We link v_f to a Continuous Time Random Walk model to quantify how v_f is controlled by in-stream (velocity, dispersion, benthic uptake rate) and hyporheic processes (exchange rate, residence times, hyporheic uptake rate). We fit the model to conservative (NaCl) and nitrate (NO₃^-) pulse tracer injections in unshaded replicate streams at the Notre Dame Linked Experimental Ecosystem Facility, which differed only in substrate size and distribution. Experiments were conducted over the first 25 d of biofilm colonization to examine how the interaction between substrate type and biofilm growth influenced modeled processes and vf_. Model fits of benthic uptake rates were ~8X greater than hyporheic uptake rates for all experiments and did not vary with substrate type or over time. High benthic uptake was attributable to filamentous green algae coverage of the streambed, which dominated total algal biomass. v_f was most sensitive to benthic uptake rate and in-stream velocity. v_f sensitivity varied with stream conditions due to its nonlinear dependence on all modeled processes. Together, these results demonstrate how integrated, reach-scale nutrient demand can be attributed to specific contributing processes in the stream.