Riverine biodiversity in space and time: Effects of network position and hydrologic variation

Background/Question/Methods

Flow regime is considered a ‘master variable’ in river ecosystems: it controls the availability of physical habitat and the dynamics of biological communities. Additionally, hydrologic connectivity between headwaters and main-stem reaches can influence biodiversity dynamics within river networks. To gain a better understanding of how flow variation and network position may individually and collectively influence metacommunity dynamics, we leveraged an intermittent stream network (Chalone Creek, Pinnacles National Park, California) that exhibits strong spatio-temporal variation in hydrology. We studied 16 stream sites over a period of strong interannual fluctuations in hydroclimate (2015-2021). At each site, we placed sensors to track hydrologic conditions (wet vs dry), used a distributed hydrologic model to simulate daily streamflow at 10-meter resolution, and ran wavelets on the model output to identify dominant wet/dry frequencies. For invertebrate community data, we first used multivariate (NMDS) analyses along with rarefaction and extrapolation methods to compare how different sampling methods (30m vs 150m Reach Wide Benthic Sampling) provided different levels of sampling coverage. On the combined community data, we used beta diversity-partitioning methods to evaluate changes in composition over space and time, and to identify ‘keystone’ sites, i.e. locations that contribute disproportionately to network-wide biodiversity.

Results/Conclusions

Hydrologic modeling results showed that three locations within the network are perennial, while the rest range in intermittency (on average flowing ~20-90% of the year). Wavelets showed that flow periodicity was driven by short-lived winter storms (1-2 week frequency) that rewetted small intermittent tributaries and by annual wet-dry cycles that weakened significantly over the multi-annual drought. Sampling methodologies were similar in the diversity they captured, with 150m samples having on average slightly higher alpha diversity counts. The studied macroinvertebrate community was highly variable over space and time, with beta diversity being more strongly driven by species replacement than by species richness gradients (overall: 67% replacement, 33% richness). Local (site-level) contributions to beta diversity ranged from 1.43-3.47%, with both perennial and headwater sites showing statistically significant contribution values. Despite being widespread, species in the families Simuliidae (Diptera), Chironomidae (Diptera), and Capniidae (Plecoptera) contributed most to network-wide beta diversity (4.03-6.79%). We contend that a better understanding of spatiotemporal variability in community composition and the identification of ‘keystone’ sites may lead to more robust prioritization of restoration and conservation efforts in dynamic riverine environments.