 1:12,000 scale orthophotos and color aerial photographs. Randomly sampled vegetation polygons were then field checked for accuracy. Next, ArcInfo GIS software was used to measure the area of each cover type around the nest and primary roosting sites. The resulting models of habitat composition for the demographic variables indicated that owls might favor those areas with a slight mix of forest types, instead of the well-documented late-successional stands. Although their habitat measures were not able to account for a few important factors, the results provide a reason to further explore this interesting conclusion that could have a profound effect on the way forests are managed.

8-Schroeder, Michael A, Aldridge, L.C. et. al. (2004). Distribution of Sage-grouse in North America. The Condor: 106: 363-376.

Previous distribution maps for Greater and Gunnison’s Sage Grouse, both currently being considered for federal listing under the Endangered Species Act, contained many inaccuracies from lack of precision to limited information on range and habitat information. These earlier maps also included areas that probably never supported sage grouse, and excluded newly discovered areas from current research. This study sought to not only create a distribution map for sage grouse of pre-European settlement based on reliable earlier maps, documents, and collected specimens, but also that of current distribution. By digitizing the data into these current maps using GIS, the authors hope analysis of many characteristics associated with sage grouse, such as habitat alteration and degradation, weather, land ownership, history, and topography, will be more readily available to managers while providing an assessment of overall habitat loss and population status.

9-Squires, John R., Todd J. Ulizio, L.F. Ruggiero, D.H. Pletscher (2006). The association between landscape features and transportation corridors on movement and habitat-use patterns of wolverines. Final Report. FHWA/MT-06-005/8171. Prepared for the State of Montana Department of Transportation. June 2006. http://www.mdt.mt.gov/research/docs/research_proj/wolverines/final_report.pdf

This study by the USFS Rocky Mountain Research Station explores how wolverines move across the landscape in response to different landscape patterns, especially major roadways. Because this species is sensitive to disturbance, has very large spatial requirements, and occur at low densities, the information on wolverine movement could hopefully be used for future highway planning. The authors used VHF and satellite transmitters placed on wolverines, as well as snow tracking individual animals, in order to characterize wolverine movement patterns relative to topography, vegetation, roads, and other landscape features. The resulting movement data was subjected to a wide array of GIS applications, such as correcting GPS error with a smoothing process in AcrView, factoring in elevation gain and loss for track distance using the ET Geowizard extension in ArcGIS 8.8, and extracting elevation, slope and aspect data from track routes on a GIS digital elevation model. Broad-scale habitat features, such as vegetation, surface roughness, and topography, were also calculated from digital elevation relief models. Vegetation layers, orthophoto quads, and hillshade topography layers were also used to identify similarities between known road-crossing areas. The results of this study indicate that wolverine movement does not appear to be sufficiently predictive to support specific recommendations in planning highway mitigation, and suggests minimizing disturbance instead of costly road improvement projects.

10-Steenhof, Karen, Mark R. Fuller, et. al. (2005). Long-range Movements and Breeding Dispersal of Prairie Falcons from Southwest Idaho. The Condor: 107: 481-496.

Little has been documented about the annual migration of Prairie Falcons, although they are well studied in both breeding and some wintering areas. Because the necessary resources for the falcons are not available year-round in the breeding area, timed migration to specific areas throughout western North America is imperative. This study used satellite transmitters to record and characterize the long-range movement patterns of radio-marked birds from the Snake River Birds of Prey National Conservation Area (NCA) in Idaho on their circular migration. Data from the transmitters of point locations throughout migration were analyzed in GIS to determine distances and direction of movements in addition to migration corridors. These estimates were used to calculate individual total migration distances, habitat, and prey selection in different seasonal use areas, and to emphasize the importance of large-scale habitat management for this species.

Email me: pskolar@yahoo.com

Or

View my photo gallery: http://www.pbase.com/pskolar