Long-Term Changes in Nutrient Loading and Stoichiometry and Their Relationships with Changes in the Food Web and Dominant Pelagic Fish Species in the San Francisco Estuary, California

Abstract

Nutrient enrichment is an important stressor in coastal ecosystems. This analysis tests the hypothesis that changes in nutrient loads, imbalances in nitrogen:phosphorus, and changes in nitrogen form, especially shifts to increasing loads of chemically reduced, rather than oxidized nitrogen, can have major impacts on food webs, from primary producers through secondary producers to fish. The application of cumulative sums of variability, the running total of deviations from normalized values over time, is a sensitive method for comparing rates of change between and among all parameters, including organisms of all trophic levels. This approach was applied to the San Francisco Estuary, California, demonstrating that abrupt changes in nutrient loads and nutrient form over the past several decades were correlated with food web changes, including pelagic fish collapse. Remediation of pelagic fish populations should be centered on reduction of nitrogen loads and reestablishment of balanced nutrient ratios delivered from point source discharges.

Keywords:

• ammonium
• CUSUM charts
• delta smelt
• eutrophication
• nutrient ratios
• nutrient stoichiometry
• pelagic organism decline
• plankton trophodynamics

ACKNOWLEDGMENTS

Many of the ideas for this analysis evolved from discussions with SCOR/LOICZ Working Group 132 on Land-Based Nutrient Pollution and Harmful Algal Blooms and the GEOHAB Core Research Project on HABs and Eutrophication. I sincerely thank H. Briceño, T. Kana, D. Hinkle, J. Alexander, C. Solomon, R. Dugdale, J. Burkholder, E. Van Nieuwenhuyse, F. Brewster, B. J. Miller, T. Mongan, D. Fullerton, R. Sitts, W. Kimmerer, and A. Müller-Solger for their assistance with, helpful discussions about, or input into, various aspects of the data presented or this analysis. Three anonymous reviewers provided very helpful critiques of an earlier version of this manuscript. This is contribution number x4414 from the University of Maryland Center for Environmental Science. Support for this work was provided by the State Water Contractors and San Luis and Delta-Mendota Water Authority and NSF grant MCB-0818276 via subcontract through Gallaudet University.