Intensive and widespread agricultural activity often is associated with elevated particulate loads and other indicators of poor water quality characteristic of degraded stream habitats. I monitored 29 sites in the River Raisin watershed, southeastern Mic higan, USA, during low-flow conditions to assess the impact of land use on water quality. Discharge, total suspended solids (TSS), turbidity, conductivity, dissolved oxygen, and pH were measured once during each season in hydrologic year 1993. Measures of particulates and ions showed marked downstream increases in the mainstem, and discharge and drainage area accounted for most of the variance. Although some influence of land use could be seen in analysis of regression residuals, land use did not consi stently explain a significant amount of remaining variation. Weaker relationships were found between discharge and drainage area with measures of particulates and ions in basin tributaries. Land use explained a greater proportion and sometimes a signifi cant amount of residual regression variance among tributaries in concentrations of particulates and ions. However, the influence of land use on water quality during non-storm conditions could not be consistently demonstrated throughout the year.

 Stormflow conditions were monitored in two adjoining small catchments within the River Raisin watershed that share similar drainage area, slope, geology and soils, but differed markedly in their history of channelization and percent agricultural land use, particularly within the riparian buffer zone. Measures of suspended particu lates and turbidity during storm events were significantly higher in the more agricultural catchment during fall, winter and spring. Conductivity was likewise higher except during winter, when rain on frozen ground and snow resulted in significant dilution of dissolved ions in both tributaries. Stormflow conditions yielded information on particulates and ions that reflect the effects of human land use activities such as agriculture on stream water quality. The inclusion of more sample sites, particularly in basin tributaries, more frequent sampling, and the inclusion of data under stormflow conditions, should be part of future water quality sampling protocols in accessing the impact of human landscape activ ities upon stream water quality.