Land cover change and the Breeding Bird Survey: Rates of change in relation to roads.

摘要

Wildlife management decisions are often based on population estimates derived from counts at sample sites by estimating the proportion of animals counted on them. One widely used methodology for sampling avian populations across North America is the United States Department of Interior Fish and Wildlife Service's Breeding Bird Survey. Reliability of the Breeding Bird Survey data has been criticized due to surveys being conducted along road-side transects. Based on BBS assumptions that land cover is the same adjacent to roads as it is away from roads the overall goal of this study is to determine if landscape changes have occurred on a broad scale over an expanded time period and if these landscape changes have been similar between road-side and non-road-side habitats, and if so, make assumptions whether or not they affect survey results of bird population metrics. In order to meet my goal, I propose to assess three primary objectives. Objective number 1 (Chapter 2) is to evaluate rates of change in landscape across 4 time periods from 1973 to 2002 in relation to roads at a Level III Ecoregion scale. Objective number 2 (Chapter 3) is to assess the change relationship between land cover change across 4 time periods and its effects on three bird categories (grassland, forest, and urban). Objective number 3 (Chapter 4) is an assessment of habitat suitability for 4 grassland bird species over five dates (1973, 1980, 1986, 1992, 2000) in the prairie pothole regions of Western Minnesota and Northwest Iowa.;Utilizing GIS analysis I assessed land cover change per ecoregion, by buffer distance (200, 400, and 800 m), and through 4 time periods. I found that land cover change occurred similarly adjacent to roads and away from roads. Significant land cover changes occurred in pockets but no broad range trends existed across the country. Land cover change did not vary when analyzed at 200, 400, and 800 m distances from roads. Most significant land cover change could have been attributed to time period.;I assessed land cover change effects on urban, grassland, and forest birds to identify whether negative changes occurred at different rates adjacent to roads than away from roads. Grassland birds experienced some negative land cover changes; but again, these changes occurred in pockets and no broad range trends existed across the country. Forest birds experienced less prevalent negative land cover changes than grassland birds. Few ecoregions showed significant negative changes across the 48 states.;I estimated densities for 4 avian species (bobolink, grasshopper sparrow, Le Conte's sparrow, and sedge wren) present in the prairie pothole region of Western Minnesota and Northwest Iowa for 5 dates. I further assessed these densities adjacent to and away from roads. My models for density estimates adjacent to roads did not differ significantly from estimates away from roads (P = 0.8873, 95% CI 0.0857 to -0.0987).;I recommend that future research be conducted in respect to off-road point counts adjacent to BBS routes to detect discrepancies between estimates on and off-road because land cover at the ecoregion landscape scale in this study did not reveal all components that could affect avian population trends. I also recommend that further research be conducted to relate local land cover change to the landscape approach of this study. In addition, I highlight priority areas of concern for future research in respect to the Breeding Bird Survey.