Sample selection disparity: Sampling only cobble overestimates the biomass of stream benthic algae

Abstract Despite the fact that scientists are aware that the streambed consists of various substrata in size, the estimation of benthic algal biomass has been conducted based almost exclusively on cobble sampling. This disparity in samples selected for the biomass estimation occurs because frame sampling collects all substrata, encompassed by the frame, including sand and stones, and is a time‐consuming method compared to single‐stone sampling. We conducted frame versus cobble sampling to test for sample selection disparity (SSD) in the estimation of benthic algal biomass. Estimates of algal biomass based on the frame sampling (area: 0.25 m 2 ) were compared with those based on the cobble sampling taken at the same sampling points in a diatom‐dominated stream. Benthic algal biomass estimated based on cobble sampling was larger than the biomass estimated with frame sampling. The contribution of cobbles to the algal biomass encompassed by the frame was considerably higher than smaller substrata. These results suggest that cobble sampling tends to result in an overestimate of the benthic algal biomass in natural streams. Because the frame sampling requires intensive labor and time, we here proposed a general model based on quick visual assessments for percentage cover of cobbles on the streambed to calibrate estimates obtained by cobble sampling.