You are here:
LANDSCAPE METRICS THAT ARE USEFUL FOR EXPLAINING ESTUARINE ECOLOGICAL RESPONSES
Citation:
Hale, S S., J F. Paul, AND M. C. Nicholson. LANDSCAPE METRICS THAT ARE USEFUL FOR EXPLAINING ESTUARINE ECOLOGICAL RESPONSES. Presented at 19th Annual Symposium of the United States Regional Association of the International Association for Landscape Ecology, Las Vegas, NV, March 30- April 3, 2004.
Description:
We investigated whether land use/cover characteristics of watersheds associated with estuaries exhibit a strong enough signal to make landscape metrics useful for predicting estuarine ecological condition. We used multivariate logistic regression models to discriminate between sub-estuaries of Delaware and Chesapeake Bays with low benthic environmental quality (BEQ, defined by chemical parameters) and those with high BEQ. A model using percent wetlands and the sum of all human uses correctly classified BEQ in 86% of the cases. Similar models discriminated between estuaries with a high index of benthic community condition and those with a low index. The benthic index model showed excellent discrimination and correctly predicted the benthic index of an independent data set 79% of the time (p < 0.05). Watersheds with higher percentages of urban and agricultural land uses were associated with low benthic index while those with higher percentages of wetlands were associated with a high index. Land use/cover metrics in riparian zones proved useful for both models, as did whole-basin metrics. Other landscape metrics that showed promise in explaining estuarine condition were percent impervious area, patch contagion, and metrics associated with watershed discharge and estuarine water residence time.