Tools and Technology Note Using Viewsheds to Determine Area Sampled by Ground- Based Radiotelemetry THOMAS R. ETHERINGTON, 1 Central Science Laboratory, Sand Hutton, York, YO41 1LZ, United Kingdom SHELLEY M. ALEXANDER, Department of Geography, University of Calgary, 2500 University Drive NW, Calgary, AB T2N 1N4, Canada ABSTRACT Geographic Information System (GIS) viewsheds have been suggested as a possible way to determine area sampled by ground telemetry. Although this would be useful information to have, there has been little use of the technique. To investigate if viewsheds could be of use, we produced a telemetry viewshed and compared the results to previously collected radiotelemetry data. Given positive initial results and potential applications, we think GIS viewsheds could be useful for radiotelemetry studies and we encourage further research in this area. ( JOURNAL OF WILDLIFE MANAGEMENT 72(4):1043–1046; 2008) DOI: 10.2193/2006-478 KEY WORDS bias, Geographic Information Systems, radiotelemetry, sampling, terrain, viewshed. Knowing the area within which radiotagged animals can be located is valuable information because it has implications for study design, data collection, and data analysis. Considering the limitations to locating animals is partic- ularly relevant when conducting ground-based radioteleme- try in rugged terrain because topography blocks radiowaves and severely reduces the area that can be sampled. To avoid the problem of blocked radiowaves, researchers can use methods such as aerial radiotelemetry or Global Positioning System (GPS) collars. However, these techniques are more expensive, and many studies rely on ground-based telemetry to relocate animals at some time. Geographic Information System (GIS) technology has been suggested as a tool to identify the area sampled by radiotelemetry (White and Garrott 1990). Using a digital elevation model (DEM), the viewshed function of a GIS identifies the area that has a topographically unobstructed line of sight to a receiver location. Because radiowaves will be blocked by intervening topography, only areas visible to multiple receiver locations could be sampled by ground- based triangulation; these areas would form a telemetry viewshed. Although potentially useful, radiotelemetry studies have rarely used viewsheds. We wanted to investigate how useful a telemetry viewshed could be to try to stimulate research in this area. We aimed to produce a telemetry viewshed that mimicked the telemetry methods of the Eastern Slopes Grizzly Bear Project (ESGBP). Comparison of the telem- etry viewshed with ground-based and aerial-based locations of radiotagged grizzly bears (Ursus arctos) would help to indicate if telemetry viewsheds could identify the area sampled by ground-based radiotelemetry. STUDY AREA We produced a telemetry viewshed for a 396-km 2 section of the Kananaskis Valley in the Rocky Mountains, Alberta, Canada. Here narrow valleys bisect steep mountain chains, with elevation ranging from 1,374 m to 3,067 m. Vegetation cover was predominantly mature forest consisting of lodge- pole pine (Pinus contorta), trembling aspen (Populus tremuloides), subalpine fir (Abies lasiocarpa), Engelmann spruce (Picea engelmannii), and white spruce (Picea glauca). METHODS We made viewsheds using ArcGIS t Spatial Analyst (Version 9.1) with a 1:20,000 DEM with cell resolution of 30 m. ArcGIS provides full details for implementing the viewshed function. Of particular relevance to radiotelemetry are the visibility parameters (Fig. 1), which can be set to match factors such as the antenna beamwidth, signal reception range, and transmitter height associated with the telemetry equipment and methods employed. Constraining visibility in this way has not been investigated previously, but will be key in making the viewshed applicable to radiotelemetry. For ground-based telemetry the ESGBP used a hand-held antenna to locate grizzly bears (Stevens and Gibeau 2005); therefore, throughout we set the OffsetA visibility param- eter to 2 m to represent the hand-held antenna height, and we set OffsetB as 1 m to approximate the height of a transmitter on a grizzly bear. We set Vert1, Vert2, Azimuth1, and Azimuth2 to allow visibility in all directions because a hand-held antenna can be rotated and lifted through all angles. We also controlled the antenna reception range, and the positions of receiver locations to ensure the angles between bearings were realistic for triangulation. However, one set of receiver locations will not provide spacing suitable for triangulation of a transmitter at both long and short distances. As the transmitter moves closer to the road the spacing needs to be smaller. Therefore, we made 3 viewsheds with different reception ranges and receiver locations. Ideally, information on reception range would be collected as part of a testing exercise (Mech 1983, White and Garrott 1990), but without this we had to estimate the range. Because very high frequency transmitters and a 3-element Yagi antenna were used in largely forested terrain, we set 1 E-mail: t.etherington@csl.gov.uk Etherington and Alexander  Telemetry Viewsheds 1043