Relationships between landscape characteristics and nonpoint source pollution inputs to coastal estuaries

Land-use activities affect water quality by altering sediment, chemical loa ds, and watershed hydrology. Some land uses may contribute to the maintenan ce of water quality due to a biogeochemical transformation process. These l and-use/land-cover types can serve as nutrient detention zones or as nutrie nt transformation zones as dissolved or suspended nutrients or sediments mo ve downstream. Despite research on the effects of individual land-use/land- cover types, very little has been done to analyze the joint contributions o f multiple land-use activities. This paper examines a methodology to assess the relationships between land-use complex and nitrate and sediment concen trations [nonpoint source (NPS) pollutants] in streams. In this process, se lected basins of the Fish River, Alabama, USA, were delineated, land-use/la nd-cover types were classified, and contributing zones were identified usin g geographic information system (GIS) and remote sensing (RS) analysis tool s. Water samples collected from these basins were analyzed for selected che mical and physical properties. Based on the contributions of the NPS pollut ants, a link-age model was developed. This linkage model relates land use/l and cover with the pollution levels in the stream. Linkage models were cons tructed and evaluated at three different scales: (1) the basin scale; (2) t he contributing-zone scale; and (3) the stream-buffer/riparian-zone scale. The contributing-zones linkage model suggests that forests act as st transf ormation zone, and as the proportion of forest inside a contributing zone i ncreases (or agricultural land decreases), nitrate levels downstream will d ecrease. Residential/urban/built-up areas were identified as the strongest contributors of nitrate in the contributing-zones model and active agricult ure was identified as the second largest contributor. The regression result s for the streambank land-use/land-cover model (stream-buffer/riparian-zone scale) suggest that water quality is highest when passive land uses, such as forests and grasslands, are located adjacent to streams. Nonpassive land uses (agricultural lands or urban/built-up areas) located adjacent to stre ams have negative impacts on water quality. The model can help in examining the relative sensitivity of water-quality variables to alterations in land use made at varying distances from the stream channel. The model also shows the importance of streamside management zones, which a re key to maintenance of stream water quality. The linkage model can be con sidered a first step in the integration of GIS and ecological models. The m odel can then be used by local and regional land managers in the formulatio n of plans for watershed-level management.