Sommer U, Aberle N, Lengfellner K, Lewandowska A
mesocosm, Kiel indoor, Kiel, Germany, spring bloom, Light, 1.4 m3
The response of the Baltic Sea spring bloomwas studied in mesocosm experiments, where temperatureswere elevated up to 6 C above the present-day sea surfacetemperature of the spring bloom season. Four of the sevenexperiments were carried out at different light levels(32–202 Wh m-2 at the start of the experiments) in thedifferent experimental years. In one further experiment, thefactors light and temperature were crossed, and in oneexperiment, the factors density of overwintering zooplanktonand temperature were crossed. Overall, there was aslight temporal acceleration of the phytoplankton springbloom, a decline of peak biomass and a decline of mean cellsize with warming. The temperature influence on phytoplanktonbloom timing, biomass and size structure wasqualitatively highly robust across experiments. The dependenceof timing, biomass, and size structure on initialconditions was tested by multiple regression analysis of they-temperature regressions with the candidate independentvariables initial light, initial phytoplankton biomass, initialmicrozooplankton biomass, and initial mesozooplankton(=copepod) biomass. The bloom timing predicted for meantemperatures (5.28 C) depended on light. The peak biomassshowed a strong positive dependence on light and a weakernegative dependence on initial copepod density. Meanphytoplankton cell size predicted for the mean temperatureresponded positively to light and negatively to copepoddensity. The anticipated mismatch between phytoplanktonsupply and food demand by newly hatched copepod naupliioccurred only under the combination of low light and warmtemperatures. The analysis presented here confirms earlierconclusions about temperature responses that are based onsubsets of our experimental series. However, only thecomprehensive analysis across all experiments highlightsthe importance of the factor light.