Proceedings
4th Microcomputer Applications in 
Fish & Wildlife Conference
October 24-27, 1999
Stateline, Nevada

 Habitat Predictability Model for Willow Flycatchers (Empidonax trailii) in Northern California, using Landsat Thematic Imagery

Christopher J. Stermer*, California Department of Fish and Game, cstermer@thegrid.net

Timothy S. Burton, California Department of Fish and Game

Richard L. Callas, California Department of Fish and Game

Dr. Lawrence Fox III, Humboldt State University

We conducted field surveys for willow flycatchers (Empidonax traillii) in 1997 and 1998, from June 15 through July 31, within the McCloud Flats in Siskiyou County, California. We used Landsat Thematic imagery to predict known and potentially suitable habitat for willow flycatchers. Vegetation classes that occurred in areas occupied by singing male willow flycatchers (territories) were used to train the imagery to identify potentially suitable habitat. National Wetland Inventory data, Digital Elevation Models, Slope gradient data, and stream layers were used as additional variables to identify potentially suitable sites. Seventy-seven willow flycatcher territories were found during our surveys. Habitat data were collected on .07 ha circular plots centered on sixty-six territories located in 1997 to characterize habitat at those sites. Riparian thickets > 2 m in height was the most abundant vegetation type, making up 53% of the vegetation within the plots. Twenty-one percent of the vegetation was a composite of live green grasses and forbs. Most (94%) of the territories were located within riparian wet meadow systems, characterized by patchy clumps of riparian thickets (predominantly willow spp.) and were associated with active streams flowing along slope gradients of < 5º. The remaining territories (6%) were found in a large wet meadow system (Bigelow Meadows), with patchy clumps of willows, saturated soils, along slope gradients of < 5º, and was not directly associated with an active stream. A habitat predictability model was developed using: (1) vegetation classes occurring within willow flycatcher habitat; (2) association of habitat to streams; (3) stream gradients; (4) landscape gradients; and (5) distribution of vegetation classes. An accuracy assessment indicated the model was 94% correct in predicting habitat potentially suitable for willow flycatchers within are study area. We concluded that Landsat Thematic imagery, when applied in conjunction with other landscape data, is an effective technique to identify willow flycatcher habitat over large landscapes.

Bernie McConnell* and Mike Fedak, Sea Mammal Research Unit, Gatty Marine Lab, St Andrews University, St. Andrews, Fife, Scotland, b.mcconnell@smru.st-andrews.ac.uk

The habitats and prey of both seals and fisheries may often overlap, thus providing a potential for competitive interactions. To develop management plans to minimize such interactions it is vital to know the foraging patterns of both ‘predators’.

We have developed Satellite Relay Data Loggers (SRDLs) to follow the movements and behaviour of seals. These telemetry devices record depth and speed and use the Argos satellite system to provide location and to relay detailed dive information. Because seals spend much of their time underwater and satellite availability is limited, the effective data rate to the satellite is low. Thus, much effort has gone into optimising the behavioural parameters transmitted, while maintaining high precision and low sampling bias. Location filters have been developed which reduce the Argos location error.

Successful deployments of many hundreds of SRDLs has lead to the, not unwelcome, problem of assimilating and interpreting large and complex temporal/spatial datasets in an ecological framework. Thus, we have developed the MAMVIS visualisation system. Based on the UNIX AVS application, MAMVIS extracts data from an Oracle database and allows the animation of track and dive behaviour within a 3-D ecological landscape. The database is also interfaced to ArcView and SAS.

We present here, as a case study, an investigation of the effects of a sandeel (Ammodytes) fishery on grey seals (Halichoerus grypus) in the North Sea, UK. We also discuss the problems inherent in extrapolating from a sample of study animals to the behaviour of populations.

Brian L. Pierce*, Scott C. Pettengill, and John T. Baccus, Southwest Texas State University, San Marcos, Texas

Line-transect sampling methodology has been criticized for the additional time required to collect perpendicular distance data and for the inaccuracy of distance estimates made for each sighting, particularly on non-linear transects. We demonstrate that new methodology and recent technological advances in DGPS, GIS, and laser range finding equipment can be applied to existing line transect theory in order to meet current management requirements. Specifically, new spotlight line-transect sampling methodology can be used to provide a more accurate estimate of population size while obtaining herd composition and spatial location data for the surveyed species. The method is specifically designed to be robust to changes in visibility which limits use of the typical spotlight strip transect method. Our method for spotlight line-transect sampling is relatively fast (0.15 ± 0.05 hours/Km SD), spatially accurate to within the limits of the ranging device and GPS used (15.2 ± 13.9m SD), consistently obtains larger sample sizes per transect (13.7 ± 9.6% beyond 200m), and returns more information per sighting (count, composition, and location) than traditional spotlight strip-transect sampling methodology. We provide results from a two year study of white-tailed deer (Odocoileus virginianus) on the Edwards Plateau region of Texas, where dense stands of Ashe Juniper (Juniperus ashei) detrimentally affect the traditional spotlight strip-transect sampling method. In addition to estimates of animal density and composition obtained from the spotlight line-transect surveys, the locational data obtained for each sighting are applicable for determining home range selection within the surveyed area (Johnson 1980) without identification of individuals (Design 1, Thomas and Taylor 1990). Our results indicate that the criticisms of increased time requirements and relative inaccuracy of perpendicular distance estimates for non-linear transects are invalid when viewed in terms of the equivalent sample size (hrs/deer/Km) and utility of the additional data obtained (habitat selection). As such, the paradigm for management of white-tailed deer can be improved through application of the new spotlight line-transect methodology which alleviates bias caused by failure to meet strip transect sampling assumptions, yields greater accuracy, provides additional data usable in species management, and is more efficient in terms of cost per unit effort.

David Ganskopp, USDA-ARS, HC-71, Burns, Oregon, david.ganskopp@orst.edu

Livestock trails frequently evolve in pastures when plant growth or establishment can not keep pace with vegetation disturbance. In some instances man-made trails are established in rangeland settings to encourage uniform use of forages or facilitate livestock passage through dense vegetation or rugged terrain. A long-term assumption has been that livestock establish pathways of least resistance between frequented areas of their pastures, but this hypothesis has never been tested. We mapped cattle trails in three 800+ ha pastures with global positioning units. A geographic information system (GIS) helped quantify characteristics of trails and the landscape and was used to plot least-effort pathways between water sources and distant points in the pastures. Characteristics of the cattle trails and pathways were compared to test the hypothesis that cattle develop least-effort routes of travel. Mean slope of pastures was 13.5%, and slope of the areas traversed by cattle trails was 8%. Actual slope of the trails was reduced to 5.2% by use of cross-slope routes. On average, cattle trails were 11% shorter (P=0.046) than least-effort pathways, and the topography traversed by cattle had a gradient about 1% less than least-effort pathways (P=0.02). Actual slope of the trails (5.5%) and pathways (5.6%) were similar (P=0.74), however, and analyses of values extracted from cost-surfaces revealed the effort to traverse trails (183) and pathways (170) were similar (P=0.07). We accepted our hypothesis that cattle establish least-effort routes between distant points, and suggest that GIS software may be useful for designing systems of livestock trails in rugged terrain.

Dr. Margaret Katherine Trani (Griep), USDA Forest Service, Roanoke, Virginia, mgriep/r8_gwjeff@fs.fed.us

R8BIRD is an Oracle system designed to facilitate the storage and analysis of information on southeastern migratory and resident avian populations and their associated habitats. The system manages information collected from multi-year breeding season surveys conducted under the Southern National Forests Landbird Conservation Strategy. R8BIRD provides a tool for assessing species demographics, temporal and spatial dynamics of regional populations, and local area diversity. Used for forest planning and analysis to identify species requiring conservation, the system can be used by other agencies and organizations involved in monitoring efforts. The system complements the monitoring process and serves as a useful tool for other forest analyses.

R8BIRD operates on the IBM RS 6000. Oracle forms (data entry screens) are provided for recording data in a standardized format. These forms include vegetative characteristics and habitat components (e.g., physiographic area, successional sere, hydrological characteristics, basal area, forest attributes); historical management activities; and bird survey data (e.g., AOU code, census time interval, sighting distance). Each form directly corresponds with the field sheets used on 14 national forests.

There are several edit tools within each of the Oracle forms. These tools allow for database queries, field editing, record scrolling, and other options. Many menu commands have equivalent button bar shortcuts. The system also contains look-up tables and drop-down windows.

R8BIRD also includes several standardized reports. These include relative abundance, species dominance and rarity, frequency distribution, and validation monitoring. Each report provides for a variety of output selections (filing, printing, and mailing report results).

Dvornich, K.M., Washington Cooperative Fish and Wildlife Research Unit, School of Fisheries, University of Washington, Seattle, WA

Grover, T.M.*, GIS and Simulation Lab, Washington State University, Spokane, WA

NatureMapping is a program initiated in 1993 to create a "public layer" of GIS wildlife information for the state Gap Analysis Project. The technological component has evolved through the Middle and High school education module developed to incorporate data collection into environmental and wildlife education. Science teachers are trained in GPS, GIS, database software, and spreadsheets to format student collected data and to standardize output for later analysis at the University of Washington. Workshops are intense hands-on experiences beginning with an example of Global Positioning System data collection. Site data are then brought into the computer lab and entered into NatureMapping data software and exported into ArcView GIS where they are linked to relevant information for comparison against habitat classification themes, digital ortho-photography, and satellite imagery. Finally, map layout and generation are demonstrated and each teacher produces a map. The teachers are encouraged to spatially analyze the data in the classroom and use the environmental context to discuss societal issues relating to wildlife habitat. Students and teachers contribute a significant amount of data to research projects while gaining valuable technological skills and heightened sensitivity to wildlife.

Examples of NatureMapping maps from the workshops will be demonstrated.

Jeffrey B. Trollinger* and Karen K. Reay, Virginia Department of Game and Inland Fisheries, Richmond, Virginia, kreay@dgif.state.va.us, jtrollinger@dgif.state.va.us

The lack of a complete and ongoing inventory of Virginia’s wildlife resources demands creative solutions. WildlifeMapping, a two-year-old citizen outreach and education program, helps fill in the ‘gaps’ by providing an opportunity for citizens, community groups and schools to contribute their wildlife observations to the Virginia’s biological databases. By providing individuals with rudimentary training in species identification and field guide use, habitat identification, map reading, and data reporting, the WildlifeMapping program opens a dialogue between the Department and the public. Virginia’s WildlifeMapping program, sponsored by the Virginia Department of Game and Inland Fisheries (VDGIF), is modeled after NatureMapping, a national outreach project developed by the University of Washington Gap Analysis Project and the Washington Department of Fish and Wildlife. To date, VDGIF personnel and 68 volunteer facilitators have trained approximately 400 WildlifeMappers across Virginia with over 3,500 observations reported to the Agency. These observations are maintained as a public data layer and used to assess wildlife distributions by filling in the ‘gaps’ where little data collection has taken place. A demographic sketch of WildlifeMappers and facilitators will be described along with an outline of data received back. A description of where data reporting sites are across the Commonwealth will be mapped. A comparison of species observation maps before and after incorporating WildlifeMapping data will demonstrate how important these data are in providing a more comprehensive view of wildlife distributions across the Virginia.

Distributional Information on Herpetofauna and the WildlifeMapping Project: Making the Connection

Karen K. Reay* and Jeffrey B. Trollinger, Virginia Department of Game and Inland Fisheries, Richmond, Virginia, kreay@dgif.state.va.us, jtrollinger@dgif.state.va.us

Amphibians and reptiles have traditionally been two groups of species that have limited distributional information. Their secretive habits combined with a lack in popular observation clubs, such as birding groups, make it difficult to collect information on wide-spread distribution patterns. Range and distribution patterns for some species are almost complete, whereas others are very incomplete. Maps of the Atlas of Amphibians and Reptiles in Virginia illustrate where people, mainly scientists, have documented a species’ occurrence with an accepted museum specimen or scientific collection record. These maps show that a particular species may have a broad ecological tolerance limit and thus a wide distribution, or relatively narrow tolerance limit and a geographically limited distribution. The Atlas also indicates where habitat favorable to reptiles and amphibians may be lost due to rapid expansion of urban centers. However, it mostly illustrates where there are gaps in collection data due to lack of surveying effort such as the Piedmont region of the state. It is the hope of the Virginia Department of Game and Inland Fisheries that the WildlifeMapping Project can assist in this effort by providing data from areas not frequently surveyed. This public surveying effort may be done through consistent weekly monitoring or records of notable sightings. With time, the volume of data generated from this volunteer effort will provide valuable insight into the distribution patterns of an elusive, but important, group of animals within Virginia.

Mark Swan, The Nature Conservancy, mswan@tnc.org

Numerous entities have expressed an interest in using datasets created by The Nature Conservancy (TNC) for its Mississippi Alluvial Plain Ecoregion Project. In order to disseminate these data and their documentation (metadata) efficiently, The Louisiana office of TNC secured a grant from the FGDC (Federal Geographic Data Committee) and the National Spatial Data Infrastructure (NSDI) to establish a Clearinghouse Node on the World Wide Web (Internet). Given that TNC has several GIS shops world-wide, this may be only the first of several TNC Clearinghouse Nodes that will comprise NSDI. A TNC representative (Louisiana's GIS Coordinator) will discuss NSDI, FGDC, metadata, and reasons for using/creating Clearinghouse Nodes. He will describe the status of the TNC Node as of Sep 1999. He will relate lessons from creating the node, provide an appraisal of how the node is used, and provide advice for NSDI collaboration.

Scott Wilson, Electronics Engineering, USGS - National Wetlands Research Center, Lafayette, Louisiana

Since colonial times 80 percent of this America’s forested wetlands have been converted to other uses. Remaining forested tracts are highly fragmented. These disconnected forests have impacted species that depend on forested wetlands, especially those species that require large continuous areas of forest to survive, including the black bear and some species of Neotropical migrant songbirds. The majority of forested wetlands remaining within the Lower Mississippi River Valley are privately owned. If species that require large areas are to survive within the Lower Mississippi River Valley, then private stewardship is essential.

To assist in the ongoing conservation of forested wetlands in the Lower Mississippi River Valley, the U.S. Fish and Wildlife Service partnered with the USGS- National Wetlands Research Center to develop a multimedia CD-ROM entitled "Black Bears and Song Birds of the Lower Mississippi River Valley." The CD-ROM was primarily designed to assist educators at the middle school level to teach conservation education, but the informational content also applies to general public outreach and education. The CD-ROM is divided into three sections. The first section contains a basic introduction to forest-dwelling songbirds, including a detailed review of nesting requirements. The second section is a basic introduction to the Louisiana Black Bear with a detailed discussion on denning and dispersal requirements. The third section introduces forest fragmentation and its effects on bears and songbirds, as well as ongoing incentive programs that have been effective in reforesting privately owned lands.

The CD-ROM has been distributed to every middle school in Louisiana, and additional copies will be made available to educators through regional workshops and in-service training throughout Louisiana. Copies have also been distributed to U.S. Fish and Wildlife field offices within the Lower Mississippi River Valley.

NOTE: * indicates presenter