Temporal variation and spatial heterogeneity of phytoplankton abundance within a water supply impoundment

The spatial heterogeneity and temporal variability of phytoplankton biomass and floristic composition was analyzed in Croton Reservoir, New York from 1984-1988 using a stepwise regression model of seven variables. The reservoir exhibited a longitudinal gradient of algal biomass, physical factors, and chemical factors from headwater to dam during the entire study period. Hydraulic residence time and reservoir volume levels played an important role in influencing the abundance and the composition of the phytoplankton, as well as affecting the extent of the physical and chemical longitudinal gradients. Although the reservoir's bathymetry and water flow superficially resembled a general model for reservoirs, analysis of the flow regimes showed that the impoundment did not have the pronounced riverine zone and clearly defined transitional zone predicted by the model. Findings suggested that, with the exception of the growing season (July-October) period, reservoir flow through rates were generally too rapid to allow for the accumulation of algal biomass. The reservoir functioned as a rapidly flushed nutrient-rich system which occasionally exhibited longitudinal algal biomass gradients due to the combined sequential effect of increased growing season residence time coupled with intense fall precipitation. The impoundment exhibited the same classic pattern of phytoplankton succession as that observed in northern temperate mesotrophic-eutrophic lakes. However, this pattern of succession did not vary spatially within the reservoir. Nutrient availability never appeared to limit algal abundance, although the relative proportions of nutrients measured corresponded well with the spatial and temporal patterns of floristic composition observed within the impoundment, and were consistent with those described by current nutrient ratio-phytoplankton competition theory. It was observed that growing season cell losses and reduced biomass may have been attributable to grazing, light-limitation, and cell sedimentation. Grazing losses appeared to be more pronounced in the reservoir's lower zone, whereas nonalgal light-limitation, and sedimentation were more pronounced in the reservoir's upper zone. This investigation utilized one of the most complete long-term data sets available for a reservoir ecosystem and its findings will be used to formulate improved water quality management strategies.