Remote sensing and landscape analysis of indicators of estuarine condition in the northeastern United States.

摘要

I focused on four questions: (1) are measures of estuarine water quality the same across study regions; (2) are spatial setting variables and intense rainfall events important in explaining variation in water quality measures; (3) what is the relationship between patterns of land cover and estuarine condition; and (4) is land-cover change a good regional indicator of estuarine condition? Estuarine data was extracted from the Environmental Monitoring and Assessment Program (EMAP) and Coastal-2000 surveys. Geographical Information System's modeling, land cover change detection, cumulative distribution functions and multiple linear regression analyses were used to quantify land-water interactions in three regions representing urban and agricultural gradients along the Virginian Biogeographic Province in the eastern coast of the United States. Spatial setting variables (riverine, lagoonal, oceanic) plus rainfall accounted for the effect of location on measurements of estuarine condition. After removing spatial setting effect from estuarine condition data, weak but significant relationships were found between total urban area (within a 4-km buffer from measuring stations) and water quality indicators for the three regions. The effect of spatial setting on indicators of estuarine condition was strikingly high in Delmarva, where distance-to-shore, and oceanic and riverine categories significantly explained variability of sediment organics (Fluorant: R2 = 0.79; P 0.001; Dieldrin: R2 = 0.91; P 0.001; PCBs: R2 = 0.96; P 0.001). Lagoonal locations also had a significant negative effect on arsenic, cadmium, and lead, and explained between 32 to 53% of their variation. Land-cover change on estuarine condition on each region showed clear differences in patterns between Delmarva and the other two areas during the 1984–92 period. In Delmarva, land cover change in buffer zones 6-, 8-, and 10-km from monitoring stations had a significant effect on the Benthic Index, silver, arsenic, cadmium, and mercury. In Cape Cod, land cover change in the same buffer zones explained approximately 15% of the variation in PAHs, dissolved oxygen, and silver. There was no effect of land cover change between 1992–99 in estuarine condition in Cape Cod and Delmarva, and significant effect (R2 = 0.38; P 0.05) in the Hudson/Raritan region.