Understanding fate and transport within fluvial systems requires accurate modeling of breakthrough curve (BTC) tails, which often display non-Fickian behaviors. However, it is unclear how anomalous processes relate to the physical and biological characteristics of the stream ecosystem. We use the Stochastic Mobile Immobile model (SMIM) to determine the impact of biofilm colonization among different substrate types on reach-scale transport velocity (V) and dispersion (D), rate of delivery to the subsurface (Λ), and retention within the subsurface (reflected by power law slope; β). During summer 2020 and 2021, we conducted a total of n=42 Rhodamine-WT releases in four experimental streams lined with contrasting substrata (sand, pea gravel, cobble, and a three-way mix) at the Notre Dame Linked Experimental Ecosystem Facility (ND-LEEF) in Indiana (USA). To explore the effect of biofilm colonization, we conducted releases under artificially shaded, early and late biofilm development, and senescent biofilm conditions. We found that replicated releases under constant conditions consistently reproduced stream BTCs and modeled transport parameters. Biofilm abundance, biofilm status (living versus dead), and substrate type produced significant variations in BTC shape and transport parameterizations. We found a non-linear relationship between algal biomass and V, where increases in biomass produced decreases in V at low biomass and increases in V at high biomass. Substrate type also predicted patterns in transport, with sand producing higher V, Λ, and β than larger substrata. These results suggest that substrate type acts as the primary driver and biofilm development the secondary control on transport in fluvial systems.