Linking community stability to spatial processes in variable environments is one of freshwater sciences continual challenges. Differences among ecological communities may be driven by differences in abiotic and biotic environmental conditions as well as dispersal connectivity. The Network Position Hypothesis (NPH) posits that these different mechanisms vary strongly in their influence on community assembly depending on the position of the community within a river network. While the NPH has been evaluated with respect to spatial β-diversity, the influence of network position on temporal β-diversity is less well understood. Using analyses of long-term monitoring data from the California Surface Water Ambient Monitoring Program, we explore how a community’s position in a branching river network can influence changes in community composition over time. We predicted that stronger abiotic filtering and reduced dispersal connectivity in headwaters would lead to reduced temporal β-diversity that is dominated by nestedness, whereas more favorable abiotic conditions and higher dispersal connectivity in mainstems lead to increased temporal β-diversity and taxonomic turnover. Consistent with predictions, we found that temporal β-diversity increased with river order, with lowest in headwaters and highest in 4- and 5- order mainstems. Headwater β-diversity was mostly composed of nestedness changes with increasing contributions of turnover in higher order mainstems. Analyses of temporal changes in trait diversity and composition were strongly consistent with analyses based on taxonomic diversity. These results provide further evidence that assembly mechanisms may differ among river network positions. Furthermore, they have direct implications for sustainability efforts, especially in the realm of biomonitoring and stream restoration.