Physiological acclimation by phytoplankton explains observed changes in Si and N uptake rates during the SERIES iron-enrichment experiment

Abstract

During the SERIES iron-enrichment experiment in the eastern subarctic Pacific, after addition of iron and its subsequent depletion, the Si:N drawdown ratio increased at approximately the time that diatoms became iron limited. Laboratory studies have reported that this results from a decrease in the rate of N uptake together with a more moderate decrease in the rate of Si uptake for iron-limited cultures compared to iron-replete cultures. However, for SERIES Boyd et al. (Limnol. Oceanogr. 50 (2005)) reported an unexplained increase in the rate of Si uptake at the onset of iron limitation and suggested that studies of nutrient uptake kinetics should be undertaken in search of an explanation. We compare the classic Michealis–Menten (MM) kinetics to the recently developed optimal uptake (OU) kinetics (the SPONGE: Smith and Yamanaka. Limnol. Oceanogr. 52 (2007)) within a variable-composition model, which employs cell quotas for each relevant nutrient, applied to the multi-element (C, N, Si, Fe) dynamics during SERIES. Using the Monte Carlo Markov Chain, we fit two versions of the model (differing only in the equations for nutrient uptake) to the available data for nutrient concentrations, chlorophyll, biogenic silica and particulate organic carbon and specific growth rates. With either uptake kinetics, the model reproduces observed concentrations well for nutrients and somewhat less well for chlorophyll. The different uptake kinetics yield greater differences in modeled elemental composition of phytoplankton and biomass of phytoplankton and zooplankton, which are not directly constrained by data. MM kinetics cannot reproduce the observed increase in Si uptake rate as a function of the decreasing trend in concentration of silicic acid, and it predicts Si limitation throughout nearly all of the experiment after iron-fertilization. In contrast, OU kinetics reproduces the increase in Si uptake rate and matches the observation-based estimate for the timing of the return to iron limitation. The key assumption of the SPONGE, that uptake rates of all nutrients depend on physiological acclimation by phytoplankton as a function of the ambient concentration of the growth-limiting nutrient, was originally formulated for modeling chemostat experiments. We show that it also agrees with the observations from this field experiment and provides an explanation for the increases in Si uptake rate and Si:N drawdown ratio.