Feeding dynamics of the copepod Diacyclops thomasi before, during and following filamentous cyanobacteria blooms in a large, shallow temperate lake

Cyanobacteria blooms are an increasing problem in temperate freshwater lakes, leading to reduced water quality and in some cases harmful effects from toxic cyanobacteria species. To better understand the role of zooplankton in modulating cyanobacteria blooms, from 2008 to 2010 we measured water quality and plankton abundance, and measured feeding rates and prey selectivity of the copepod Diacyclops thomasi before, during and following summertime cyanobacteria blooms in a shallow, eutrophic lake (Vancouver Lake, Washington, USA). We used a combined field and experimental approach to specifically test the hypothesis that copepod grazing was a significant factor in establishing the timing of cyanobacteria bloom initiation and eventual decline in Vancouver Lake. There was a consistent annual succession of zooplankton taxa, with cyclopoid copepods (D. thomasi) dominant in spring, followed by small cladocerans (Eubosmina sp.). Before each cyanobacteria bloom, large cladocerans (Daphnia retrocurva, Daphnia laevis) peaked in abundance but quickly disappeared, followed by brief increases in rotifers. During the cyanobacteria blooms, D. thomasi was again dominant, with small cladocerans abundant in autumn. Before the cyanobacteria blooms, D. thomasi substantially consumed ciliates and dinoflagellates (up to 100% of prey biomass per day), which likely allowed diatoms to flourish. A shift in copepod grazing toward diatoms before the blooms may have then helped to facilitate the rapid increase in cyanobacteria. Copepod grazing impact was the highest during the cyanobacteria blooms both years, but focused on non-cyanobacteria prey; copepod grazing was minimal as the cyanobacteria blooms waned. We conclude that cyclopoid copepods may have an indirect role (via trophic cascades) in modulating cyanobacteria bloom initiation, but do not directly contribute to cyanobacteria bloom decline.