Growth of Coccolithophores Controlled by Internal Nutrient Stores in Light- and Nutrient-Limited Batch Reactors: Relevance for the BIOSOPE Deep Ecological Niche of Coccolithophores.

Perrin Laura, Sorbonne Universit├ęs (UPMC, univ. Paris 6) -CNRS-IRD-MNHN, LOCEAN-IPSL, PARIS, France, Ian Probert, Station Biologique de Roscoff, Roscoff, France, Gerald Langer, University of Cambridge, Department of Earth Sciences, Cambridge, United Kingdom and Giovanni Aloisi, LOCEAN, UMR 7159, CNRS-UPMC-IRD-MNHN, Paris, France
Coccolithophores are unicellular, calcifying marine algae that play a fundamental role in the oceanic carbon cycle. Recent research has focused on investigating the effect of ocean acidification on cellular calcification. However, the success of this important phytoplankton group in the future ocean will depend on how cellular growth reacts to changes in a combination of environmental variables. We carried out batch culture experiments in conditions of light- and nutrient- (nitrate and phosphate) limitation that reproduce the in situ conditions of a deep ecological niche of coccolithophores in the South Pacific Gyre (BIOSOPE cruise, 2004). We modelled nutrient acquisition and cellular growth in our batch experiments using a Droop internal-stores model. We show that nutrient acquisition and growth are decoupled in coccolithophores; this ability may be key in making life possible in oligotrophic conditions such as the deep BIOSOPE biological niche. Combining the results of our culture experiments with those of Langer et al. (2013), we used the model to obtain estimates of fundamental physiological parameters such as the Monod constant for nutrient uptake, the maximum growth rate and the minimum cellular nutrient quota. These parameters are characteristic of different phytoplankton groups and are needed to simulate phytoplankton growth in biogeochemical models. Our results suggest that growth of coccolithophores in the BIOSOPE deep ecological niche is light-limited rather than nutrient-limited. Our work also shows that simple batch experiments and straightforward numerical modelling are capable of providing estimates of physiological parameters usually obtained in more costly and complicated chemostat experiments.