Assessments of landscape influences on streams frequently identify agricultural land use in stream catchments as a major stressor as well as an important predictor of biological community composition. Despite agriculture’s frequent association with stream condition, the mechanisms by which agriculture affects streams vary with stream types as well as with different kinds of crops and farming practices on the landscape. Stream temperature, for example, is affected by subsurface geology which dictates groundwater delivery to streams; stream shading, which is affected by forested buffers and stream size; and runoff from fields, which depends on cropping practices, soil types, slope, and artificial drainage. These factors interact to dictate stream water temperature which is a major driver of fish species presence. Michigan is one of the most agriculturally diverse states in the US, while also supporting diverse and valuable freshwater fisheries, making it an ideal region to evaluate relationships between complex agricultural systems and fish communities. This presentation proposes preliminary analyses and a theoretical framework to disentangle the myriad impacts from agriculture on fish communities, in watersheds dominated by agriculture in the southern Lower Peninsula of Michigan. We leverage available landscape datasets (such as the new National tile drainage dataset) to understand how fish species and functional groups respond to different agricultural influences on fish habitat.