Historical modification of land cover has resulted in stream macroinvertebrate communities shifting to generalist taxa in forested, grassland, and agricultural watersheds. Conversion from undeveloped to agricultural land cover alters water quality, organic matter inputs, retention, and habitat heterogeneity in streams. Erosional and depositional habitat heterogeneity is essential for supporting unique macroinvertebrate niches, although sampling protocols designed to sample riffle-only habitats could miss taxa exclusive to depositional habitats. The United States Environmental Protection Agency (USEPA) multihabitat Rapid Bioassessment Protocol (RBP) collects invertebrates from multiple habitats but aggregates all samples from a single site, so communities are not associated with habitats. We will quantify macroinvertebrate assemblages from both erosional and depositional habitats to better understand the biotic communities in each habitat type and how they may influence measures of stream condition. We will assess five quantitative biomass-weighted macroinvertebrate samples in depositional and riffle habitats across 24 streams along a gradient of agricultural land cover. Our study will provide information on often-undersampled habitats in macroinvertebrate bioassessment. We hypothesize that depositional habitat types (i.e., debris accumulations, undercut banks) will support more taxonomically and functionally diverse communities than riffles. Due to depositional habitat diversity providing niches for macroinvertebrates, we also predict and will present data on whether streams with greater depositional habitat diversity have greater species richness. Additionally, we predict and will present data on whether depositional habitats will support more biomass-weighted clinger and shredder taxa than riffles. Mixed pasture-row crop land-cover within the watershed ranged from 10% to 76%, and Shannon habitat diversity ranged from 0 (one habitat type) to 1.3 (different proportions of all five habitat types). Mean richness across 7 preliminary sites indicated that riffles (mean= 44, standard deviation = 8.50) and depositional habitats (mean = 43, standard deviation = 12.49 ) contained similar amounts of taxa. Our study will determine the unique taxonomic and functional assemblages that riffle and depositional habitats provide to stream bioassessments, and how these assemblages change along a gradient of agricultural land cover and depositional habitat heterogeneity.