PARENT SESSION
Oral Session - Quantitative Relationships Between Landscape Processes and Patterns and Wildlife - Afternoon Session Chair(s): Robinson, Vincent¹, ¹ University of Toronto at Mississauga, Mississauga, ON, Canada
Friday, April 2, 2004 1:00 PM - 2:40 PM Apollo Room 1-2

Choice of a classification scheme can lead to ambiguous relationships between landscape metrics and population processes. ^*HOWERTER, DAVID W. , ¹ Institute for Wetland and Waterfowl Research, Stonewall, MB, Canada

ABSTRACT- Landscape ecologists often attempt to elucidate relationship between the spatial patterning of habitats and vital rates of wildlife species. Often, thematic mapping products from remotely sensed images form the basis for these analyses, and a wide variety of metrics have been developed to quantify patch- and landscape-level spatial heterogeneities. Much previous work has been done examining how landscape metrics vary with changing grain and extent of thematic-mapping products; however, little previous research has examined the ramifications of choice of a map classification scheme on landscape metrics and, hence, on relationships between metrics and population processes. The problem of choice of a map classification is particularly acute when attempting to model landscape effects on habitat generalists. In this paper, I provide a case study using a large sample (n > 3,600) of nests from upland-nesting ducks collected at 15 sites throughout the aspen parklands of south-central Canada. Ducks in this naturally patchy environment nest in a variety of vegetation types and, similarly, predators of duck nests forage in multiple vegetation types. A precise definition of what constitutes a habitat patch, therefore, becomes difficult. Thus, I examined a variety of habitat classification schemes. Depending on the scheme selected, relationships between hatching rates and individual landscape metrics appeared as positive, negative, or neutral. This ambiguity has obvious practical ramifications for researchers and habitat managers alike. Information-theoretic model selection techniques may help identify a parsimonious habitat classification scheme(s) that relates well to the vital rates of interest

KEY WORDS: vital rates, map classification, aspen parkland, model, nests