Predictors of Algae Growth in Estuary Systems: A Case Study in Suisun Bay, California

Abstract

Eutrophication is a widespread ecological issue affecting freshwater and marine ecosystems globally. California’s coastal regions are particularly vulnerable to eutrophication due to agricultural intensity and dense population centers, while various underlying abiotic factors drive the occurrence of eutrophication. This paper focuses on understanding the drivers and patterns of eutrophication in the Sacramento-San Joaquin River Delta, specifically emphasizing individual major algal growth predictors and their collective influence on total algae concentration. USGS water quality data from Van Sickle Island in Suisun Bay, California, collected from June 2020 to May 2023, was analyzed using multivariate linear regression analysis. The study found that the predictors of temperature and turbidity were positively related to chlorophyll concentration. Although considered a predictor, nitrate concentration showed a weaker correlation with chlorophyll. The analysis revealed that the monthly average data for each algae growth predictor best fit the relationship with chlorophyll. These findings suggest that turbidity may be the dominant factor influencing algal growth in the Suisun Bay estuary system, while nutrient availability, specifically nitrates, may not be a primary driver. The results highlight the need to consider waterbody temperature and turbidity as important variables in developing effective management and mitigation practices around eutrophication.

Keywords: eutrophication, algal growth predictors, water quality management, Sacramento-San Joaquin River Delta

Introduction

Eutrophication describes the sudden or gradual enrichment of water bodies with excess nutrients, particularly nitrogen and phosphorus, over time. It is a major driver of ecological shifts and community disturbances influencing freshwater and marine ecosystems globally. Eutrophication events can occur naturally due to shifts in temperature and precipitation, but have increased in frequency and intensity in recent times due to factors associated with heightened runoff from anthropogenic activities [1] such as urbanization [2], intensified agriculture [3], and elevated water surface temperatures [4], [5], induced by anthropogenic climate change. Algae are primary producers, converting sunlight, carbon dioxide, and nutrients in the water into organic matter through photosynthesis [6]. This provides food for organisms at higher trophic levels, such as zooplankton and fish, and oxygen to support the survival of macroinvertebrates [7], [8]. Algal growth is limited by nutrient availability in an aquatic environment, but excessive nutrient enrichment can favor unrestrained growth of algal species that release toxic compounds into the waterbody. These harmful algal blooms can negatively impact the health of both aquatic life and humans exposed to the toxins, thus posing a public health risk [9]. Uncontrolled growth can also lead to an increase in organic matter and sedimentation in the waterbody regardless of algal species, which can result in adverse changes to the physical, chemical, and biological characteristics of the ecosystem, including reduced oxygen levels [10], altered species composition [11], and disruptions to the food web. Therefore, it is essential to understand the factors driving observed eutrophication.

Eutrophication is increasingly problematic in the Sacramento-San Joaquin River Delta of Northern California [12]. This follows a state-wide trend of harmful algal blooms in coastal and estuarine environments related to wastewater discharge from agricultural and urban sources [11], [13]. Contents of such discharge include nitrogenous products such as urea, which are correlated with harmful algal blooms along California’s coastlines [14]. The concentration of dissolved nutrients, chlorophyll-a, and organic contaminants is highest at coastal and estuary sites where agricultural wastewater discharges are dominant [15]. At the water source or watershed level, agricultural operations, temperature, and turbidity patterns may drive observed algal growth in estuary ecosystems [16]. Examining these variables can help establish potential indicators and the drivers behind eutrophication at the site-specific scale.

Here, we aimed to assess patterns of eutrophication in one estuary site within the Sacramento-San Joaquin River Delta, focusing on putative predictors of algal growth and how they collectively relate to the total concentration of algae. We analyzed estuary water quality data to investigate the following questions:

1. Which water quality parameters are most associated with the patterns of measured algal concentration, approximated by chlorophyll-a measurements?

2. Is eutrophication occurring in the water body?

The ultimate goal of this research is to inform the efficacy of management practices and policies to reduce nutrient loads, protect important estuary ecological communities, and conserve water resources.

Methods

Site Description

This study used existing water quality data from June 3, 2020, to May 19, 2023, collected offshore near Van Sickle Island (38.04° N, 121.88° W). Van Sickle Island is located in Suisun Bay near Pittsburg, California, just downstream of the confluence of the Sacramento and San Joaquin Rivers (Figure 1). Van Sickle Island is a small, inland island at the head of the Suisun Bay, a large and shallow bay fringed with tidal marsh and located within the brackish portion of the San Francisco Estuary. Suisun Bay is host to a diversity of estuarine organisms, including endangered fish species like the delta smelt (Hypomesus transpacificus), and phytoplankton like diatoms and blue-green algae (i.e. cyanobacteria) [17]. Suisun Bay receives inflow from irrigated agriculture, wastewater, and urban runoff traveling downstream from the Sacramento and San Joaquin Rivers and can be used to explain baseline water quality in Suisun Bay. Upstream of Suisun Bay is San Francisco Bay, where eutrophication and harmful algal blooms are well-documented [18], [19], [20].

Water Quality Acquisition

To analyze the relationship between algal growth and potential predictors, we downloaded time-series water quality data from the United States Geological Survey (USGS) for Van Sickle Island (Site: 11455508) from June 3, 2020, to May 19, 2023 [21]. We considered the following water quality parameters: chlorophyll fluorescence, nitrate plus nitrate, temperature, and turbidity (Table 1).

The data was collected every 15 minutes autonomously at the Van Sickle site with sensors provided and maintained by the USGS. Gaps in the data were present, varying from single 15-minute readings to more significant gaps due to malfunctions in the sensors and periods of non-collection where sensors are removed for maintenance. For example, a period of no chlorophyll fluorescence data exists from August 15, 2021 to October 6, 2021. Chlorophyll fluorescence, temperature, and turbidity data were collected since September 16, 2016, but nitrate plus nitrite data was not collected until May 20, 2020. Finally, large subsets of data for each parameter collected in the past year were marked as provisional by the USGS, meaning the data has yet to be approved due to additional field inspection and measurements required.

Chlorophyll fluorescence measures the amount of light re-emitted by photosynthetic organisms containing chlorophyll-a in a given volume of water and represents a proxy for algal concentration in the waterbody [22], [23]. The parameter is hereafter abbreviated as “chlorophyll.” Nitrate plus nitrite is a measurement of dissolved nitrates (NO₃^- ) available for organisms in the waterbody, which is known to influence algae growth dynamics [24]. Suisun Bay is a nitrogen-limited waterbody, as opposed to phosphorus or potassium-limited; therefore, NO₃^- can be used as a proxy for the concentration of overall limiting nutrients in the waterbody [19]. Nitrate plus nitrite is hereafter abbreviated as “nitrate.” Temperature was incorporated into the analysis as algal species have optimal temperature ranges for growth [25], and temperature affects the rates of chemical and biological processes. Turbidity measures the cloudiness of the water and is influenced by suspended sediment and organic matter, which we considered a proxy measurement of light availability in the system for algae photosynthesis [26].

Methods of Data Analysis

To analyze the relationship between algal growth and potential predictors (nitrate, temperature, and turbidity), we performed multivariate linear regression analysis in Microsoft Excel v. 2311  using the data analysis add-in feature with chlorophyll as our outcome variable and nitrate, temperature, and turbidity as predictor variables. The following equation represents the mechanics of the regression analysis:

pC = m₀ + m_NN + m_T1T₁ + m_T2T2

Variables are defined as follows:

• pC = predicted chlorophyll
• m₀ = intercept
• m_N = slope coefficient of nitrate
• N = nitrate concentration
• m_T1 = slope coefficient of temperature
• T₁ = temperature
• m_T2 = slope coefficient of turbidity
• T₂ = turbidity

The raw data was cleaned in Excel by removing metadata and blank spaces from missing data and clipping the data range from June 3, 2020, to May 19, 2023. Chlorophyll, nitrate, temperature, and turbidity were normalized using a natural log transformation. Linear regressions on water quality parameters were performed on daily, weekly, and monthly averages of the dataset, respectively. Resulting slope coefficients, p-values, and R-squared values were evaluated to determine the relative strength of each predictor and confidence in the observed relationships. The slope coefficient represents the change in the mean of the outcome variables per unit change in the predictor variable, taking into account the effect of the other predictor variables. Alternatively, the p-value was used to assess the level of confidence in our results:

Confidence Level = 1 - p_value

The R-squared value represents the multiple correlation between chlorophyll and the three predictor variables.

Results

We analyzed 87,904 measurements for each water quality parameter using multivariate linear regression analysis. We found a positive relationship between each predictor and chlorophyll, and the predictors in terms of highest to lowest relative strength are as follows: temperature, turbidity, and nitrate. Temperature and turbidity had substantially higher coefficients than nitrate, ranging from a difference of 0.260 - 0.390 for temperature and 0.189 - 0.264 for turbidity across time scales (Table 2). The monthly data had the largest slope coefficients; however, the monthly data also exhibited high p-values. Reducing the granularity of the time scale of the data resulted in stronger slope coefficients; however, this was coupled with an increase in p-values. Weekly and monthly averages exhibited the highest p-values. Monthly averages have the highest p-values for temperature and turbidity, the predictors that appear to have the greatest influence on chlorophyll (Table 2).

Aggregating the contributions of the predictors as explanatory variables of chlorophyll, we found that the monthly averages of our predictors explained the largest proportion of the variance in the monthly average of chlorophyll, compared to the relationship between daily and weekly averages. The R-squared value for monthly is 0.130, which is 0.0592 and 0.0555 higher than daily and monthly, respectively, suggesting that the relationship between chlorophyll and nitrate, temperature, and turbidity is best characterized at a monthly time-scale in our data. We observed 97.4% and 99.5% confidence levels for temperature and turbidity at the weekly timescale, respectively, both of which exceed the 95% critical threshold for statistical significance. In comparison, the values dropped to 89.9% and 91.3% at the monthly scale. We nevertheless found the predictors to best fit the data at the monthly timescale. Plotting the monthly time-series of the normalized data for each variable reveals that temperature and turbidity generally fluctuate similarly to chlorophyll. Additionally, there appears to be a staggered relationship between the peaks of nitrate and chlorophyll on a monthly basis (Figure 2). Qualitatively, the time-series shows a delayed response in chlorophyll increase, following a peak in nitrate concentration. While no statistically significant correlation was observed between these two variables, there may be a relationship between chlorophyll and nitrate in the Van Sickle Island system, in addition to observed temperature and turbidity linkages.

Overall, we observed an increasing relationship between our predictors and chlorophyll at longer timescales, although this was likely due to lower sample sizes. Although our coefficients increased, our confidence levels decreased, indicating more uncertainty in the relationship. Additionally, comparing the measured chlorophyll levels with the predicted chlorophyll levels calculated using the multivariate linear regression equation underlying this analysis, we qualitatively see the greatest predictive capacity for chlorophyll when chlorophyll is close to its mean value (Figure 3). During spring spikes in chlorophyll, specifically in March 2021 and 2022, we see the greatest deviation between measured and predicted values, indicating the limited predictive power of the model to account for sharper increases and decreases in chlorophyll.

Discussion

Our findings suggest that temperature and turbidity are associated with the observed changes in chlorophyll, and nitrate concentration to a lesser extent. Previous studies have attributed observed chlorophyll variance to water temperature [27], [28] due to the influence of ambient water temperature on algae photosynthetic rates [29]. Water temperature is also driven by abiotic factors, such as season, daylight hours, and outflow, which could exert a stronger effect on phytoplankton. Subsequently, stronger positive correlation with chlorophyll concentration indicates that algal growth in this estuarine system may be largely affected by turbidity, rather than temperature. The positive relationship between turbidity and chlorophyll is counterintuitive because we may expect algal productivity to decrease as turbidity increases due to a reduction in light availability in the waterbody, limiting algal photosynthesis and enabling a greater input of nutrients without eutrophication [30], [31], [32], [33]. Nevertheless, [34] found turbidity to be the most important variable in modeling chlorophyll-a concentration based on a strong positive relationship. We can interpret the observed pattern as concurrent increases in chlorophyll and turbidity due to increased algal biomass in the waterbody [35].

Previous studies on San Francisco Bay water quality and Suisun Bay in recent decades have collectively charted a decline in primary production relative to nutrient levels since the 1980s, proposing clam grazing, light limitation, and ammonium (NH₄) input as potential explanations for this trend [20], [36], [37], [38]. NH₄ is a more energetically efficient source of nitrogen for phytoplankton compared to nitrate, contributing to greater primary production rates and the export of nitrate from the Sacramento-San Joaquin Delta through San Francisco Bay and out to the ocean [39], [40]. In Suisun Bay, the anthropogenic nutrient input from agriculture and urban runoff may drive the low chlorophyll concentrations we observed because nutrient loading is mainly in the form of NH₄, and high NH₄ concentration can restrict phytoplankton from using nitrate (NO₃^-) [41], [42]. The repression of nitrate uptake by NH₄ could explain why we observed a weak correlation between nitrate and chlorophyll for Van Sickle Island. However, the NH₄ inhibition hypothesis was mainly tested downstream of the wastewater treatment plant in Sacramento and thus may not hold true in Suisun Bay [43]. Unlike for nitrate, the USGS does not have a continuous NH₄ time series at the Van Sickle Island site, making it difficult to investigate the NH₄ and nitrate relationship. While previous research and our analysis have not fingerprinted eutrophication occurrence within Suisun Bay, NO₃^- transport to the ocean may result in coastal algal blooms [44].

Notably, there are linkages between temperature, turbidity, and nitrate that are mediated by variables, such as salinity and dissolved oxygen, that were not considered in this study and may be responsible for the observed water quality conditions [28]. Our model explored correlations between chlorophyll and potential algal growth predictors, but did not establish causation for the observed chlorophyll trends. Additionally, we could not accurately predict the spring spikes in chlorophyll concentration through linear regression, which could be due to additional abiotic influences or potential biotic interactions outside the scope of our analysis. Importantly, when phytoplankton density is high, there is more nitrogen uptake and a corresponding decline in overall nitrogen concentration, indicating a potential time lag between chlorophyll and nitrate that was not considered. To address these limitations, future studies could analyze the delayed reaction in the algal population response to nitrate, temperature, and turbidity to further synthesize the processes facilitating the observed chlorophyll-a concentration.

Van Sickle Island falls within expected natural trends for a freshwater body in terms of nitrate concentrations and chlorophyll-a content [45]. The observed variability in our data is also standard for an estuary system, as nitrate and chlorophyll values fluctuate due to river water discharge rates or the regular cycle of nutrient digestion and microbial death [46]. In developing our regression model, we made several assumptions about algal populations and the water quality parameters collected at Van Sickle Island. The most critical assumption is chlorophyll as a proxy for algal concentration. While algal concentration and chlorophyll concentration are positively correlated [46], chlorophyll-a concentrations do not account for other significant nitrate-limited markers of algal growth, such as chlorophyll-b and phycocyanin, which are indicators of green algae and cyanobacteria concentrations respectively [47]. We assumed chlorophyll-a to be sufficient because Suisun Bay is dominated by predominantly green algae species [40]. This study also assumes that Van Sickle Island represents all water inputs into Suisun Bay. However, the catchment surrounding the bay is more extensive, and inputs can occur downstream of the Van Sickle Island monitoring station (Figure 1). Finally, our analysis only considered nitrate, temperature, and turbidity when there are innumerable amounts of factors that can impact algal growth and facilitate variability in chlorophyll-a.

Overall, our results align with previous research on the predictors of algae growth in estuarine systems. Currently, Suisun Bay is a healthy water body according to Environmental Protection Agency (EPA) guidelines. The EPA has a nitrate drinking water standard of 10 mg/L, though concentrations as low as 0.3 mg/L can facilitate algal blooms [48]. Chlorophyll-a criteria exist for lakes and reservoirs, while, unlike total nitrogen, there is no direct regulation of chlorophyll-a in estuaries and rivers. There are currently no requirements for intervention in preventing algal blooms for the Van Sickle Island site. However, continued and increased sustainable nutrient management in the greater Sacramento-San Joaquin Delta can limit potential downstream impacts of nutrient loading and allow the Delta to continue to provide suitable habitat and resources for wildlife and ecosystem services for humans.

Conclusion

Our study suggests that temperature and turbidity are the best predictors of chlorophyll concentration at Van Sickle Island. Due to the reduction in sample size at the weekly and monthly timescales, we saw higher p-values highlighting the uncertainty in the observed relationships. Nitrate was not a strong predictor of chlorophyll, potentially due to a delay between nitrate concentration and chlorophyll response and anthropogenic inputs of NH₄ into the water body. We determined that Van Sickle Island is experiencing baseline levels of chlorophyll-a, indicating eutrophication is not occurring. Although Van Sickle Island has not exhibited strong spring or summertime algal blooms since 2020, there is a defined seasonal pattern in algae growth. Similar water bodies in California influenced by agricultural and urban runoff may be experiencing strong seasonal algal blooms from nutrient loading. Therefore, site-specific water quality assessments can be applied in other locations in Suisun Bay and other freshwater bodies to elucidate what parameters are facilitating algal blooms. These analyses can be used to characterize the nature of the algal blooms in terms of the algal species involved, and whether they may be toxic to the aquatic ecosystem or pose human health risks. This project contributes to the broader picture of characterizing existing water quality conditions in estuary ecosystems and diagnosing potential environmental problems. This study adds to the scientific understanding of the drivers of eutrophication in estuary systems and highlights the need to understand the relative strength of algal density predictors on a per-site basis. Future research on water quality issues can allow for targeted mitigation to limit potential ecosystem and health impacts as these issues intensify due to climate change.

 

About the Author: Benjamin Narwold

Benjamin Narwold is a recent graduate of UC Davis with a B.S. in Environmental Science and Management. He is a member of the Global Ecology and Sustainability Lab, directed by Professor Rebecca Hernandez, where he has researched the water quality impacts of floating photovoltaics on freshwater bodies since 2022. Since September 2023, Ben has worked on Kauai as an Associate Scientist for Global Algae: researching algae for food and biofuel. Aside from floating solar and algae, Ben enjoys ultimate frisbee, basketball, and surfing.

Supporting Information

References

1. Harper D. 1992. What is eutrophication? In: Eutrophication of Freshwaters: Principles, problems [Internet]. Springer Dordrecht. (1):1–28. doi:10.1007/978-94-011-3082-0_1

2. Brans KI, Engelen JMT, Souffreau C, De Meester L. 2018. Urban hot-tubs: Local urbanization has profound effects on average and extreme temperatures in ponds. Landsc Urban Plan. 176:22–9.

3. Sutton MA, Bleeker A, Howard CM, Erisman JW, Abrol YP, Bekunda M, et al. 2013. Our nutrient world. The challenge to produce more food & energy with less pollution. Edinburgh: Centre for Ecology & Hydrology. Available from: library.wur.nl/WebQuery/wurpubs/434951

4. Ho JC, Michalak AM, Pahlevan N. 2019. Widespread global increase in intense lake phytoplankton blooms since the 1980s. Nature [Internet]. 574(7780):667–70. doi:10.1038/s41586-019-1648-7

5. Dai Y, Yang S, Zhao D, Hu C, Xu W, Anderson DM, et al. 2023. Coastal phytoplankton blooms expand and intensify in the 21st century. Nature [Internet]. 615(7951):280–4. doi:10.1038/s41586-023-05760-y

6. Round FE. 1981. The Ecology of Algae. CUP Archive.

7. Reynolds CS, Walsby AE. 1975. Water-Blooms. Biol Rev. 50(4):437–81.

8. Wetzel RG, Likens G. 2000. Limnological Analyses. Springer Science & Business Media.

9. Grattan LM, Holobaugh S, Morris JG. 2016. Harmful algal blooms and public health. Harmful Algae. 57:2–8.

10. Breitburg D, Levin LA, Oschlies A, Grégoire M, Chavez FP, Conley DJ, et al. 2018. Declining oxygen in the global ocean and coastal waters. Science. 359(6371). doi:10.1126/science.aam7240

11. Diaz RJ, Rosenberg R. 2008. Spreading Dead Zones and Consequences for Marine Ecosystems. Science. 321(5891):926–9.

12. Watson EB, Rahman FI, Woolfolk A, Meyer R, Maher N, Wigand C, et al. 2022. High nutrient loads amplify carbon cycling across California and New York coastal wetlands but with ambiguous effects on marsh integrity and sustainability. PLOS ONE. 17(9):e0273260.

13. Jassby AD. 2005. Phytoplankton Regulation in a Eutrophic Tidal River (San Joaquin River, California). San Franc Estuary Watershed Sci [Internet]. Available from: escholarship.org/uc/item/9jb2t96d

14. Kudela RM, Lane JQ, Cochlan WP. 2008. The potential role of anthropogenically derived nitrogen in the growth of harmful algae in California, USA. Harmful Algae. 8(1):103–10.

15. Handler NB, Paytan A, Higgins CP, Luthy RG, Boehm AB. 2006. Human development is linked to multiple water body impairments along the California coast. Estuaries Coasts. 29(5):860–70.

16. Grimes PM. 1996. Urbanization and Water in the Northern San Joaquin Valley [dissertation]. University of California, Davis.

17. Taberski K, Dougdale R, Parker A, Marchi A. 2010. Spring Phytoplankton Bloom Progression in Suisun Bay. San Francisco Bay Region Workplan Monitoring.

18. Cloern JE, Schraga TS, Nejad E, Martin C. 2020. Nutrient Status of San Francisco Bay and Its Management Implications. Estuaries Coasts. 43(6):1299–317.

19. Wilkerson FP, Dugdale RC, Hogue VE, Marchi A. 2006. Phytoplankton blooms and nitrogen productivity in San Francisco Bay. Estuaries Coasts. 29(3):401–16.

20. Dugdale RC, Wilkerson FP, Hogue VE, Marchi A. 2007. The role of ammonium and nitrate in spring bloom development in San Francisco Bay. Estuar Coast Shelf Sci. 73(1):17–29.

21. USGS Water Data. Suisun Bay a Van Sickle Island NR Pittsburg CA [Internet]. 2020. Available from: https://waterdata.usgs.gov/monitoring-location/11455508/

22. Krause G, Weis E. 1991. Chlorophyll fluorescence and photosynthesis: the basics. Annu Rev Plant Biol. 42(1):313–49.

23. Guo H, Yao J, Sun Z, Duan D. 2015. Effects of salinity and nutrients on the growth and chlorophyll fluorescence of Caulerpa lentillifera. Chin J Oceanol Limnol. 33(2):410–8.

24. Pozzobon V, Cui N, Moreaud A, Michiels E, Levasseur W. 2021. Nitrate and nitrite as mixed source of nitrogen for Chlorella vulgaris: Growth, nitrogen uptake and pigment contents. Bioresour Technol. 330:124995.

25. Raven JA, Geider RJ. 1988. Temperature and algal growth. New Phytol. 110(4):441–61.

26. Jewson DH, Taylor JA. 1988. The influence of turbidity on net phytoplankton photosynthesis in some Irish lakes. Freshw Biol. 8(6):573–84.

27. Denman KL. 1976. Covariability of chlorophyll and temperature in the sea. Deep Sea Res Oceanogr Abstr. 23(6):539–50.

28. Apple JK, Smith EM, Boyd TJ. 2008. Temperature, Salinity, Nutrients, and the Covariation of Bacterial Production and Chlorophyll-a in Estuarine Ecosystems. J Coast Res. (10055):59–75.

29. Singh SP, Singh P. 2015. Effect of temperature and light on the growth of algae species: A review. Renew Sustain Energy Rev. 50:431–44.

30. Irigoien X, Castel J. 1997. Light Limitation and Distribution of Chlorophyll Pigments in a Highly Turbid Estuary: the Gironde (SW France). Estuar Coast Shelf Sci. 44(4):507–17.

31. Boyd CE. 2015. Particulate Matter, Color, Turbidity, and Light. In: Water Quality: An Introduction [Internet]. Cham: Springer International Publishing. doi:10.1007/978-3-319-17446-4_5

32. Tilzer MM. 1988. Secchi disk — chlorophyll relationships in a lake with highly variable phytoplankton biomass. Hydrobiologia. 162(2):163–71.

33. Bilotta GS, Brazier RE. 2008. Understanding the influence of suspended solids on water quality and aquatic biota. Water Res. 42(12):2849–61.

34. Hollister JW, Milstead WB, Kreakie BJ. 2016. Modeling lake trophic state: a random forest approach. Ecosphere. 7(3):e01321.

35. Hansson LA. 1992. Factors regulating periphytic algal biomass. Limnol Oceanogr. 37(2):322–8.

36. Ball MD, Arthur JF. 1979. Planktonic chlorophyll dynamics in the northern San Francisco Bay and Delta. San Franc Bay Urban Estuary Pac Div Am Assoc Adv Sci San Franc Calif. 265–85.

37. Cloern JE. 1987. Turbidity as a control on phytoplankton biomass and productivity in estuaries. Cont Shelf Res. 7(11):1367–81.

38. Alpine AE, Cloern JE. 1992. Trophic interactions and direct physical effects control phytoplankton biomass and production in an estuary. Limnol Oceanogr. 37(5):946–55.

39. Parker AE, Hogue VE, Wilkerson FP, Dugdale RC. 2012. The effect of inorganic nitrogen speciation on primary production in the San Francisco Estuary. Estuar Coast Shelf Sci. 104–105:91–101.

40. Wang Z, Chai F, Dugdale R, Liu Q, Xue H, Wilkerson F, et al. 2020. The interannual variabilities of chlorophyll and nutrients in San Francisco Bay: a modeling study. Ocean Dyn. 70(8):1169–86.

41. Dortch Q. 1990. The interaction between ammonium and nitrate uptake in phytoplankton. Mar Ecol Prog Ser Oldendorf. 61(1):183–201.

42. Novick E, Senn D. 2014. External nutrient loads to San Francisco Bay. San Franc Estuary Inst Richmond CA USA.

43. Cloern JE. 2021. Use Care When Interpreting Correlations: The Ammonium Example in the San Francisco Estuary. San Franc Estuary Watershed Sci [Internet]. 19(4). Available from: escholarship.org/uc/item/1xz922jm

44. Kamykowski D, Zentara SJ. 2005. Changes in world ocean nitrate availability through the 20th century. Deep Sea Res Part Oceanogr Res Pap. 52(9):1719–44.

45. Atique U, An KG. 2019. Reservoir Water Quality Assessment Based on Chemical Parameters and the Chlorophyll Dynamics in Relation to Nutrient Regime. Pol J Environ Stud. 28(3):1043–61.

46. Gameiro C, Cartaxana P, Cabrita MT, Brotas V. 2004. Variability in Chlorophyll and Phytoplankton Composition in an Estuarine System. Hydrobiologia. 525(1):113–24.

47. Dahm C, Parker A, Adelson A, Christman M, Bergamaschi B. 2016. Nutrient Dynamics of the Delta: Effects on Primary Producers. San Franc Estuary Watershed Sci. 14.

48. Shaw BH, Mechenich C, Klessig LL. 2004. Understanding lake data. University of Wisconsin–Extension, Cooperative Extension.