PARENT SESSION
Contributed Oral Session 64: GIS / Remote Sensing and Landscape Ecology
Tuesday, August 9, 1:30 PM - 5:00 PM, Meeting Room 519 B, Level 5, Palais des congrès de Montréal
Using a landscape-scale approach to understanding the phosphorus transfer system at multiple scales.
Dorioz, Jean-Marcel*,1, Wang, Deane2, 1 Institut National de la Recherche Agronomie, Thonon-Les-Bains, France2 Rubenstein School of Environment and Natural Resources, Burlington, VT, USA
ABSTRACT- Conceptualizing the problem of phosphorus in the environment as a biosocial transfer system operating at multiple scales has led us to ask a variety of questions about possible differences in processes and dynamics at several scales. Clearly, the information available at difference scales from individual fields, to small sub-watersheds, to mid-sized watersheds, to large basins varies enormously and thus constrains our ability to describe processes and patterns at these scales. Our work to collect data in the Lake Champlain Basin (Vermont, New York, Quebec) and the Basin of Lake Leman (France and Switzerland) suggests that variability in P transfer remains high at all scales reflecting variation in both how individuals practice land management (agriculture, forestry, urban/suburban landscape management), and how this social relationship to the land might change over larger areas as topography, geology, economy, and cultural practices vary. Using multiple regression models to fit P export estimates to major land uses (e.g., agriculture, urban, wetland, with adj. R2 of .88) in mid-sized watersheds (3,000 to 21,000 ha) reveals that average deviations from model are about 21% of predicted P exported with the maximum deviation of 55%. Using the same approach to look at an individual watershed (14,000 ha with sub-watersheds ranging in size from 150 to 1400 ha) reveals similar average deviations from the predicted model value (adj. R2 of .87). If spatial variation was due to field-scale variation in the small sub-watersheds, then the larger area aggregation in watersheds, should reduce variation in P export at that scale. The ability of these kind of observations to identify higher than average P emission sites at multiple scales provides both land managers and scientists with a spatial tool to target management resources and research infrastructure at these higher emission sites. Because these sites may produce the bulk of the P exported, this spatially targeted approach can result in better understanding of problem areas for P emission from both an ecological and social perspective as the P transfer invariably results from an interaction of the two. The changing role of urban development in P transfer as the area of study increases also becomes evident as we compare export models at multiple scales.
Key words: phosphorus, watershed, regression, model