Search

ABSTRACT Prochlorococcus and Synechococcus are the two dominant picocyanobacteria in the low-nutrient surface waters of the subtropical ocean, but the basis for their coexistence has not been quantitatively demonstrated. Here, we combine in situ microcosm experiments and an ecological model to show that this coexistence can be sustained by specialization in the uptake of distinct nitrogen (N) substrates at low-level concentrations that prevail in subtropical environments. In field incubations, the response of both Prochlorococcus and Synechococcus to nanomolar N amendments demonstrates N limitation of growth in both populations. However, Prochlorococcus showed a higher affinity to ammonium, whereas Synechococcus was more adapted to nitrate uptake. A simple ecological model demonstrates that the differential nutrient preference inferred from field experiments with these genera may sustain their coexistence. It also predicts that as the supply of NO3− decreases, as expected under climate warming, the dominant genera should undergo a nonlinear shift from Synechococcus to Prochlorococcus, a pattern that is supported by subtropical field observations. Our study suggests that the evolution of differential nutrient affinities is an important mechanism for sustaining the coexistence of genera and that climate change is likely to shift the relative abundance of the dominant plankton genera in the largest biomes in the ocean. IMPORTANCE Our manuscript addresses the following fundamental question in microbial ecology: how do different plankton using the same essential nutrients coexist? Prochlorococcus and Synechococcus are the two dominant picocyanobacteria in the low-nutrient surface waters of the subtropical ocean, which support a significant amount of marine primary production. The geographical distributions of these two organisms are largely overlapping, but the basis for their coexistence in these biomes remains unclear. In this study, we combined in situ microcosm experiments and an ecosystem model to show that the coexistence of these two organisms can arise from specialization in the uptake of distinct nitrogen substrates; Prochlorococcus prefers ammonium, whereas Synechococcus prefers nitrate when these nutrients exist at low concentrations. Our framework can be used for simulating and predicting the coexistence in the future ocean and may provide hints toward understanding other similar types of coexistence.

Crocosphaera and Cyanothece are both unicellular, nitrogen-fixing cyanobacteria that prefer different environments. Whereas Crocosphaera mainly lives in nutrient-deplete, open oceans, Cyanothece is more common in coastal, nutrient-rich regions. Despite their physiological similarities, the factors separating their niches remain elusive. Here we performed physiological experiments on clone cultures and expand upon a simple ecological model to show that their different niches can be sufficiently explained by the observed differences in their photosynthetic capacities and rates of carbon (C) consumption. Our experiments revealed that Cyanothece has overall higher photosynthesis and respiration rates than Crocosphaera. A simple growth model of these microorganisms suggests that C storage and consumption are previously under-appreciated factors when evaluating the occupation of niches by different marine nitrogen fixers.

Stable isotope probing (SIP) combined with nano-scale secondary ion mass spectrometry (nanoSIMS) is a powerful approach to quantify assimilation rates of elements such as C and N into individual microbial cells. Here, we use mathematical modeling to investigate how the derived rate estimates depend on the model used to describe substrate assimilation by a cell during a SIP incubation. We show that the most commonly used model, which is based on the simplifying assumptions of linearly increasing biomass of individual cells over time and no cell division, can yield underestimated assimilation rates when compared to rates derived from a model that accounts for cell division. This difference occurs because the isotopic labeling of a dividing cell increases more rapidly over time compared to a non-dividing cell and becomes more pronounced as the labeling increases above a threshold value that depends on the cell cycle stage of the measured cell. Based on the modeling results, we present formulae for estimating assimilation rates in cells and discuss their underlying assumptions, conditions of applicability, and implications for the interpretation of intercellular variability in assimilation rates derived from nanoSIMS data, including the impacts of storage inclusion metabolism. We offer the formulae as a Matlab script to facilitate rapid data evaluation by nanoSIMS users.