Summer phytoplankton assemblages and their environmental correlates in the Southern California Bight

Abstract
Weekly observations of chemical and physical variables, and of phytoplankton abundances, were made over a 21-wk period at 3 stations in the Southern California Bight. Principal component analysis was employed to resolve 4 phytoplankton assemblages among the 25 taxa with the highest variances for their log transformed abundances. Two of these components were described by the abundance of taxa characteristic of upwelling assemblages.sbd.both dominated by diatoms. A 3rd assemblage was dominated by the red tide dinoflagellate, Gonyaulax polyedra, along with some small diatoms and a coccolithophorid. Canonical correlation analysis of the 4 phytoplankton components against 13 environmental variables revealed distinct sets of temperature-salinity-nutrient conditions associated with periods of abundance of each assemblage. Two were interpretable, on physical grounds, as upwelling situations, and these were associated with periods of abundance of the phytoplankton assemblages that were identified, on floristic grounds, as upwelling assemblages. The red tide assemblage was associated with nonupwelling conditions. The pattern of correlations of 89 other taxa, besides the 25 employed in the principal component analysis, with the 4 principal components and the 4 environmental canonical variates were also consistent with this interpretation. Comparison with phytoplankton assemblages described in other studies reveals substantial consistency in broad outline, but many differences in detail, especially with respect to the presence and absence of species. The species assemblages defined by the principal components analysis exhibited episodes of abundance of a duration of 2-3 wk at a given location. Current meter records, from nearby stations, but from another year, suggest that a persistence time of 2-3 wk at a stationary site corresponds to a spatial patch scale of 20-40 km. These same current meter records show approximately 50% coherence in low-frequency currents at a separation of 25 km, indicating a possible common scale for spatial coherence of currents and spatial extent of phytoplankton blooms in this system. Implications of the analysis are discussed in terms of hypotheses concerning the structure and dynamics of plankton communities.