Macro- to fine-scale spatial and temporal distributions and dynamics of phytoplankton and their environmental driving forces in a small montane lake in southern California, USA

A wireless network of buoys, two autonomous robotic boats, and an autonomous tethered vertical profiling system were used to characterize phytoplankton dynamics and spatiotemporal changes in chemical and physical forcing factors in a small montane lake (Lake Fulmor, Idyllwild, California). Three deployments each year were conducted in 2005 and 2006 to examine seasonal changes in the structure of the lake and phytoplankton assemblage, as well as fine-scale temporal and spatial variations. The buoys yielded fine-scale temporal patterns of in situ fluorescence and temperature, while the vertical profiling system yielded two-dimensional, cross-sectional profiles of several parameters. The autonomous vehicles provided information on fluorescence and corresponding temperature patterns across the surface of the lake. Average, lake-wide chlorophyll concentrations increased 10fold seasonally, and strong anoxia developed in the hypolimnion during the summer. The latter process dramatically affected vertical chemical gradients in the 5 m water column of the lake. Small-scale spatial (,1 m) and temporal (minutes) heterogeneity in fluorescence were surprisingly large. These variations were due predominantly to vertical mixing of the phytoplankton assemblage and to phytoplankton vertical migratory behavior. Large peaks in fluorescence at 0.5-m occurred at very short time intervals (minutes) during all deployments, and appeared to be due to upward mixing of deeper dwelling eukaryotic phytoplankton during early–mid-summer, or downward mixing of surface-associated cyanobacteria during late summer. The development, application, and continued sophistication of environmental sensor platforms in ocean science have become common themes in research programs investigating water quality and large-scale oceanographic processes (Dickey 2003; Glasgow et al. 2004; Dickey and Bidigare 2005). These sensor platforms have provided valuable environmental data for modeling biogeochemical cycles of important chemical constituents, particularly over long temporal or broad geographical scales (Dickey 2004). More recently, emphasis has been directed towards the application of embedded networks of sensors that can obtain and collate information from several locations to provide a detailed characterization of the ecosystem under study (Szewczyk et al. 2002; Pottie and Kaiser 2005; Fries