Leveraging high-frequency measurements to predict seasonal cyanobacterial blooms in an oligotrophic lake

Background/Question/Methods

Cyanobacteria have become more abundant in freshwater lakes worldwide, with adverse consequences for ecosystem function. These increases are widely expected to continue into the future, due to the ability of pelagic cyanobacteria to outcompete eukaryotic phytoplankton in warmer temperatures and strongly stratified conditions.  However, cyanobacterial blooms are not driven solely by division in the water column: pelagic cyanobacteria can also increase via recruitment from dormant stages in the benthos.  In fact, recruitment from the benthos can provide a critically important inoculum for pelagic populations, including blooms.

     This talk will describe our ongoing efforts to understand and predict recruitment and pelagic population dynamics of the large colonial cyanobacterium Gloeotrichia echinulata in Lake Sunapee, New Hampshire, USA.  Since 2005, we have quantified surface abundances and recruitment from littoral sediments at multiple sampling sites at weekly intervals from May through September.  We have also sampled surface abundances at daily intervals at one near-shore site since 2008.  Currently, we are building models that relate these Gloeotrichia measurements to high-frequency measurements made by littoral HOBO temperature-light loggers and the lake’s pelagic Global Lake Ecological Observatory (GLEON) buoy.

Results/Conclusions

We found extensive spatial and temporal variability in both Gloeotrichia recruitment and pelagic abundances across the 10-year dataset.  Recruitment contributed strongly to the pelagic population in some years, but not others.  Analyses linking recruitment and pelagic abundance to the high frequency temperature and light measurements suggest several interesting findings.  For example, recruitment rates within any given summer tend to be best predicted by multiplicative functions of light and temperature, while among-year variation in recruitment appears to be more strongly related to lake mixing, as defined by deeper thermoclines, lower Schmidt stability, and greater diurnal variability in water temperature.  Additionally, population growth rates calculated from the daily Gloeotrichia data increase at higher lake temperatures, but growth was also strongly density-dependent at densities >1 colony/L, indicating that Gloeotrichia blooms may be self-limiting in ways not reported for other cyanobacteria.  Next steps include bridging the daily and weekly data, and the recruitment and pelagic abundance data, to develop comprehensive population models.  Taken together, our work demonstrates that integrating high-frequency sensor data with manual field collections over long time periods can provide valuable insights into cyanobacterial population dynamics.