Long-term temporal and spatial trends in phytoplankton biomass and class-level taxonomic composition in the hydrologically variable Neuse–Pamlico estuarine continuum, North Carolina, U.S.A

Phytoplankton diagnostic photopigments in near-surface waters (<0.5 m) were identified and quantified by high-performance liquid chromatography beginning in April 1994 in the Neuse River Estuary and in October 1999 in the Pamlico Sound, North Carolina. Photopigment concentrations were analyzed using ChemTax to determine the class-specific biomass of the dominant phytoplankton groups. Long-term annual and seasonal trends in phytoplankton biomass and composition were characterized along the river–estuarine continuum and compared to river flow rates, which were variable because of droughts, uncharacteristic seasonal rainfall patterns, and, since 1996, an increase in the frequency of tropical storms and hurricanes. We tested the hypothesis that temporal and spatial patterns of phytoplankton biomass and composition were largely controlled by changes in river flow rate through associated changes in salinity and residence time or through physical transport and advection of phytoplankton classes with river flow along the estuary. Significant interannual, seasonal, and spatial variability in phytoplankton biomass and composition was observed and coincided with variability in river flow rates. The five dominant phytoplankton classes (Chlorophyceae, Cryptophyceae, Cyanobacteria, Bacillariophyceae, and Dinophyceae) were physically displaced downstream during periods of elevated river flow. Dinoflagellates were reduced in abundance during high flow conditions, especially in the upper Neuse River Estuary. The abundance of cyanobacteria was also reduced throughout the system during elevated river flow conditions, although chlorophytes were more abundant. Changes in hydrology can be a useful indicator of seasonal phytoplankton distribution and higher level compositional changes along hydrologic gradients in these and similar systems. Because of their small size, rapid nutrient uptake and growth rates, specific growth requirements, and susceptibility to grazing, phytoplankton are particularly responsive to variations in environmental condition (Reynolds 1984; Stolte et al. 1994). This sensitivity, combined with their ability to rapidly proliferate in waters affected by anthropogenic (agricultural, urban, and industrial watershed activities) and natural (droughts and elevated rainfall events) disturbances, implies that phytoplankton can be useful indicators of environmental change in freshwater, estuarine, and marine environments (Paerl 1988; Reynolds et al. 1993; Richardson 1997). Deviations from typical phytoplankton abundance and compositional patterns through time and space can thus be used to detect the occurrence of ecological change in estuaries (Malone et al. 1988; Harding 1994). In systems that have been eutrophic for many years and continue to be affected by anthropogenic and climatic perturbations, determining the normal cycles of phytoplankton abundance and composition from which to detect these deviations proves to be difficult. In addition, in many estuarine systems, few data exist on phytoplankton community structure prior to these impacts. In order to distinguish between typical and uncharacteristic temporal and spatial phytoplankton cycles and thus to be able to detect past, present, and future ecological change, long-term, continuous information on phytoplankton community structure is necessary to develop a sitespecific reference or baseline condition of phytoplankton community dynamics (Harding and Perry 1997). Previously, phytoplankton biomass and composition data were acquired by microscopic analysis of water samples. This technique proved to be time-consuming, analyst-specific, and prone to errors (Tester et al. 1995; Mackey et al. 1996). More recently, the use of highperformance liquid chromatography (HPLC) to identify and quantify concentrations of phytoplankton diagnostic photopigments (chlorophylls and carotenoids) and the subsequent analysis of these photopigment concentrations by mathematical computations (multiple linear regression or steepest-descent algorithm using ChemTax), has become 1 Corresponding author ([email protected]). 2 Department of Biological Sciences, University of South Carolina, Columbia, South Carolina 29208. Acknowledgments We appreciate the technical assistance and input of past and present laboratory technicians and students, especially Alan R. Joyner, Tom Gallo, Nathan S. Hall, Benjamin L. Peierls, Richard S. Weaver, and Pamela R. Wyrick. Research discussed in this paper was partially supported by the National Science Foundation (DEB 9815495 and OCE 9905723), the U.S. Department of Agriculture NRI Project 00-35101-9981, the U.S. EPA STAR Projects R82-5243-010, R83-0652, and R82867701, the NOAA/North Carolina Sea Grant Program R/ MER-43, and the North Carolina Department of Natural Resources and Community Development/University of North Carolina Water Resources Research Institute (Neuse River Estuary Modeling and Monitoring Project, ModMon, and the Ferry-based Water Quality Monitoring Program, FerryMon). Limnol. Oceanogr., 51(3), 2006, 1410–1420 E 2006, by the American Society of Limnology and Oceanography, Inc.