Purpose and Scope

The purpose of this report is to provide information about how the major drivers of nutrient distributions interact across Delta aquatic landscapes during periods of low river flows (spring, summer, and fall). The first set of major drivers is the magnitude and location of nutrient loadings, which interact with flow to determine nutrient concentrations; this set of major drivers was compiled to improve understanding of how planned upgrades to the SRWTP may affect desired environmental outcomes. The second set of major drivers is the timescales of nutrient transport in and across the Delta. Timescales of nutrient transport vary with flow conditions but also interact with channel geometry, tidal mixing, exports from the south Delta, and other water withdrawals in ways that influence residence times and transport timescales disproportionately across the Delta. Residence times and transport timescales interact to affect the time over which natural internal cycling can act to alter nutrient concentrations and forms. The third set of drivers is the interactions of nutrients with features of Delta aquatic landscapes such as herbaceous tidal marsh, submerged aquatic vegetation, and large expanses of intertidal or subtidal sediments, all of which can be expected to exert some level of demand on available nutrients. Fourth, we explored the extent to which pelagic phytoplankton production acted as a driver of nutrient concentrations and was driven by nutrient concentrations and forms, with special emphasis on how nutrients are related to the formation of harmful algal blooms—those that produce toxins that adversely affect aquatic organisms, as well as beneficial algal blooms, and those that result in production of algae particularly beneficial to imperiled Delta pelagic aquatic food webs.

The study comprises three broad-area, high resolution water-quality-mapping surveys of the Delta that used continuous underway high frequency sampling and measurements onboard a high-speed boat to characterize spatial variation across the full extent of the Delta on three successive days, with surveys conducted in May, July, and October of 2018. Information about the concentration and spatial distribution of all major nutrient types and forms is collected simultaneously with analogous information about all major classes of phytoplankton and associated water-quality conditions. The results show substantial variation across space and time, providing an unprecedented snapshot of the dynamic environmental processes that shape the ways nutrients interact with and affect aquatic habitats in the Delta.

How To Use this Report

This report describes three comprehensive, high resolution spatial surveys that integrate nutrient concentrations, phytoplankton abundances, and phytoplankton taxonomic distributions with associated water-quality conditions across most of the Delta. This report highlights some of the main findings we believe are of interest to stakeholders, managers, and scientists responsible for managing and studying the Delta, but the limited scope of the report precludes discussion of many topics for which collected data may be highly informative. Data associated with this report will be available in various machine-readable forms in a published data release (Bergamaschi and others, 2020; https://doi.org/10.5066/P9FQEUAL). In addition, a publicly available, interactive, online data analysis and visualization exploration portal was created to facilitate exploration of data by interested parties (https://ca.water.usgs.gov/bay-delta/2018-delta-wide-mapping-surveys.html), and maps and other visualizations developed by this project have been published as a USGS Data Report (Soto Perez and others, 2023). Using this portal, users may query the surveys in a variety of ways to test hypotheses, examine relations between different parameters, and assess spatial trends and seasonal differences. The data exploration portal is intended to be an immersive experience that allows users to gain greater understanding of the complex interactions that shape Delta aquatic environments, and this report is intended as a companion to the portal, allowing the reader to challenge and further explore the highlighted findings. Thus, in addition to highlighting the major findings of the study, another major aim of this report is to orient readers to ways that these data can be explored interactively to improve understanding of temporal and spatial patterns, links between nutrient inputs to the system, nutrient transformation (including internal sinks and sources) within the system, and phytoplankton abundance and community composition. Hyperlinks and references to tabs within the tool (for example, tab 1, Nutrient Parameter Maps; Bergamaschi and others, 2020) are included with each figure citation to the location within the data portal wherein the figure may be reproduced.

Definition of Regions and Reaches for Analysis

Aside from evaluation of the Delta as a whole, the survey data were divided into smaller subsets for analysis in two primary ways. First, to evaluate intra-annual temporal and spatial trends and to identify controls on nutrients, phytoplankton, and other water-quality parameters at the regional scale, we defined eight study regions based on hydrodynamically distinct zones identified by Jabusch and others (2016). Broadly speaking the Delta has three tidal transition zones (north Delta, Mokelumne system, and south Delta), two strongly, tidally forced systems (western and central Delta), two dead-end channel systems (CSC and Suisun Bay marsh), and the brackish part of the estuary that connects freshwater portion of the Delta to San Francisco Bay. The eight regions defined in this study and used in the online visualization tool include (1) north Delta tidal transition zone, (2) Mokelumne system tidal transition zone, (3) south Delta tidal transition zone, (4) San Joaquin tidal transition zone, (5) central Delta tidal transition zone, (6) western Delta tidal transition zone, (7) Cache Slough complex channel system, and (8) Suisun Bay (fig. 3). A ‘transition zone’ is where transport transitions from being mostly unidirectional, like in rivers, to mostly tidal. These eight regions were used by Jabusch and others (2016) to evaluate data collection efforts and nutrient status and trends analyses in the Delta; these regions were originally developed to represent restoration opportunity areas based on an understanding of ecological function, physical drivers, existing constraints, and elevation gradients, while also accounting for hydrologic features, watershed boundaries, and Delta Simulation Model II (DSM2) modeling requirements.

For the purposes of this report and the associated data visualization, we also defined nine specific study reaches within these hydrodynamic regions. These study reaches were chosen to represent dominant flow paths that can be used to examine changes in nutrient concentrations and forms along with changes in phytoplankton abundances and species composition; these study reaches represent a range of transport timescales from a few days to several weeks. These reaches are defined as the (1) Sacramento River, (2) lower Sacramento River, (3) Liberty Cut, (4) Shag Slough, (5) upper San Joaquin River, (6) Old River, (7) Middle River, (8) Hog Slough, and (9) Sycamore Slough (fig. 3). Data for these nine study reaches are plotted against river mile (mi), with the northern- or western-most point designated mi zero for each reach (tab 4, Nutrient Maps and Study Reaches).

Sampling and Analysis

The mapping surveys described herein comprise measurements made continuously onboard the survey mapping vessel, those made on samples collected while underway, and samples collected while stopped at specific sampling locations. This high resolution boat-based mapping approach has been described previously by Downing and others (2016) and Stumpner and others (2020). The methods of sampling and analysis of the resulting data are described below.

Data Processing

Data from onboard continuous instruments not directly logged to the flow-through data collection system were merged based on timestamp to the nearest second. All data directly logged to the flow-through data collection system were processed using pandas software library (0.220; McKinney, 2010) in Python (3.7.3, Van Rossum and Drake, 2009) to remove periods of compromised data (for example, flow blockages and bubbles), to apply instrument corrections and unit conversions, and to apply a centered 20-second median filter to the time series.

Onboard NO₃ data measured with the SUNA V2 were processed to remove the salinity-dependent interference of bromide and dissolved organic matter (DOM) using the Sakamoto and others (2009) method (as implemented in the Seabird UCI software). Final NO₃ values were obtained by regressing instrument response against NO₃ concentrations obtained from laboratory measurements of discrete samples collected through the course of the day. Field and laboratory NO₃ measurements were typically well correlated with a linear fit. For the regressions, the SUNA NO₃ concentration data were the explanatory variable used to compute the response variable of the laboratory NO₃ concentration. Individual sample results with more than three standard deviations from the regression were considered outliers and removed from the regression. Larger differences between SUNA and laboratory NO₃ concentration data were assumed to arise from (1) actual spatial heterogeneity in environmental NO₃ concentrations between water entering the flow-through system and water captured in the sample bottle, (2) uncertainty inherent in sample handling and laboratory analysis, or (3) potential interferences of turbidity and DOM that impact the algorithm for determining nitrate concentration from the ultraviolet (UV) absorption spectrum. Summary statistics and regressions for individual days are identified in table 3.

Ammonium data were processed by first correcting the voltage (V) output of the instrument for baseline drift based on the instrument response during periodic measurements of organic-free deionized water and then using a regression model of the V response to a standard concentration. A coefficient of determination (R²) of 0.97 or greater was considered acceptable. The NH₄-data timestamp was corrected to account for the time lag associated with the microfluidic processing before integration with other data sources.

Dissolved inorganic nitrogen (DIN) concentrations were calculated as the sum of NO₃, NO₂, and NH₄ for continuous data and discrete sample data. Laboratory dissolved organic nitrogen (DON) concentration was calculated by subtracting NH₄, NO₃, and NO₂ data from the TDN data (in other words, DON=TDN−DIN). Continuous DOC and DON concentrations were developed by regressing onboard fDOM data against the DOC and DON concentrations measured in discrete samples. Data exceeding three times the standard deviation of the residuals were considered outliers and were not included in the model. Continuous total dissolved nitrogen was calculated as the sum of DIN (NO₃ + NH₄) and DON.

Data were spatially aligned to a common spatial framework to facilitate comparisons among dates and to minimize the effects of differential data density on the spatial interpolation calculations by assigning median-filtered data by proximity to a centerline vector comprising points located at the centerline of all channels navigated, spaced at about 150-m intervals. Interpolated water-quality maps were created from the spatially aligned data using ESRI’s ArcGIS Pro 2.2.1 Spline with Barriers tool (Terzopoulos and Witkin, 1988). The raster output was smoothed with ESRI’s ArcGIS Pro 2.2.1 Focal Statistics tool, and interpolated values were extracted to the spatially aligned points to create continuous spatially aligned data across the mapping domain.

Hydrodynamic Context

The Delta is characterized by a complex mixture of deep tidal channels that were formed when the area was drained and leveed for habitation and agriculture in the 19th century (Thompson, 1957). The deep channelization causes complex hydrodynamic interactions as tidal energy propagates through the system, affecting virtually all aquatic processes in the Delta. The complex hydrodynamics alters residence times, causes high levels of mixing, and transports material landward and seaward via tidal pumping (Gross and others, 2019). Adding to this hydrodynamic complexity is the large-scale export of water from the south Delta via the California State Water Project and the Federal Water Project, which causes a net north-to-south flow during periods of pumping (Kraus and others, 2017a). In some areas, like the CSC and dead-end sloughs, local water withdrawals are large enough to create net landward flows, transporting constituents in directions that are typically considered “upstream.”

The major driver of hydrologic conditions in the Delta is the status of river inflows, with the Sacramento River being the largest source of water (Jassby, 2008). The May 2018 survey was conducted approximately 1 month following the peak winter flows on the Sacramento River and 2 weeks after the recession of those flows to baseline values (fig. 4). Flows during the 3 days of the May 2018 survey corresponded to a seasonal minimum in Sacramento River discharge of about 8,000 cubic feet per second (ft3/s) at Freeport, Calif. (USGS site number 11447650; U.S. Geological Survey, 2021; fig. 4A), with only November 2018 flows being lower that year. Discharge at Freeport was twice as high during the July 2018 survey (about 16,500 ft3/s) and 30 percent higher (about 10,500 ft3/s) during the October 2018 survey (fig. 4A).

The other major drivers of transport and dispersion across the Delta are the export pumping facilities in the south Delta, which draw Sacramento River water across the Delta from the north to the south, diverting water from flowing west to San Francisco Bay and out to the Pacific Ocean (not shown). Exports can be quite high in comparison to river inflows, at times exceeding 14,000 ft3/s. Exports were at a local minimum during the May survey (about 2,500 ft3/s), with higher exports (about 4,000 ft3/s) immediately prior. Exports during the July and October surveys were about 7,000 ft3/s, with higher exports (greater than 10,000 ft3/s) dominating the intervening period (Dayflow–California Natural Resources Agency, 2021).

To understand the effect of exports on transport and dispersion in the Delta, we examined the magnitude of exports relative to the magnitude of Delta outflows exiting to the west toward San Francisco Bay (Exports: dayflow). The value approaches zero when exports are small relative to outflow and theoretically increases to a value greater than one when exports exceed outflow to the ocean. In water year 2018, the ratio of the two values varied from a minimum of 0.04 during spring high flows to a maximum of 0.62 during summer low flows (Dayflow–California Natural Resources Agency, 2021); a water year represents the 12-month period from October 1^st of one year to September 30^th of the following year and is designated by the year in which it ends. The ratio varied markedly between the three study periods, with values of approximately 0.15 during the May survey, 0.40 during the July survey, and 0.60 during the October survey, indicating that the cross-Delta flows increased progressively throughout the year (fig. 4).

Hydrologic transport in the Delta is also affected by the Delta Cross Channel, a managed flow-control structure that redirects Sacramento River main stem water into the central Delta tidal zone via the North Fork Mokelumne and South Fork Mokelumne Rivers (figs. 2, 3). The gates are opened typically in late spring and closed in fall. When the gates are closed, Sacramento River water is increasingly directed into the western distributaries and down the main stem into the Rio Vista reach, though flows to the central Delta also occur via the ungated Georgiana Slough. When the Delta Cross Channel gates are open, the Mokelumne River system tidal transition zone is more directly influenced by the north Delta tidal transition zone (fig. 3). In 2018, the Delta Cross Channel was closed during the May survey but open during the July and October surveys.

Finally, understanding the effects of agricultural withdrawals helps to understand the distribution and sinks of nutrients in the Delta. When dead-end, tidal slough systems have a total volume much greater than the volume of water exchanged tidally, water withdrawals and evaporation can cause a slow landward transport of water that is unable to mix with water in the surrounding channels. There are many examples of this across the Delta. The Toe Drain at Yolo By-Pass draws Sacramento River water north into Prospect Slough thus affecting the nutrient concentrations found in Liberty Island, Little Holland Tract, and the surrounding channels and wetlands (fig. 2). Other notable examples are the Sacramento River Deep Water Ship Channel, upper Cache Slough in the western portion of the CSC, and the three dead-end sloughs—Beaver, Hog, and Sycamore—which extend to the east from the South Fork Mokelumne River. These areas act as entrapment zones, trapping nutrients and phytoplankton and extending residence times (fig 3; Downing and others, 2016; Gross and others, 2019; Stumpner and others, 2020).

Nutrients in the Delta

The Delta is formed by the convergence of the Sacramento River to the north, the San Joaquin River to the south, and other smaller rivers to the east (fig. 2). Land use in the Delta and immediately upstream is dominated by agriculture and urban use (Saleh and Domagalski, 2015). The sources and magnitudes of nutrient loading to the Delta can vary in response to storms, seasonal changes in discharge, agricultural practices, and other processes, with most of the Delta nutrient loading coming from the Sacramento and San Joaquin Rivers (Kratzer and others, 2011). On an annual basis, upstream landscapes provide most of the nitrogen loading in the Delta, whereas approximately 25 percent of the total nitrogen loading and 20 percent of the total phosphorus loading in the Delta are from municipal wastewater (Domagalski and Saleh, 2015; Saleh and Domagalski, 2015), and these point sources comprise most of the nutrient loading during non-winter months (Novick and others, 2015). The largest discharger in the Delta, SRWTP, releases most of its nitrogen as NH₄ into the Sacramento River just downstream of Freeport, accounting for more than 90 percent of the total annual NH₄ loading to the north Delta (Jassby, 2008). Other major point sources of nutrients to the Delta include the Stockton Regional Wastewater Control Facility (RWCF), which discharges near the Stockton Deep Water Channel; the Central Contra Costa Sanitary District Treatment Plant (DTP); and the Fairfield-Suisun Sewer DTP, which discharge into Suisun Bay and Suisun marsh, respectively.

Using a box model approach, Novick and others (2015) estimated that the Sacramento River—including inputs from the SRWTP—contributes most of the NH₄ (95 percent), DIN (65 percent), and TDN (65 percent) entering the Delta during the summer months (June–October). That modeling effort assumed that internal nutrient loading within the Delta from point sources, bottom sediments, atmospheric deposition, and wetlands) is minimal, arising largely from Delta-island drainage (Richardson and others, 2022). Novick and others (2015) estimated that between 15 and 30 percent of the total inputs to the Delta were lost during transit, with greater reduction occurring in average or low-flow years compared to high-flow years. Nevertheless, outflow from the Delta is the largest source of nutrients to the San Francisco Bay (Novick and others, 2015).

Spatial and Seasonal Variation in Nutrient Concentrations

The survey measurements included continuous high resolution measurements of nutrient concentrations, including NO₃ (representing the sum of nitrate and nitrite), NH₄, fDOM (a proxy for DOC), and DON. As described in the methods, from these data we calculated dissolved inorganic nitrogen (DIN = NO₃ + NH₄), DON, and total dissolved nitrogen (TDN = DIN+DON). Water samples for laboratory analysis of PO₄ were collected every 2 mi, and the data were interpolated for analysis. Below we discuss the highlights for the measured nutrients by survey date.

Nutrient Distributions during the May 2018 Survey

The May 2018 survey revealed a nutrient-rich system, with median DIN concentrations of 24 micromolar (μM) and median TDN concentrations of 39 μM (fig. 5; tab 8, Side by Side Parameter Maps; Bergamaschi and others, 2020). The range of concentrations (0.0–64.2 μM DIN; 6.1–70.7 μM TDN), however, was lower than in other months (fig. 6; tab 6, Nutrient Data Explorer; Bergamaschi and others, 2020), indicating that the distribution in May was the legacy of elevated nitrogen concentrations present in winter runoff (Downing and others, 2017). Detached and high-productivity areas of the Delta exhibited much lower concentrations of DIN (less than 10 μM), particularly in northern Cache Slough and upper Disappointment Slough. The Delta Cross Channel was closed during this survey, and consequently, DIN concentrations were also low on the North Fork Mokelumne and South Fork Mokelumne Rivers. DIN was substantially depleted in upper Grizzly Bay, which is well connected to the higher concentration areas immediately to the south and west, indicating a high local demand for nitrogen.

The median concentration of NH₄ during the May survey was 2.2 μM and accounted for 11 percent of DIN—the lowest of all surveys (fig. 7; Bergamaschi and others, 2020). Likely as a result of the low river inflows during the May survey, NH₄ was not highly dispersed; observed concentrations of NH₄ were highest in the Sacramento River immediately downstream of the SRWTP discharge point and diminished with distance in the distributary channels to the east and west. High concentrations of NH₄ also were seen in the main stem of the lower Sacramento River at the intersection with Cache Slough. This area is a hydrodynamic junction where the comparatively low-volume upper Sacramento River meets the much-higher-capacity tidal river formed by the lower Cache Slough and lower Sacramento River. Consequently, there is an abrupt lowering of seaward transport at this junction and a concomitant increase in residence time that permits NH₄ to accumulate as it is slowly mixed by tides north into the CSC and southwest toward Suisun Bay. Similarly elevated NH₄ concentrations also were observed at the junction of Miner Slough with lower Cache Slough as a result of a similar process; Miner Slough, like the main stem of the Sacramento River, contains high concentrations of NH₄ derived from the SRWTP (figs. 2, 3, 7: tab 8, Side by Side Parameter Maps; Bergamaschi and others, 2020). Nevertheless, NH₄ concentrations were attenuated landward into the CSC and seaward into the lower estuary, with lower concentrations measured in Grizzly Bay (fig. 3). Finally, elevated NH₄ concentrations were measured in the Stockton Deep Water Channel, likely of an authigenic source.

DON was markedly elevated during the May survey, with a median concentration of 12 μM, comprising 30 percent of TDN (fig. 8; tab 8, Side by Side Parameter Maps; Bergamaschi and others, 2020). DON concentrations were most elevated in areas near wetlands, such as the mouth of Montezuma Slough, upper Cache Slough, Franks Tract State Recreation Area (SRA), and Mildred Island. These areas also do not have elevated TDN concentrations, indicating that there is a local process that transforms nitrogen species from inorganic to organic forms or that releases DON (for example, plant leachates or decomposition of detritus).

The median PO₄ concentration (2.5 μM) during the May 2018 survey was similar to the median concentrations during surveys in July (2.2 μM) and October (2.6 μM) of 2018, but the distribution was more uniform than in the other surveys (Bergamaschi and others, 2020). The highest PO₄ concentrations were measured near the SRWTP and Stockton RWCF and in the northern CSC (fig. 9; tab 8, Side by Side Parameter Maps; Bergamaschi and others, 2020).

Nutrient Distributions during the July 2018 Survey

The median TDN concentration (26 μM) in July was the lowest of the three surveys (fig. 5; tab 8, Side by Side Parameter Maps; Bergamaschi and others, 2020), but the distribution of TDN and its component forms was substantially different in July as a result of changes in flows, exports, and Delta Cross Channel operations. River flows were more than twice as high in July than in May, but water transported west to Suisun Bay remained similar because exports from the south Delta more than doubled (fig. 4). This meant that proximity to local sources, flow routing, and transit times had a larger effect on the observed nutrient distributions than did river flows. With the Delta Cross Channel open, a larger fraction of nutrient loads transported by the Sacramento River was routed into the Mokelumne River system and the central Delta, and a concomitantly smaller fraction was delivered into the lower Sacramento River and the CSC, meaning SRWTP-derived nutrients were distributed through a larger area; this would tend to lower the median nutrient concentration and increase transit times through the Delta, allowing additional time for transformations to occur.

There are many examples of how flow routing affected the spatial distribution of nutrients. The differences between the North Fork Mokelumne and South Fork Mokelumne Rivers are one example showing the effects of mixing and residence time. The TDN concentrations measured in the northern sections of the North Fork Mokelumne and South Fork Mokelumne Rivers are similar to TDN concentrations measured in the Sacramento River (approximately 35 μM), but the contributions of the component nitrogen species progressively change along the different flow paths, with NH₄ measured throughout the duration of the lower-residence-time in the North Fork Mokelumne River but strongly attenuated in the South Fork Mokelumne River, diminishing to below 3 μM at the southern extent of the reach from a concentration of more than 20 μM near the Delta Cross Channel (fig. 7; tab 8, Side by Side Parameter Maps; Bergamaschi and others, 2020). Nitrate and DON concentrations progressively increase along this reach indicating that the NH₄ is being nitrified and transformed into DON during the approximate 3-day travel time between the Delta Cross Channel and the junction of the North Fork Mokelumne River with the San Joaquin River; however, there may be a demand exerted by the three dead-end sloughs along this reach that affects the extent of change in the reach (figs. 8–11; tab 8, Side by Side Parameter Maps; Bergamaschi and others, 2020).

A large-scale example of how routing changes may impact nutrient transport and distribution was evident along the main stem of the Sacramento River extending from Rio Vista to Suisun Bay. Median concentrations of TDN, NO₃, NH₄, and PO₄ were all lower in the Sacramento–San Joaquin River confluence area—the area just east of Suisun Bay where the Sacramento and San Joaquin Rivers meet—than farther inland. For example, TDN concentrations in the confluence area were about 40 percent of the concentrations observed in the Sacramento River (figs. 5–10; tab 8, Side by Side Parameter Maps; Bergamaschi and others, 2020). This example is consistent with loss during transit due to uptake and denitrification. Nevertheless, concentrations for all of these constituents were higher in Grizzly Bay than in the confluence, indicating that local sources to Grizzly Bay were contributing approximately 40 percent of the TDN and DIN in this region during this time.

Increases in transit time and contribution of local sources similarly affected the distribution of nutrients in the south-central Delta. Total dissolved nitrogen concentrations in Old River were less than one-third of concentrations measured in the Sacramento River, indicating that considerable attenuation occurred during transit, with DON progressively increasing to eventually dominate the total nitrogen pool as water transited south. Middle River, however, clearly showed the influence of the Stockton Deep Water Channel, with elevated concentrations of NO₃, NH₄, and PO₄, and elevated concentrations continuously observed in the direction of the pump intakes, highlighting the importance of nutrient inputs to the south-central Delta from the San Joaquin River and Stockton RWCF.

Nutrient Distributions during the October 2018 Survey

Compared to July, the October survey occurred following a period of lower river flows entering the Delta and higher water exports to the south; exports in October from the south Delta amounted to more than one-half of the flows entering the estuary (fig. 4). This resulted in an increased proportion of San Joaquin River water in the south-central Delta and a strong transport gradient from north to south across the central Delta. Mean flow southward in Old River (USGS site number 11313405; U.S. Geological Survey, 2021) was more than 4,000 ft3/s around the time of the survey, down from peak export flows of more than 6,000 ft3/s.

Although system-wide median concentrations for TDN, NO₃, NH₄, and PO₄ were similar to those reported in other surveys, lower conveyance rates caused steeper concentration gradients than previously observed (figs. 5–11; tab 8, Side by Side Parameter Maps; Bergamaschi and others, 2020). Concentrations of NO₃ more than 150 μM were measured in the San Joaquin River near Stockton, corresponding to TDN concentrations exceeding 180 μM. These high concentrations can be tracked into Middle River and south to the State and Federal Water Project export pumps. Similarly, elevated concentrations were measured in the Sacramento River, which had TDN concentrations above 100 μM and NH₄ concentrations above 90 μM. The junctions of the Sacramento River and lower Cache Slough and of Miner Slough and lower Cache Slough also had elevated nutrient concentrations, particularly for NH₄.

Ammonium was depleted in the high-residence time and high-productivity areas previously discussed: the north CSC, Disappointment Slough, and Grizzly Bay (figs. 2, 3, 7; tab 8, Side by Side Parameter Maps; Bergamaschi and others, 2020). Also, measured concentrations of NH₄ were low in Old River, consistent with longer transit times from wastewater treatment plant (WWTP) sources to reach this area. In Suisun Bay (fig. 2), except for the depleted high-productivity area in the northeast corner of Grizzly Bay, concentrations of NH₄ were generally high in October and were more than 15 μM in the south, demonstrating the contribution of wastewater discharge from local sources; lower NH₄ concentrations (about 5 μM) measured in the Sacramento–San Joaquin River confluence area during this period indicate that inputs from the Delta likely are not causing this elevation.

Total dissolved nitrogen and component species were more elevated in detached areas during the October survey than in May and July, showing substantial depletion only in the northern CSC and in the Old River corridor (figs. 3, 5B). Interestingly, TDN showed little sign of substantial attenuation during transit down Old River to the pumps, although NH₄ and PO₄ demonstrated longitudinal profiles consistent with biological transformation (figs. 5–11; tab 8, Side by Side Parameter Maps; Bergamaschi and others, 2020). This attenuation is likely the result of high export rates increasing transport through the reach such that transit times are too low to allow for substantial transformation, thus complicating efforts to adequately quantify changes in concentration. Transitioning from summer to fall results in a slowing of biological processes across the Delta aquatic landscapes due to lower temperatures and solar radiation. As discussed in the following section, biological transformation in October was still considerable as evidenced by the strong attenuation in the northern CSC.

Phytoplankton in the Delta

The status of phytoplankton communities in the Delta has been a focus of research because of notable declines in abundance in recent decades (Jassby, 2008; Cloern and Jassby, 201012), which have been linked to declines in zooplankton abundances—the prey of many endemic pelagic fish species (Winder and Jassby, 2011). There is particular interest in the abundance and distribution of diatoms within the Delta because they are thought to be the phytoplankton class most beneficial to imperiled aquatic food webs (Glibert and others, 2014b; Galloway and Winder, 201523). Some researchers have suggested that elevated concentrations of NH₄ in the north Delta and estuary have resulted in conditions that reduce overall primary production and limit the production of diatoms in favor of blue-green algae, which are thought to be less beneficial to Delta food webs (Parker and others, 2012a; Parker and others, 2012b; Glibert and others, 2014b27; Lehman and others, 2015), though other research indicates controls on phytoplankton are more complex (Senn and Novick, 2014; Ward and Paerl, 2016; Kraus and others, 2017b; Senn and others, 2020; Stumpner and others, 2020). Another issue contributing to interest in phytoplankton abundance and community composition is the increasing prevalence of harmful algal blooms, which in fresh waters comprise principally blue-green algae of various types (Harke and others, 2016). Recent research in the Delta has shown an increasing incidence of Microcystis, a blue-green algae capable of forming toxins (Lehman and others, 2013, 202243).

Spatial and Seasonal Variation in Phytoplankton Abundance and Community Structure

The surveys included characterization of the spatial distribution of phytoplankton abundance and phytoplankton community structure at the class level for the purpose of relating their abundance to associated concentrations of nutrients and other water-quality conditions. The classes characterized were diatoms, green algae, blue-green algae, and cryptophytes. We discuss each survey below.

In addition to the high resolution mapping, samples also were collected at specific locations to characterize the size distribution of chlorophyll (Chl) and to allow for identification and enumeration of phytoplankton at the species level by microscopic examination. The Chl size distribution provides information useful for assessing the relative benefit of phytoplankton to aquatic food webs because organisms smaller than 0.5 μm—typically mostly blue-green algae—are thought to provide less overall direct benefit to copepods, a mainstay of endemic pelagic fish diets (Kimmerer and others, 2018). Direct identification and enumeration of phytoplankton provides confirmation of field measurements and provides species-level abundance information for the larger phytoplankton species.

Phytoplankton Distributions during the May 2018 Survey

Phytoplankton abundances, as indicated by median Chl concentration across the whole study area, were similar for all three surveys at about 3–4 micrograms per liter (μg/L), but there were distinct spatial differences among months (fig. 12; tab 4, Nutrient Maps and Study Reaches and fig. 13; tab 8, Side by Side Parameter Maps; Bergamaschi and others, 2020). During the May survey, Chl concentrations across the Delta were generally low (less than 10 μg/L) except in the western Delta and Suisun Bay areas in Grizzly and Honker Bays (fig. 3) where concentrations of blooms reached 20–50 μg/L. These areas with elevated Chl concentrations also had low NO₃ concentrations (fig. 10; tab 1, Nutrient Parameter Maps and fig. 12; tab 4, Nutrient Maps and Study Reaches; Bergamaschi and others, 2020), indicating phytoplankton production was likely responsible for NO₃ drawdown.

Across the whole study area, based on FluoroProbe data, diatoms accounted for 12–94 percent of the Chl for the May survey with a median value of 36 percent—the highest of all 2018 surveys (fig. 11; tab 8, Side by Side Parameter Maps; fig. 12; tab 1, Nutrient Parameter Maps; Bergamaschi and others, 2020). System wide, there was a strong positive correlation between Chl concentrations and diatom populations (fig. 14; tab 5, Parameter v. Parameter; Bergamaschi and others, 2020). In some local areas, diatoms comprised the largest fraction of the phytoplankton community, particularly in Suisun Bay (greater than 75 percent) and in the northern areas of Grizzly and Honker Bays, where there appeared to be an intense bloom during the May survey and the median contribution of diatoms exceeded 90 percent. The Sacramento River also had a high fraction of diatoms (median=58 percent of total Chl) despite the high NH₄ concentrations present in the reach downstream of SRWTP’s discharge location. Diatoms were not the dominant phytoplankton group present in some areas that were showing elevated Chl concentrations and nutrient depletion, such as in the upper CSC, where blue-green algae comprised greater than 40 percent of the phytoplankton community. Even so, diatoms in the CSC were present in considerable abundance, accounting for about 30 percent of the total Chl. Diatoms and blue-green algae also had comparatively similar contributions to the Chl signal in the Sacramento River Deep Water Ship Channel (fig. 2) during this period, together accounting for 90 percent of total Chl, though Chl concentrations were not elevated. The southern reaches of Suisun Bay had a similar distribution in May, with blue-green algae making up about 50 percent and diatoms about 40 percent of the total Chl concentration of 3 μg/L. Concentrations of Chl in the south Delta were typically less than 10 ug/L, with populations of green algae comprising 30–50 percent of the total Chl in Old and Middle Rivers.

Phytoplankton Distributions during the July 2018 Survey

The distribution of phytoplankton was markedly different in the July survey compared to the May and October surveys, with several areas exhibiting locally elevated Chl concentrations (fig. 12; tab 4, Nutrient Maps and Study Reaches; figs. 13, 15; tab 1, Nutrient Parameter Maps; Bergamaschi and others, 2020). Compared to the median July value of 4.3 μg/L for the whole study area, the central and south Delta regions had median Chl concentrations that were about 50 percent greater (6.3 and 6.7 μg/L, respectively), with locally high values reaching 10 and 15 μg/L, respectively (fig. 12; tab 4, Nutrient Maps and Study Reaches; fig. 13; tab 8, Side by Side Parameter Maps; fig. 15; tab 7, Phytoplankton Data Explorer; Bergamaschi and others, 2020). Chlorophyll concentrations also were higher than the median value in Middle River, Disappointment Slough, and most of the South Fork Mokelumne River, particularly Hog Slough. There also were higher concentrations of Chl in the lower Sacramento and San Joaquin Rivers, near the confluence. Chlorophyll concentrations in the Sacramento River, the North Fork Mokelumne River, and the northern CSC were all below the entire Delta and Suisun Bay study area’s July 2018 median value of 4.3 μg/L. The northern CSC DIN concentrations were particularly low (less than 6 μM) during the July survey, indicating possible nutrient limitation in this higher water-residence time area. However, phytoplankton abundances appeared generally low even in areas exhibiting high DIN concentrations.

There were distinct spatial patterns in the phytoplankton community composition in the July survey compared to the May and October surveys. Across the Delta and Suisun Bay, blue-green algae comprised 50 percent of the total Chl based on the median concentrations. In the south Delta, blue-green algae accounted for 50–90 percent of the elevated Chl concentrations as compared to 26–55 percent in the May survey (fig. 13; tab 8, Side by Side Parameter Maps; Bergamaschi and others, 2020). In the upper San Joaquin River, in an area replete with nutrients, blue-green algae accounted for nearly 80 percent of the total Chl. Blue-green algae also accounted for nearly all of the Chl in the low-nutrient areas of the northern CSC.

Despite the predominance of blue-green algae across the full study area in the July survey, there also were localized areas where diatoms dominated the phytoplankton Chl signal (fig. 12; tab 4, Nutrient Maps and Study Reaches; Bergamaschi and others, 2020). These diatom-rich areas were associated with the higher (greater than 10 μg/L) Chl concentrations measured in the confluence area and the South Fork Mokelumne River, where locally diatoms accounted for greater than 70 percent of the total Chl. The abundance of diatoms was not clearly related to any of the parameters measured during the survey, including TDN, DIN, PO₄, temperature, turbidity, depth, or any other potentially explanatory variable.

Phytoplankton Distributions during the October 2018 Survey

Phytoplankton abundances during the October survey were somewhat lower (mean concentration of 2.9 μg Chl/L) than seen in the May and July surveys, with only a few areas exhibiting concentrations approaching bloom conditions: the northern corner of Grizzly Bay, the southern end of the Stockton Deep Water Channel, and the eastern sloughs off the South Fork Mokelumne River (fig. 12; tab 4, Nutrient Maps and Study Reaches; Bergamaschi and others, 2020). Though lower than bloom conditions, modestly elevated phytoplankton concentrations were observed generally in Grizzly Bay, in the Sacramento–San Joaquin River confluence area and in the northern CSC; the lowest concentrations were observed in the Sacramento River, Mokelumne River, and most of the central and south Delta.

Green algae were more prominent in the October survey than the May and July surveys, accounting for a median of 38 percent of the total Chl across the study area (fig. 13; tab 8, Side by Side Parameter Maps; Bergamaschi and others, 2020). Green algae were particularly abundant in the south Delta, where together with blue-green algae, they accounted for greater than 90 percent of the total Chl, split nearly evenly; they also comprised half or more of the phytoplankton community in the very distal reaches of the northern CSC, in Grizzly Bay, and in the San Joaquin River near the Stockton Deep Water Channel—areas with elevated Chl. Abundance of Chl was generally inversely related to the fraction of blue-green algae and directly related to the fraction of diatoms in the phytoplankton community during this period (fig. 12; tab 4, Nutrient Maps and Study Reaches; fig. 13; tab 8, Side by Side Parameter Maps; Bergamaschi and others, 2020).

The area in Grizzly Bay with elevated Chl nearing bloom concentrations (20 μg/L) also had high abundances of diatoms, comprising up to 50 percent of the Chl. This was despite NH₄ concentrations above 4 μM, the nominal concentration thought to inhibit formation of diatom blooms (Dugdale and others, 2007; Wilkerson and Dugdale, 2016).

Landscape-Scale Observations

The purpose of this study was to improve understanding of how hydrodynamics, landscape features, and aquatic primary productivity interact to drive nutrient cycling and transport in the Delta, providing insights into the underlying processes most directly responsible for the conditions at the time of this study, and thus most likely to dominate following changes to nutrient inputs and biogeochemical processes. One major anticipated change includes upgrades to wastewater treatment plants (Senn and others, 2020), but the study also informs our understanding of how other nutrient-management actions, large-scale wetland restoration, drought, levee failure, and changes to water management might affect phytoplankton communities in the Delta.

In addition to nutrient loading rates, which are the primary driver of nutrient concentrations found in the Delta, this analysis distinguished among three other large-scale drivers that act to shape the distribution and effects of nutrients: hydrodynamics, interaction with landscape features, and aquatic productivity. Hydrodynamics variously affects timescales of transport in and across the Delta, responding to changes in flow conditions, but also interacting with channel geometry, tidal mixing, exports from the south Delta, and other water withdrawals in ways that influence water-residence and transport times. These effects are disproportionately arrayed across the Delta, substantially affecting the range of times over which natural internal cycling can act to alter nutrient concentrations and forms. Interaction with Delta aquatic landscapes such as herbaceous tidal marsh, submerged aquatic vegetation, and large expanses of intertidal or subtidal sediments, all highly productive landscapes, are expected to exert some level of demand on available nutrient supplies in support of the high productivity. Finally, phytoplankton also require nutrients to sustain production and thus are a potential nutrient sink, but the amount and form of nutrients also affects the type of phytoplankton and rates of production, potentially affecting the occurrence of harmful and beneficial algal blooms.

To assess the relative effects of these drivers on nutrient concentrations and distributions, we examined regions and reaches within the Delta where we observed, or expected to observe, relatively substantial transformations in the forms and concentrations of nutrients. This was done by using the explicitly designated regions and reaches defined above (fig. 3) and used in the visualizations (https://ca.water.usgs.gov/bay-delta/2018-delta-wide-mapping-surveys.html#Contents; Bergamaschi and others, 2020), as well as by identifying additional localized gradients and hot spots that were apparent during a given survey (for example, Grizzly Bay, the northern reaches of the CSC, and the confluence area). Landscape features we examined included open-water areas, inter-tidal or sub-tidal shallow areas, and tidal wetlands such as the open-water areas of Liberty Island and Franks Tract Franks Tract SRA, the intertidal areas in Grizzly and Honker Bays, and the large tidal wetland extents in the CSC, and Sherman Island (fig. 3). We also focused on areas with characteristic hydrodynamics, such as reaches with extended transport times, areas with long water-residence times, and channels with known mixing zones. Example areas with extended transport times include the lower Sacramento River and Old River. Channels with long water-residence times are typically dead-end sloughs such as Shag Slough and Hog Slough (fig. 3). Locations with characteristic mixing zones also are associated with these dead-end sloughs, at the limit of tidal exchange.

Landscape-Scale Variation in Nutrient Distributions

Some areas within the Delta stand out at the regional scale, having persistent nutrient and Chl distributions that span all survey months. Certainly, the most persistent of the features are the elevated nutrient concentrations near the SRWTP, Central Contra Costa Sanitary DTP, and Stockton RWCF discharge locations (figs. 2, 5–13; tab 8, Side by Side Parameter Maps; Bergamaschi and others, 2020). In all three surveys, high concentrations of NH₄ and PO₄ were observed in the Sacramento River in the reach immediately downstream of the SRWTP discharge location, with concentrations exceeding 30 μM NH₄ and 2.5 μM PO₄. Concentrations of NO₃ generally increased with distance in the Sacramento River downstream of Freeport as NH₄ was oxidized to NO₃ via nitrification. Elevated NO₃ and PO₄ concentrations characterized the waters near the Stockton RWCF discharge location, with NO₃ concentrations ranging between 30 and 165 μM and PO₄ concentrations ranging between 4 and 20 μM. There was also some evidence of elevated NH₄ concentrations near Stockton, particularly in July, that might be attributed to fluxes from the sediments.

Although the physical processes in Suisun Bay are much different than the Sacramento and San Joaquin Rivers, there also were persistently elevated nutrient concentrations in the vicinity of the Central Contra Costa Sanitary DTP’s discharge location. The physical processes and hydrodynamics in Suisun Bay near this discharge location result in substantial mixing and dilution that rapidly lower the overall concentration of any inputs; bi-directional tidal discharge is more than 10 times larger in this region compared to the Sacramento River discharge, and the volume of water available for mixing is concomitantly much greater. Nevertheless, concentrations of NO₃ and NH₄ were elevated in the vicinity of the Central Contra Costa Sanitary DTP’s discharge location together with concentrations of PO₄ and TDN, except for in May, when a large phytoplankton bloom appears to have drawn down available nutrients.

The CSC in the north Delta also exhibited similar nutrient distributions across all months. The CSC comprises Liberty Island, Little Holland Tract, Prospect Slough, Shag Slough, and related channels (figs. 2, 3). As such, the CSC is characterized by large areas of open water and substantial areas of tidal wetlands, with few external inputs during the survey periods other than Sacramento River water being advectively and dispersively transported from the south up into the area. During all months, NH₄ is virtually depleted within a few hundred meters of the southern boundary, and NO₃ concentrations are strongly depleted from the southern boundary of the CSC from the south to the north (figs. 7, 10; tab 8, Side by Side Parameter Maps; Bergamaschi and others, 2020). In contrast, PO₄ concentrations more than doubled along the same south to north transect. These persistent nutrient distributions in the CSC highlight the importance of wetlands and long water-residence times. Although NH₄ is readily assimilated or nitrified, given sufficiently long exposure times, NO₃ also undergoes uptake and transformation, declining by an average of nearly 80 percent from the south to the north of the CSC (fig. 11; tab 8, Side by Side Parameter Maps; Bergamaschi and others, 2020). A recent study of water-residence times in the CSC (Downing and others, 2016) indicates that timescales of this transformation are on the order of several weeks. However, that the observed northward decline of DIN does not equate with a loss of nitrogen to denitrification and uptake—increases in DON along this transect offset the losses in DIN, resulting in an average TDN loss of 30 percent (fig. 16; tab 4, Nutrient Maps and Study Reaches; Bergamaschi and others, 2020).

Because we observed net changes in concentrations, we cannot definitively assign changes in NO₃ concentration to specific processes such as uptake, denitrification, nitrification, mineralization, or benthic fluxes because these can all occur simultaneously. However, these results indicate that NO₃ is lost to uptake and denitrification and transformed into DON in relatively equal measure based on mass-balance data. Nevertheless, the ecological role of DON is not well understood because its biological lability is uncertain. Further, the provenance and age of the DON found in the upper Cache Slough is unknown; rising concentrations to the north may represent accumulation of wastewater-derived material in the same way PO₄ accumulates in the area, or DON may be produced over longer timescales than the attenuation of available DIN. More than 80 percent of available TDN in the area is DON; therefore, understanding the source and ecological role of DON would likely improve understanding nitrogen dynamics and its relation to phytoplankton communities in the Delta.

Although we observed strong gradients and extensive regional-scale effects in the CSC during all three surveys, we did not consistently see similarly strong gradients in nutrients and phytoplankton in other locations where they were expected. For example, in the vicinity of Franks Tract SRA, we expected to see relatively strong gradients associated with this large open-water feature. Franks Tract SRA is a large, shallow, open-water area in the central Delta characterized by relatively clear waters and extensive populations of submerged aquatic vegetation (Ta and others, 2017; fig. 2). Although the boat-based survey was not able to pass through Franks Tract SRA because of navigation hazards, tidal exchange is substantial between Franks Tract SRA and the surrounding channels that were included in the survey. Given that the volume of tidal exchange exceeds the volume of surrounding channels, we expected to observe strong gradients in nearby channels, yet only modest potential effects were observed in May and October (figs. 5–11; tab 8, Side by Side Parameter Maps; Bergamaschi and others, 2020). Therefore, we conclude that biogeochemical processes in Franks Tract SRA did not appreciably affect nutrient concentrations and distributions during these two surveys. In July, when water temperatures were warmer than in May (25 °C versus 19 °C), there was limited evidence that this region may have been a sink for DIN. To the west and north of Franks Tract SRA, DIN concentrations were about 25 μM; to the south in Old River, the measurably lower concentrations (15 μM) might have been caused by extended residence times that resulted from changes in the balance between exports and outflows (fig. 6; tab 6, Nutrient Data Explorer; Bergamaschi and others, 2020). We also expected to observe substantial DIN loss during the estimated weeks-long transit down Old River from Franks Tract SRA to Clifton Court Forebay (fig. 2), but none was observed in any month of the study (fig. 17; tab 4, Nutrient Maps and Study Reaches; Bergamaschi and others, 2020). These findings indicate that interaction with wetlands, submerged aquatic vegetation beds, and open-water areas characterizing the central Delta does not appreciably alter nutrient concentrations during intermediate water-residence times; comparisons of nutrient concentrations and residence times between the central Delta and the CSC indicate that longer residence times have a larger effect on nutrient concentrations.

Observed nutrient gradients from these broad-area surveys provide substantial evidence that water-residence time is a more important determinant of nutrient distributions than other landscapes features such as open water, wetland density, and the presence of submerged aquatic vegetation. The relation between water-residence time and nutrient concentrations is also evident at the local scale, and focusing on the local scale allows us to identify the separate effects of transformation and loss. For example, downstream of Freeport, the Sacramento River exhibited a modest, but progressive downstream increase in NO₃ in all three survey periods, with values increasing 2–3 μM over a 10-mi river reach (fig. 18; tab 4, Nutrient Maps and Study Reaches ; Bergamaschi and others, 2020). This downstream increase in NO₃ has been previously observed in continuous-station data collected on this same river reach and is attributable to the transformation of wastewater-derived NH₄ to NO₃ via nitrification (Kraus and others, 2017c). As described above, during the extended water-residence times seen in the dead-end Hog and Sycamore Sloughs, NH₄ and NO₃ concentrations were reduced to near the detection level within 1 mi of the junctions of Hog and Sycamore Sloughs with the Mokelumne River, much like the attenuation seen in Shag Slough in the CSC (fig. 19; tab 4, Nutrient Maps and Study Reaches; Bergamaschi and others, 2020); Bergamaschi and others, 2020).

An unexpected finding that is evident from this spatially rich dataset is that Suisun Bay (including Honker and Grizzly Bays) receives substantial nutrient loadings from sources other than the Delta. This is most distinct for the July 2018 surveys where NO₃, NH₄, DIN, and TDN concentrations were all lower in the confluence area than in Suisun Bay (figs. 5–11; tab 8, Side by Side Parameter Maps; Bergamaschi and others, 2020). Although some of this nitrogen may originate from the Central Contra Costa Sanitary DTP, inputs entering this region from Suisun marsh to the north and San Francisco Bay to the west also are plausible.

Plankton blooms also can shape the distribution of nutrients across Delta landscapes. As noted earlier, there was an unusually intense phytoplankton bloom in the landward part of Grizzly Bay during the May sampling, with spatially averaged Chl concentrations nearing 50 μg/L and elevated DO and pH values, which are all indicators of intense bloom activity (figs. 20–21; tab 8, Side by Side Parameter Maps; Bergamaschi and others, 2020). Throughout the spatial extent of the bloom, NO₃ and DIN were substantially reduced—from more than 30 μM to approximately 3 μM. Given that this geographic area is tidally well mixed and has low characteristic water-residence times, the rate of nutrient loss is substantial, and there is evidence that the bloom reduced ambient nutrient concentrations across Suisun Bay. Other studies have shown similar levels of nutrient loss during bloom events in other Delta locations (Bergamaschi, 2018), but the frequency with which such events occur is unknown (Glibert and others, 2014a).

Synthesis of Findings

The spatial surveys conducted during this study allowed for the evaluation of the relative magnitude of the many drivers affecting nutrient concentrations and distributions across the Delta. The study showed that nutrient concentrations and losses within and across the Delta are predominantly controlled by hydrodynamic processes that effect transport, mixing, and water-residence time. At the broadest scale, the distribution of nutrients across the Delta is controlled by the locations of major nutrient sources and the transport pathways to the San Francisco Bay and to the south Delta State and Federal Water Project export facilities. In addition to well recognized inputs to the Sacramento and San Joaquin Rivers from WWTPs and agriculture, inputs to Suisun Bay also were apparent when examining this dataset (Bergamaschi and others, 2020). Clear gradients radiating from these input locations were found along major hydrologic flow paths, with the location of the gradients shaped by the local hydrodynamic interaction between tidal currents and local channel geometries, which together affected the extent of mixing and water-residence time. Consequently, nutrient concentrations and distributions varied seasonally in relation to river discharge.

We did not find evidence of any appreciable nutrient point sources internal to the Delta aside from the WWTPs. Nevertheless, given that the nature of the surveys was to identify spatial variation, we cannot rule out contribution from non-point source releases that occur over a broad area, such as fluxes from the sediments or decomposition of detrital material. However, magnitudes of internal sources for nitrogen did not appear to be sufficient to result in clear net increases of nutrient concentrations during transport through any of the reaches studied. This does not mean there are no internal nutrient sources in the Delta, but rather indicates that internal sinks (i.e., uptake and transformation) balance out these sources during transport.

The study also provided information about the relative importance of different mechanisms for transformation and loss of nutrients within the Delta. Nitrogen transformations from NH₄ to NO₃ were shown to occur over a timescale of days, such as was observed in the Sacramento River, whereas substantial net loss of total DIN to uptake or denitrification occurred over much longer timescales. Extensive losses occurred only in areas with limited exchange and concomitantly long water-residence times. For example, features such as dead-end sloughs and the northern CSC exhibited low DIN concentrations and strong concentration gradients in all seasons, elevating in concentration with proximity to nearby tidal channels; concentration gradients were also evident in the Mokelumne River system. The DIN concentrations coming from the Sacramento River in July and October were lower in the South Fork Mokelumne River compared to the North Fork Mokelumne River, which may be attributed to the three large dead-end sloughs that extend off the South Fork Mokelumne River. The dead-end sloughs were characterized by steep declines in DIN concentrations over a short distance. These persistent gradients signify that such sloughs and similar areas are a constant sink for DIN in the Delta, but understanding the magnitude of the sink was beyond the scope of this study.

Although much of the emphasis of this study was on nitrogen, the concentration and distribution of wastewater-derived phosphate (as PO₄) also was examined. The environmental chemistry of PO₄ is different from nitrogen because, although nitrogen may be lost from the system as N₂ through denitrification, PO₄ has no analogous loss pathway. Also, examination of sources and sinks for PO₄ is more challenging because concentrations of PO₄ are typically much lower than nitrogen species, which makes direct observation of uptake and loss more difficult. Though the study documented the wastewater sources of PO₄ and the distribution of this material through hydrodynamic flow paths, we did not identify any major net internal sinks or sources for PO₄. We did find an increase of PO₄ to three or more times the average concentrations in areas with long water-residence times such as in the CSC on all three dates, indicating that dissolved PO₄ exceeded the local demand.

The comprehensive measurements collected in this study revealed the extent to which DON generally comprises a substantial fraction of total nitrogen in the water column, becoming the dominant form of dissolved nitrogen in areas with longer residence times. Given the locations where high concentrations are reported, the source of this material is presumptively wetlands, including sediments and aquatic vegetation, but formation from other sources cannot be ruled out. Future study of the ecological significance of this available nitrogen would provide valuable insights, especially considering prospective future decreases in DIN inputs due to WWTP upgrades.

Large landscape features such as open-water areas, tidal wetland extent, submerged aquatic vegetation beds, and inter-tidal zones did not exert a clear net demand on available DIN that was sufficient to appreciably alter its distribution and concentration. One hypothesis guiding this study was that areas characterized by substantial primary production, such as herbaceous marsh and submerged aquatic vegetation, would deplete the surrounding waters of nutrients, but this hypothesis was not supported by the results. This may be because hydrologic transport of nutrient-rich waters into many of these areas is sufficient to maintain relatively low water-residence times and thus provide a continuous supply of DIN. Alternatively, even with substantial nutrient uptake, sufficient nutrients may be made available through recycling and/or through remineralization of decaying plant and soil organic matter (i.e., nutrient sinks were not greater than sources). Many areas of the Delta, otherwise depleted in nutrients and isolated from active supply, exhibited NH₄ and NO₃ concentrations near or below detection limits, which is consistent with active authigenic production followed by rapid uptake. These findings indicate that the demand exerted on nutrient supplies by such large-scale landscape features may not appreciably alter the effects of prospective source control actions and wetland restorations on nutrient concentrations in the major channels of the Delta.

A major finding of this study was that phytoplankton blooms can substantially deplete DIN concentrations throughout their spatial extent and affect concentrations in surrounding areas more than we originally expected. This depletion was particularly pronounced in Grizzly Bay in May, when a large bloom comprising more than 85-percent diatoms accounted for a nearly 90-percent depletion of available inorganic nitrogen, demonstrating that phytoplankton growth can be a substantial sink for nitrogen in the Delta and Suisun Bay. Given that phytoplankton—particularly diatoms—are appreciably grazed by zooplankton and clams, uptake of nitrogen into phytoplankton biomass may be an efficient means of transporting nutrients to the lower food web.

Although phytoplankton production and uptake clearly influenced nutrient concentrations, it was not apparent from the data collected during these three surveys that nutrient concentrations or forms were a dominant control on phytoplankton. However, the lack of bloom conditions during the surveys makes it challenging to query this dataset for clues as to what triggers and sustains phytoplankton growth. Our data show that diatoms can flourish in Suisun Bay despite the presence of appreciable NH₄, and that blue-green algae and green algae are found across the system, often dominating the Chl biomass. The new capabilities demonstrated in this study—simultaneous high resolution surveys of all major nutrient forms, Chl concentrations, water-quality parameters, and phytoplankton community structure—will provide information about the formation and food-web value of phytoplankton community, including beneficial algae like diatoms and harmful algae like cyanobacteria.

The spatially and temporally rich dataset generated by this study provides Delta managers the ability to gain new insights into the distribution of nutrient concentrations and forms, what factors drive nutrient gradients, and how nutrients along with other factors shape the phytoplankton community. The collection of this spatially high resolution dataset allows us to examine regional and local patterns in nutrients and phytoplankton in relation not only to point source inputs, but also to landscape-scale features such as dead-end sloughs with long water-residence times, shallow-water habitats, and interaction with different kinds of restored wetlands. Modelers can use these data to develop improved representation of biogeochemical processes to help assess prospective conditions relevant not only for nutrients, but also for water quality, phytoplankton, and aquatic vegetation.

Additional broad-area, high resolution surveys of the Delta that occur during dry, wet, and average water years and that capture phytoplankton bloom conditions would provide a means to better assess the interactions between nutrient supplies and desired environmental conditions. Moreover, survey data collected following WWTP upgrades would capture how reductions in point source nutrient inputs and the change in the form of nitrogen from NH₄ to NO₃ would affect nutrient and phytoplankton distributions in the Delta and Suisun Bay. Such data would allow us to more definitively identify whether DIN inputs to the CSC and Suisun Bay can sustain and improve local production, whether a reduction in NH₄ concentrations would lead to improved growth of diatoms, or whether lower nutrient availability would hinder the growth of harmful algal blooms. Collecting these data over multiple years also would capture effects associated with wetland restoration efforts, managed flow actions, and changes in hydrologic flow paths. If more data of this type were collected, we could gain a better understanding of nutrient status and trends, identify sources and sinks, and ascertain the roles that hydrologic and biological drivers play in determining nutrient distributions, concentrations, and effects in the Delta.