UC Davis

Eastern Boundary Upwelling Systems (EBUS) are highly productive biomes, which provide benefit to society and support local ecosystems. Although EBUS total area is small when compared to other pelagic ecosystems, a growing body of literature demonstrate that climate impacts on EBUS will have disproportionately large consequences for human society. Like other EBUS, the California Current System (CCS) is experiencing a number of inextricably linked stressors: acidification, oxygen stress (hypoxia), altered food webs, and warming temperatures. Each stressor has the potential to change species interactions; alter the abundance and distribution of organisms; and can even result in mortality for certain organisms. Wind forcing and freshwater input drive change in the coastal zone, and result in heterogeneous expression of multiple stressors in time and space. River-flow and winds are both anticipated to change in magnitude and timing due to human- and climate-induced changes, which drive associated impacts to physical and biogeochemical processes in estuaries and continental shelves. A step towards better understanding drivers of multiple stressor interactions includes analysis of subsurface observations to identify relationships and trends in shelf waters. In this work we focus on the physical dynamics which influence dissolved oxygen (DO) over the shelf off northern California and Washington, to better understand the physical dynamics that influence hypoxia. In the CCS, and other EBUS, high productivity is supported by coastal wind-driven upwelling that supplies the shelf with nutrient-rich waters. However, high rates of productivity in the coastal zone may operate at the expense of (1) decreasing aragonite and calcite saturation states and decreasing dissolved oxygen (DO) concentrations because the water that is upwelled to the continental shelf also has reduced DO levels, lower pH, and higher concentrations of dissolved inorganic carbon (DIC); and (2) high productivity maintains a high standing stock of particulate organic carbon (POC), which builds a respiration signal in the water column and at the sediment/water interface and results in a decline in DO. These are two mechanisms that make EBUS, including the CCS, prone to hypoxia and acidification, which threaten ecosystems and the communities they support. The coastal waters of Washington (and southern British Columbia) have the highest primary productivity in the CCS, but this high productivity is not co-located with the strongest upwelling-favorable alongshore winds (which occur off northern California). This mismatch has been explored (e.g, by Hickey and Banas 2008), and results point to additional mechanisms that facilitate the region's high productivity beyond the traditional focus of the coastal wind field. The work presented in this dissertation, to explore the physical dynamics which influence DO over the shelf in two regions of the CCS, was motivated by (1) the link between productivity and hypoxia (and the mismatch of productivity/wind forcing); (2) reports of extreme low DO observed off Washington in the summers of 2017 - 2019; and (3) a lack of subsurface DO time-series observations off northern California (where peak upwelling wind stress occurs). Two chapters of this dissertation focus on the central CCS (off northern California), a region for which which DO time-series are scarce; and one chapter addresses the northern CCS (off Washington) and is comprised of an analysis of a 10-year record of DO, temperature and salinity at multiple sites along with wind and river discharge data to better understand the timing and severity of shelf hypoxia.

Although the CCS is one of the most highly observed ocean regions in the world, there has been comparatively limited research on subsurface DO and carbonate system parameters in the central CCS off northern California (from 37°N to 42°N). Since long time-series of subsurface DO are relatively scarce, management decisions are made without a proper understanding of regional risk. Scientifically we are left wondering: (1) what DO levels occur in a section of the CCS that experiences the strongest upwelling favorable winds (~8x stronger than the Pacific Northwest)?; and (2) how DO levels respond to upwelling and relaxation events in this subregion? We collected time-series mooring data (temperature, salinity and DO), which are used to describe patterns and timing of hypoxia, and explore how DO levels respond to upwelling and relaxation events in the northern California coastal upwelling region. A deeper, mid-shelf site (~54m) is located in the Gulf of the Farallones, offshore San Francisco Bay, and within the more stratified upwelling shadow south of Point Reyes. The lowest DO concentrations and most persistent hypoxia were present at the mid-shelf site. At the shallower sites, (~18 and 30m), results show highly variable DO values with brief hypoxic events outside the core upwelling season (i.e., strongest winds and coldest water did not associate with the lowest DO levels). At the deeper, mid-shelf site, two distinct modes of variability were observed. During the first mode, upwelling events related to DO decline and relaxation events to increasing DO. During the second mode, the opposite occurs: upwelling related to an increase in DO and relaxation events to declining DO. At the shallower inner-shelf sites, and for the entire time-series, upwelling generally relates to DO decline and relaxation events to increasing DO. The importance of source water is clear during the first half of the mid-shelf deployment, and the overall trend and second half shows the importance of local drawdown. We also explored the seasonality of DO over a submarine bank (Cordell Bank) located at the shelf-break off northern California. Results show a recurrent seasonal cycle in temperature and DO. The similarity of seasonal patterns of temperature across years (2014 - 2018) is interesting, especially given the diverse set of oceanographic conditions the CCS experienced from 2014 – 2019. Although the coolest water occurs over the bank early in the upwelling season, the DO minimum occurs later, towards the end of the upwelling season and often during the relaxation season. Deviations from the seasonal trend observed are likely attributed to a combination of physical and biogeochemical processes working together. Specifically, we hypothesize that the interplay of wind-driven mixing and surface productivity can explain internanual differences, but additional work is needed to fully understand the role of these drivers. At Cordell Bank, DO concentrations were often below the threshold of mild hypoxia (2.45 ml/L), but only one instance of intermediate hypoxia was observed (1.4 to 2.45 ml/L) during the relaxation season (July 2017). Overall DO off northern California appears higher (fewer hypoxic events) than those observed in the northern CCS, but the upwelling favorable winds are also eight times stronger off northern California. Finally, observed DO is also lower than predicted using two source waters (PEW and PSUW) thus pointing to the likely importance of local drawdown (and potentially presence of more than two source water masses).

We focused on the physical dynamics which influence shelf DO in the northern CCS (off Washington), in attempt to better understand the physical dynamics that influenced the extreme low DO observed in the summers of 2017 - 2019. Mooring data (2011 - 2020) are used to describe the timing and severity of hypoxia off Washington. River and winds data (1991 - 2020) are also used to better understand the coastal environment and drivers of low DO. From this work we found significant interannual variability in DO. The 2016-2019 low DO period is statistically associated with spicier water, suggesting a link between source waters impacted by the El Niño, very, very low North Pacific Gyre Oscillation Index, marine heat wave presence and low DO over the shelf. When compared to historical DO records, summertime hypoxic exposure appears to have worsened on the Washington shelf. However, 2011 – 2020 shows significant interannual variability without a clear downward trend. We also observed a north-south trend with lower DO in the south, which can be explained by several hypotheses, ranging from shelf width to canyons to stratification. However, the relationship between stratification, surface salinity and DO is complex. Within periods with similar surface salinity, more stratification is related to lower DO, but overall higher stratification is caused by lower surface salinity and is associated with higher DO. This likely demonstrates a complex relationship between the presence of river water advected northward by downwelling-favorable winds and vertical mixing driven by downwelling. Additional work to further assess the impact of timing of the Columbia and Fraser Rivers relative to wind events is important to understand DO and carbonate chemistry off Washington, and help make predictions for future climate conditions.