Using convective mixing in mesocosms to study climate-driven shifts in phytoplankton community distributions

With climate change predicted to alter water column stability and mixing across the world's oceans, a mesocosm experiment was designed to ascertain how a natural phytoplankton community would respond to these changes. As a departure from other mesocosm experiments, we used heating and cooling to produce four different climate-inspired mixing scenarios ranging from well-mixed water columns representative of typical open turbulence (& epsilon; = 3 x 10(-8) m(2)/s(3)) through to a quiescent water column with stable stratification (& epsilon; = 5 x 10(-10) m(2)/s(3)). This method of turbulence generation is an improvement on previous techniques (e.g., grid, shaker, and aeration) which tend to produce excessive dissipation rates inconsistent with oceanic turbulence observations. Profiles of classical physical parameters used to describe turbulence and mixing (turbulent dissipation rate, buoyancy frequency, turbulent eddy diffusivity, Ozmidov scale) were representative of the profiles found in natural waters under similar mixing conditions. Chlorophyll-a profiles and cell enumeration showed a clear biological response to the different turbulence scenarios. However, the responses of specific phytoplankton groups (diatoms and dinoflagellates) did not conform to the usual expectations: diatoms are generally expected to thrive under convective, turbulent regimes, while dinoflagellates are expected to thrive in converse conditions, i.e., in stable, stratified conditions. Our results suggest that responses to mixing regimes are taxon-specific, with no overwhelming physical effect of the turbulence regime. Rather, each taxon seemed to very quickly reach a given vertical distribution that it managed to hold, whether actively or passively, with a high degree of success. Future studies on the effects of climate change on phytoplankton vertical distribution should thus focus on the factors and mechanisms that combine to determine the specific distribution of species within taxa. Our convection-based mesocosm approach, because it uses a primary physical force that generates turbulence in open waters, should prove a valuable tool in this endeavor.