Spatial and temporal trends of diatom flux in British Columbian fjords

Publisher Website
Google Scholar
Abstract
The monthly flux of diatom taxa at four sites in two fjords of British Columbia, Canada, was determined over a 3-year period by particle interceptor traps. Saanich Inlet is a small basin with little horizontal density gradient while Jervis Inlet experiences seasonal runoff and strong vertical stratification. The basic pattern of seasonal succession is the same in both fjords, but the timing and volume of production differ. Production is spread over a longer part of the year in Saanich Inlet, resulting in a more diverse assemblage, while in Jervis Inlet there is a single production maximum in May, dominated by Skeletonema costatum. Most taxa occur simultaneously at all sites, but many show a preference for one fjord, or one area within a fjord. Several species of Chaetoceros recur consistently in inner Saanich Inlet each fall, suggesting re-seeding by benthic spores. In contrast, Rhizosolenia setigera is more common in the outer reaches of both fjords, and is probably introduced from outside. In Saanich Inlet interannual variability is relatively low, except for the fall season, which may differ from year to year in both volume and taxonomic composition. Jervis Inlet is both spatially and temporally more variable: spring production may differ from year to year by an order of magnitude. In this fjord the degree of stratification, controlled by freshwater influx, appears to determine the timing and composition of the phytoplankton assemblage. Large volumes of runoff may suppress the spring bloom, and prolonged stratification through the summer produces an unusual assemblage which may be light limited. El Niño of 1986–1987 had an opposite effect in the two fjords. In Saanich Inlet the unusual degree of sunshine during fall resulted in the largest bloom of the entire interval. Unusually warm air temperatures prevented the build-up of winter snow around Jervis Inlet. As a result, the normal late-spring stabilization of the water column from snowmelt did not occur, and the spring bloom was eliminated.