Linking Spatial Variations in Water Quality with Water and Land Management using Multivariate Techniques.

摘要: Most studies using multivariate techniques for pollution source evaluation are conducted in free-flowing rivers with distinct point and nonpoint sources. This study expanded on previous research to a managed “canal” system discharging into the Indian River Lagoon, Florida, where water land management is single most important anthropogenic factor influencing quality. Hydrometric use data of four drainage basins were uniquely integrated analysis 25 yr monthly quality collected at seven stations determine impact spatial variability Cluster (CA) classified monitoring groups (CA groups). All parameters identified by discriminant showed patterns among CA groups. Two-step principal component analysis/factor (PCA/FA) was (i) alone (ii) conjunction rainfall, flow, data. The results indicated that PCA/FA unable identify factors associated activities. addition hydrometric revealed close associations nutrients color storm-water retention pasture citrus lands; total suspended solids, turbidity, NO 3 + 2 flow Lake Okeechobee releases; specific conductivity supplemental irrigation supply; dissolved O wetland preservation. practical implication emphasizes importance basin-specific ongoing pollutant loading reduction ecosystem restoration programs.