Anderson, CR, Siegel, DA, Kudela, RM, Brzezinski, MA (2009). Empirical models of toxigenic Pseudo-nitzschia blooms: Potential use as a remote detection tool in the Santa Barbara Channel. HARMFUL ALGAE, 8(3), 478-492.
The Santa Barbara Channel, CA is a highly productive region where wind-driven upwelling and mesoscale eddies are important processes driving phytoplankton blooms. In recent years, the spring bloom has been dominated by the neurotoxin-producing diatom, Pseudo-nitzschia spp. In this paper, we relate a 1.5-year time series of Pseudo-nitzschia spp. abundance and domoic acid concentration to physical, chemical, and biological data to better understand the mechanisms controlling local Pseudo-nitzschia spp. bloom dynamics. The data were used to define the ranges of environmental conditions associated with Pseudo-nitzschia spp. bloom development in the Santa Barbara Channel. The time series captured three large toxic events (max. particulate domoic acid concentration, pDA similar to 6000 ng L-1; max. cellular domoic acid concentrations, cDA similar to 88 pg cell(-1)) in the springs of 2005-2006 and summer 2005 corresponding to bloom-level Pseudo-nitzschia spp. abundance (>5.0 x 10(4) cells L-1). In general, large increases in Pseudo-nitzschia spp. abundance were accompanied by increases in cDA levels, and cDA peaks preceded pDA peaks by at least one month in both the springs of 2005 and 2006. Statistical models incorporating satellite ocean color (MODIS-Aqua and SeaWiFS) and sea surface temperature (AVHRR) data were created to determine the probability that a remotely sensed phytoplankton bloom contains a significant population of toxic Pseudo-nitzschia spp. Models correctly estimate 98% of toxic bloom situations, with a 7-29% rate of false positive identification. Conditions most associated with high cDA levels are low sea surface temperature, high salinity, increased absorption by cDOM (412 nm), increased reflectance at 510/555 nm, and decreased particulate absorption at 510 nm. Future efforts to merge satellite and regionally downscaled forecasting products with these habitat models will help assess bloom forecasting capabilities in the central CA region and any potential connections to large-scale climate modes. (C) 2008 Elsevier B.V. All rights reserved.