INSIGHT REVIEW NATURE Vol 437 15 September 2005 doi 10 1038 nature04158 Marine microorganisms and global nutrient cycles Kevin R Arrigo1 The way that nutrients cycle through atmospheric terrestrial oceanic and associated biotic reservoirs can constrain rates of biological production and help structure ecosystems on land and in the sea On a global scale cycling of nutrients also affects the concentration of atmospheric carbon dioxide Because of their capacity for rapid growth marine microorganisms are a major component of global nutrient cycles Understanding what controls their distributions and their diverse suite of nutrient transformations is a major challenge facing contemporary biological oceanographers What is emerging is an appreciation of the previously unknown degree of complexity within the marine microbial community To understand how carbon and nutrients such as nitrogen and phosphorus cycle through the atmosphere land and oceans we need a clearer picture of the underlying processes This is particularly important in the face of increasing anthropogenic nutrient release and climate change Marine microbes which are responsible for approximately half of the Earth s primary production play an enormous role in global nutrient cycling In this review I will highlight the four exciting aspects of marine microbial ecology that are receiving a great deal of attention and may prove to be crucial to a revised understanding of marine and global nutrient cycling The first involves the explanations for and consequences of the variable nutrient stoichiometry of phytoplankton The second is the emerging concept that phytoplankton growth can be limited by more than one resource The third concerns the upward revision of estimates of marine nitrogen fixation and the fourth is the discovery that fixed nitrogen in the ocean can be lost through anaerobic ammonium oxidation anammox reactions Although distinct these topics all represent examples of the marine microbial community modulating the coupling between the cycles of nutrients and carbon Consequently they all have the capacity to fundamentally alter our perceptions of global nutrient cycles and their response to environmental change Non Redfield behaviour of phytoplankton In the early part of the twentieth century Alfred Redfield noticed that the elemental composition of plankton was strikingly similar to that of the major dissolved nutrients in the deep ocean1 On the basis of these observations Redfield proposed that the nitrate phosphate NO3 PO4 ratio of 16 1 in the sea was controlled by the requirements of phytoplankton which subsequently release nitrogen N and phosphorus P to the environment at this ratio as they are broken down remineralized Redfield s initial observations have been confirmed numerous times and the notion of a Redfield ratio describing the stoichiometry of both phytoplankton and seawater remains a fundamental tenet shaping our understanding of marine ecology biogeochemistry and even phytoplankton evolution The Redfield ratio has been extended to include other elements most notably carbon C and it links these three major biogeochemical cycles through the activities of marine phytoplankton Unfortunately a clear mechanism explaining the observed magnitude of the Redfield C N P ratio of 106 16 1 for either phytoplankton or the deep ocean has been elusive It has long been recognized that conditions exist under which phytoplankton stoichiometry diverges from the canonical Redfield ratio Furthermore a number of processes drive oceanic nutrient inventories away from the Redfield ratio including changes in exogenous nutrient delivery2 and microbial metabolism3 for example nitrogen fixation denitrification and anammox see below These processes are sometimes manifested as variations in N a measure of the degree of N deficit or excess relative to P for a given water mass4 What governs variations in phytoplankton nutrient stoichiometry and given that variation why is the Redfield N P ratio observed in the deep ocean so universal At the most basic level the C N P stoichiometry of extant phytoplankton reflects the elemental composition retained from their early evolutionary history5 In the case of eukaryotic phytoplankton the two major superfamilies differ markedly in their cellular C P and N P ratios with the green superfamily exhibiting significantly higher ratios than the red green C P 200 and N P 27 red C P 70 and N P 10 However all observed C N P stoichiometries cannot be explained by the evolutionary lineage of an organism The highly dynamic stoichiometry often exhibited by unicellular algae reflects their ability to store nutrients in internal pools switch between enzymes with different nutrient requirements and modify osmolyte composition6 7 Lower frequency variations in C N P stoichiometry are related to changes in the structural elements of the phytoplankton cell A major breakthrough in our understanding of cellular C N P stoichiometry came with the realization that different cellular components have their own unique stoichiometric properties Most notably resource light or nutrients acquisition machinery such as proteins and chlorophyll is high in N but low in P whereas growth machinery such as ribosomal RNA is high in both N and P8 9 Because these components make up a large proportion of cellular material changes in their relative proportions have a marked effect on bulk cellular C N P stoichiometry Why then might the proportions of these components change 1 Department of Geophysics Stanford University Stanford California 94305 2215 USA 2005 Nature Publishing Group 349 INSIGHT REVIEW NATURE Vol 437 15 September 2005 Figure 1 Three different phytoplankton growth strategies and their resulting cellular N P ratios The allocation of resources and resulting N P ratios for the survivalist and the bloomer are from the optimization model of Klausmeier et al 10 Interestingly the model does not predict an optimal N P ratio of 16 our hypothetical generalist under any of the environmental conditions tested This indicates that the Redfield N P ratio of 16 observed in nature is simply an average value that reflects an ecological balance between the survivalists and bloomers in a population The survivalist Has a high N P ratio 30 Can sustain growth when resources are low Contains copious resource acquisition machinery The bloomer Has a low N P ratio 10 Adapted for exponential growth Contains a high proportion of growth machinery The