Stream restoration effectively alters functional diversity and composition of riparian plant communities in the southern Rocky Mountains, U.S.A.

Channel incision degrades ecosystems by lowering water tables and disconnecting floodplains. Stream restoration often aims to reverse these impacts. However, projects typically receive minimal monitoring, and treatment effectiveness has not been validated. We used trait‐based analysis to evaluate whether two stream restoration techniques—beaver dam analogs (BDAs) and plug‐and‐ponds—raised water tables and increased overbank flooding, whether these altered environmental filters facilitated recovery of riparian plant communities, and how reassembly impacted the representation of traits that influence ecosystem function. We report on a before‐after‐control‐impact study and Bayesian analysis that estimated the probability that treatments affected riparian plant functional diversity and composition. We found a high probability (0.99 and 0.97, respectively) that BDAs decreased functional dispersion by ≥50% and plug‐and‐ponds decreased dispersion by ≥30%. Both treatments increased the relative abundance of high moisture use plants, wetland plants, and plants with high anaerobic tolerance. For example, BDAs increased the relative abundance of obligate wetland plants by 100%, and plug‐and‐ponds increased the relative abundance of facultative wetland plants by 105%, on average. These results suggest treatments modified environmental filters and recovered riparian plant communities. Ecosystem function was likely altered as the streamside plant community reassembled. Small increases in functional divergence suggest both treatments increased resource use efficiency, and we found a high probability of small treatment effect sizes (<20%) related to changes in community‐level C:N and nitrogen fixation. Our results demonstrate trait‐based analysis can detect a rapid response to restoration and offer a cost‐effective monitoring approach to compare treatments across space and time.