A prolonged bloom of Karenia mikimotoi was observed during 2006 in Scottish waters. This bloom is thought to be unique in the region in terms of its large spatial extent. From its first detection in the west of the country, the bloom moved clockwise around the coast eventually reaching the east coast and the Shetland Isles to the north. The bloom resulted in extensive mortalities of benthic organisms including annelids and molluscs and some species of fish. Farmed fish mortalities were absent but gill damage was reported. The availability of satellite remote sensing, phytoplankton counts from multiple sites, meteorological data and some water chemistry, as well as information on the physical characteristics of the sampling sites, provided an extensive spatial and temporal data set. Analysis of remotely sensed chlorophyll-a data from Aqua-MODIS indicated that this parameter is a useful early warning indicator of K. mikimotoi in shelf waters off the Scottish west coast, and suggested that the bloom developed in this region prior to its advection to coastal waters. An earth observation (EO) based harmful bloom classifier for K. mikimotoi recognised areas of elevated K. mikimotoi cell density but generated false positives in areas of high reflectance. Data were also used to evaluate, in Scottish waters, various hypotheses that exist to explain the formation of K. mikimotoi blooms including phototaxis, nutrient availability, cell transport and elevated water temperature. Specifically, we sought to evaluate if routinely collected environmental data (water temperature, insolation, wind strength and direction, and sea-loch aspect) could be used as a predictor of bloom magnitude near aquaculture facility locations, which typically lie within fjordic sea lochs. Path analysis was used to derive intuitive models linking environmental drivers to bloom magnitude. Once the effects of latitude such as northward water cooling were taken into account, only rainfall was a significant predictor of bloom magnitude at the sampling sites. Therefore, while the offshore development and progression of a bloom may be predicted from satellite information, it is likely that local hydrodynamic influences are crucial in determining its magnitude at coastal aquaculture sites. Crown Copyright (C) 2008 Published by Elsevier B.V. All rights reserved.
