San Francisco Estuary and Watershed Science

Cyanobacterial harmful algal blooms (CHABs), often comprising the cyanobacteria genus Microcystis, are an increasing water-quality concern in the large, hydrologically complex, Sacramento–San Joaquin Delta (Delta). Monitoring CHABs in estuarine environments is challenging because of the dynamic physical, chemical, and biological gradients across the freshwater–saltwater continuum. Here we provide an overview and synthesis of environmental drivers known to influence CHAB formation. We analyzed long-term environmental data— including flow, turbidity, temperature, salinity, and nutrients compiled from routine monitoring stations—to assess variability across seven defined regions of the Delta. Our results highlight significant differences in environmental conditions among regions. The upper San Joaquin River (USJR) region has previously been recognized as one of the areas in the Delta most prone to CHAB events, and in our analyses exhibited the highest summer water temperatures, elevated nutrient concentrations, and low turbidity all conditions that favor CHAB development. In contrast, the lower San Joaquin River (LSJR) and lower Sacramento River (LSR) regions had cooler temperatures and higher turbidity, which are generally less favorable for bloom formation. While water temperatures in the East Delta (ED) were relatively cool, they increased more rapidly than in other regions, with an observed warming trend in June of approximately 0.2 °C per year. The ED was also the region that had the least amount of environmental monitoring relative to the other study regions and had no consistent nutrient data. Enhanced monitoring in the ED and continued monitoring in the other regions will be critical for tracking climate-driven changes in environmental conditions that support CHAB proliferation. 

Cyanobacteria harmful algal bloom (CHAB) events, often dominated by the cyanobacteria genus Microcystis, are a major water-quality issue in the large, hydrologically complex, Sacramento–San Joaquin Delta (Delta). To improve future CHAB-specific monitoring, we analyzed historical data to characterize spatial and habitat-based patterns in CHAB occurrence. We identified three primary habitat types that support CHABs to various degrees: (1) dead-end channels and marinas, which tend to support the highest densities and most frequent blooms; (2) flooded islands, shallow side channels, and sloughs, where CHABs occur less frequently, and generally at lower densities, and (3) main channels, where CHABs occur but typically at lower intensities and without the surface accumulations observed in dead-end channels. Data from routine monitoring programs and special studies were used to evaluate CHAB variation across Delta regions and among these habitat types. Monitoring metrics included chlorophyll-a (chl-a), Microcystis abundance determined by microscopy and quantitative polymerase chain reaction (qPCR), microcystin toxin concentrations, and Microcystis Visual Index (MVI) scores. July through October was identified as peak CHAB season, with CHAB severity greatest in the South Delta, Central Delta, and San Joaquin River regions. CHABs and associated toxins were infrequent in the Sacramento River, Cache Slough Complex, and East Delta regions, with microcystin concentrations typically below detection (0.15 µg L–1) and always below the California health-based warning trigger of 6.0 µg L–1. Dead-end channel habitats exhibited significantly higher microcystin concentrations than other habitat types, with nearly 40% of samples exceeding the California caution trigger of 0.8 µg L–1, compared to 10% in flooded islands and 7% in main channels. Based on these findings, we recommend enhanced sampling in dead-end channels and continued CHAB monitoring to identify and refine risk forecasts under changing climate conditions.

In 2023 we placed Delta Smelt (Hypomesus transpacificus) into field enclosures for 6 weeks to assess whether an enclosure study could effectively detect Delta Smelt responses to habitat differences during the summer–fall period. Prior enclosure studies have noted high biofouling accumulation during warm months, so we simultaneously tested two biofouling mitigation techniques (cleaning vs. replacement) to better understand their efficacy and potential effects on the fish inside. Fish were held at two sites in the San Francisco Estuary: the Sacramento River near Rio Vista, and Montezuma Slough near Belden’s Landing. Both biofouling mitigation measures were employed at each site. We evaluated water quality and zooplankton inside and outside of the enclosures as well as algal and macroinvertebrate accumulation on the enclosures, and assessed fish survival, health, and diets. We observed significantly improved fish health metrics at Rio Vista compared to Belden’s Landing, potentially associated with higher-quality food, lower water temperatures, and lower water velocities; however, it is unclear whether these differences were in response to the geographic location or to localized enclosure effects, because the fish were unable to seek refuge or forage. We also observed differences in enclosure biofouling and associated macroinvertebrates between the two mitigation measures and saw higher mortality in replaced enclosures but no differences in fish health among survivors. Macroinvertebrates were detected on nearly all enclosures and often made a sizeable proportion of smelt diet biomass, regardless of location or mitigation method, indicating that the enclosures may provide an altered food environment. This study demonstrates the complexity of implementing and interpreting Delta Smelt enclosure studies, and highlights limitations of this tool. Our data offers physiological clues that enhance our understanding of the response of cultured Delta Smelt to differing habitats, but uncertainties highlight that enclosures studies should not be used in isolation to address large-scale management questions.

San Joaquin River-origin Central Valley Chinook Salmon are a species of conservation concern because they play an essential role in the health and function of California’s aquatic ecosystems yet experience low survival as juveniles pass through the Sacramento–San Joaquin Delta (Delta) on their way to the Pacific Ocean. We analyzed 8 years of juvenile Chinook Salmon acoustic telemetry data near the State Water Project and Central Valley Project in the South Delta to inform how hydrodynamic and environmental conditions affect fish movement. The behavior of Chinook Salmon in this area was related to simulated outputs from a high-resolution hydrodynamic model that used a continuous-time multi-state Markov model. A total of 839 fish were detected near the facilities, and the median time fish spent in the area was 33 hours. Flow conditions near the water projects strongly influenced juvenile Chinook Salmon movement patterns. For fish detected at the Central Valley Project trash racks and outside the radial gates to Clifton Court Forebay at the State Water Project, the strongest effects on movement were from upstream-directed net flow, time of day, and turbidity. Export operations drew fish movement into the water export facilities when exports were high and at night. Movements in West Canal between Highway 4 and the radial gates were influenced by tidal flow and net flow. Fish primarily moved downstream on the ebb tide, especially when exports were lower. Study fish near the water projects experienced high predation, which reduced sample size and increased uncertainty. Our study identified factors that drive juvenile Chinook Salmon movements near the water projects and can be used to inform water operations and management decision-making that may improve salmon survival in the Delta.

Evaluating patterns in the distribution of genetic variation across heterogeneous landscapes provides perspectives on how environmental features contribute to molecular evolution, and how human activities influence populations. Impassable barriers, such as hydroelectric dams, and human-directed movement or propagation of individuals across riverscapes have had strong impacts on the genetic diversity observed in natural populations. We considered the population structure and distribution of adaptive genetic variation in the salmonid species Oncorhynchus mykiss throughout Putah Creek, a watershed near Napa Valley, California that was transformed by the 1950's construction of multiple dams and a large artificial reservoir (Lake Berryessa). In addition, hatchery-raised O. mykiss has been introduced throughout the watershed over many years, which may have introduced genetic variation into existing Putah Creek populations. To explore the population structure and distribution of adaptive genetic variation in Putah Creek, we analyzed microhaplotypes from eight O. mykiss populations above Lake Berryessa and two populations below Lake Berryessa and compared them with 40 reference populations from California Central Valley, coastal, inland, and hatchery rainbow trout lineages. We found distinct patterns of neutral and adaptive variation between populations above Lake Berryessa and those below. Populations below Lake Berryessa resembled various Central Valley populations and hatchery rainbow trout strains, while those above were more similar to coastal O. mykiss populations. Additionally, Putah Creek populations below Lake Berryessa possessed significantly different proportions of adaptive variants associated with life-history traits compared to populations above Lake Berryessa, consistent with studies in other populations located above and below barriers to migration.

Long-term monitoring programs may face protocol or equipment changes for various reasons, and they must determine how to implement change while maintaining the historical consistency and significance of their data. A case study on the US Fish and Wildlife Service’s (USFWS’s) Enhanced Delta Smelt Monitoring program (EDSM) demonstrates what factors a long-term monitoring program might consider when substituting new equipment to collect environmental data, such as turbidity. Turbidity can influence the health, survival, and habitat quality of fish and other organisms. Thus, reliable collection of turbidity data is critical. Turbidity data are commonly presented in two units: nephelometric turbidity units (NTU) and formazin nephelometric units (FNU). The EDSM historically collected turbidity data using NTU units, like other monitoring programs in the San Francisco Estuary. Changes in reported turbidity values by other agencies, and proposed criteria relevant to the protection of the federally threatened Delta Smelt, have shifted focus to FNU values. Before transitioning to only collecting FNU values, the EDSM performed side-by-side collections of both FNU and NTU values for 2 years. Multiple linear regression supported a linear relationship between FNU and NTU values when the covariates of temperature, dissolved oxygen, specific conductance, stratum, and month were included. The discrepancies in FNU and NTU values could be explained by varying light wavelengths, differences in sample collection method, or sensor fouling. The model also demonstrated predictive capabilities for future unit conversion. These findings support the EDSM collecting turbidity in a single unit while maintaining the ability to correlate with historical environmental data, if needed. Other long-term monitoring programs could make similar considerations when faced with programmatic or equipment changes to ensure minimal effect on their historical datasets.