Australia’s tropical waters represent predicted ‘hotspots’ for nitrogen (N2) fixation based on empirical and modelled data. However, the identity, activity and ecology of diazotrophs within this region are virtually unknown. By coupling DNA and cDNA sequencing of nitrogenase genes (nifH) with size-fractionated N2 fixation rate measurements, we elucidated diazotroph dynamics across the shelf region of the Arafura and Timor Seas (ATS) and oceanic Coral Sea during Austral spring and winter. During spring, Trichodesmium dominated ATS assemblages, comprising 60% of nifH DNA sequences, while Candidatus Atelocyanobacterium thalassa (UCYN-A) comprised 42% in the Coral Sea. In contrast, during winter the relative abundance of heterotrophic unicellular diazotrophs (δ-proteobacteria and γ-24774A11) increased in both regions, concomitant with a marked decline in UCYN-A sequences, whereby this clade effectively disappeared in the Coral Sea. Conservative estimates of N2 fixation rates ranged from <1 to 91 nmol l−1 day−1, and size fractionation indicated that unicellular organisms dominated N2 fixation during both spring and winter, but average unicellular rates were up to 10-fold higher in winter than in spring. Relative abundances of UCYN-A1 and γ-24774A11 nifH transcripts negatively correlated to silicate and phosphate, suggesting an affinity for oligotrophy. Our results indicate that Australia’s tropical waters are indeed hotspots for N2 fixation and that regional physicochemical characteristics drive differential contributions of cyanobacterial and heterotrophic phylotypes to N2 fixation.

How the functional traits (FTs) of phytoplankton change with temperature is important for understanding the impacts of ocean warming on phytoplankton mediated biogeochemical fluxes. This study quantifies the thermal performance curves (TPCs) of FTs in the cosmopolitan model diatom, Thalassiosira pseudonana, to advance understanding of trade-offs between physiological (photoacclimation, carbon fixation, nitrate, phosphate and silicate uptake) and morphological traits (cell volume and frustule silicification). We show that each FT has substantial phenotypic plasticity and exhibits a unique TPC, varying in both shape and thermal optimum, and diverging from the growth response. The TPC for growth was symmetric with a thermal optimum (Topt) of 18 °C. In comparison, the TPC for primary productivity was warm-skewed with a Topt around 21 °C, whereas frustule silicification decreased linearly with increasing temperature. Together, this suggests that the optimal temperature for overall fitness is a balance of trade-offs in the underlying functional traits. Moreover, these results demonstrate that growth is not necessarily an accurate estimate of overall biogeochemical performance and that temperature change will likely influence elemental fluxes such as carbon and silicon. Finally, we show that temperature-driven changes in individual traits e.g. photoacclimation, can mimic responses experienced under other environmental stressors (high light) and so a multi-trait assessment is essential for accurate interpretation of the cellular impact of warming. This study also reveals that multi-trait analysis, in the context of TPCs, provides insight into the cellular physiology regulating the whole cell response and has the potential to provide better estimates of how diatom-mediated biogeochemical fluxes are likely to be impacted in the context of ocean warming. Analyzing the response of multiple traits more comprehensively over other environmental gradients may therefore provide a useful framework to advance understanding of how taxon-specific functional traits will respond to multifaceted ocean change.

Predicting spatiotemporal distributions of phytoplankton biomass and community composition heavily relies on experimental studies that document how environmental conditions influence population growth rates. In unicellular phytoplankton, the net population growth rate is the difference between the cell division rate and the death rate. Along with predation and disease, phytoplankton mortality arises from abiotic stress. Although the effect of temperature on the net population growth rate is well understood, studies examining thermally induced death rates in phytoplankton are scarce. We investigated how cell division and death rates of the diatom Phaeodactylum tricornutum varied within its thermal tolerance limits (thermal niche), and at temperatures just above its upper thermal tolerance limit. We show that death rates were largely independent of temperature when P. tricornutum was grown within its thermal niche, but increased significantly at temperatures that approached or exceeded its upper thermal tolerance limit. Furthermore, the sensitivity of mortality increased with the duration of exposure to heat stress and was affected by the pre‐acclimation temperature. Heat waves can be expected to significantly affect phytoplankton mortality episodically. The increasing frequency of heat waves accompanying global warming can be expected to drive changes in phytoplankton community structure due to interspecific variability of thermal niches with potential implications for food web dynamics and biogeochemical cycles.

The distribution of marine phytoplankton will shift alongside changes in marine environments, leading to altered species frequencies and community composition. An understanding of the response of mixed populations to abiotic changes is required to adequately predict how environmental change may affect the future composition of phytoplankton communities. This study investigated the growth and competitive ability of two marine diatoms, Phaeodactylum tricornutum and Thalassiosira pseudonana, along a temperature gradient (9–35°C) spanning the thermal niches of both species under both high‐nitrogen nutrient‐replete and low‐nitrogen nutrient‐limited conditions. Across this temperature gradient, the competitive outcome under both nutrient conditions at any assay temperature, and the critical temperature at which competitive advantage shifted from one species to the other, was well predicted by the temperature dependencies of the growth rates of the two species measured in monocultures. The temperature at which the competitive advantage switched from P. tricornutum to T. pseudonana increased from 18.8°C under replete conditions to 25.3°C under nutrient‐limited conditions. Thus, P. tricornutum was a better competitor over a wider temperature range in a low N environment. Being able to determine the competitive outcomes from physiological responses of single species to environmental changes has the potential to significantly improve the predictive power of phytoplankton spatial distribution and community composition models.