A deterministic model for N₂ fixation at stn. ALOHA in the subtropical North Pacific Ocean

Abstract

Marine N₂ fixation by diazotrophic microorganisms is a key process in biogeochemical cycling and yields an important input of new nitrogen into the tropical and subtropical surface ocean. However, it is only poorly accounted for in current numerical models. We present a simple biological model that explicitly includes N₂ fixation by diazotrophic phytoplankton. The model employs a mechanistic parameterization of N₂ fixation based on physiological responses of Trichodesmium to physical conditions of the environment. The model is conceived to allow shifts in nitrogen versus phosphorus control of the plankton community by resolving the biogeochemical cycles of both elements. Typical N : P ratios were assigned to the different functional groups to capture variations in the N : P stoichiometries of inorganic and organic matter pools. The biological model was coupled to a 1D physical model of the upper ocean. A simulation was performed at stn. ALOHA (22°45′N, 158°W) in the subtropical North Pacific Ocean, where intense blooms of Trichodesmium occurred during the last decade. The model captures essential features of the biological system including the vertical structure and seasonal course of chlorophyll, the seasonal cycle and interannual differences in diazotrophic biomass, the mean vertical particle flux, and an oscillation in the relative importance of nitrogen versus phosphorus. We regard this model as a step towards a mechanistic tool to assess the magnitude of marine N₂ fixation and to explore hypotheses on its effect on carbon sequestration from the atmosphere.

Introduction

The importance of marine N₂ fixation by diazotrophic phytoplankton in the present ocean has been stressed in several recent studies (Carpenter and Romans, 1991; Capone et al., 1997; Falkowski, 1997; Karl et al., 1997; Capone and Carpenter, 1999; Karl et al., 2001a). N₂ fixation by Trichodesmium, which is dominant diazotroph in tropical and subtropical waters, yields an important input into the global nitrogen cycle (Capone et al., 1997), as it supplies a significant fraction of new nitrogen in the subtropical Pacific (Karl et al., 1997), the tropical North Atlantic (Carpenter and Romans, 1991), and the Arabian Sea (Capone et al., 1998). Furthermore, the aperiodic appearance of large Trichodesmium blooms observed during the Hawaii ocean time-series (HOT) program has been recognized as an important source of biogeochemical variability in the subtropical Pacific. The massive occurrence of Trichodesmium changed the elemental composition of organic matter pools (Karl et al., 1992) and potentially switches the system from a nitrogen to a phosphorus limited state (Karl et al., 1997). The admitted importance of marine N₂ fixation has prompted speculation about its effect on the oceanic draw-down of atmospheric CO₂ by altering the export of carbon from surface waters (Hood et al., 2000).

N₂ fixation acts as a negative feedback mechanism in the global nitrogen cycle and might directly affect the partitioning of carbon between the atmosphere and the ocean (Falkowski, 1997). On long time scales of thousands of years, N₂ fixation, which is selected for by low N : P ratios and repressed by high N : P ratios, represents a mechanism to constrain ecological stoichiometry close to the Redfield ratio (Lenton and Watson, 2000). In the present ocean, phosphate is available in slight excess relative to nitrate (Tyrrell, 1999) and N₂ fixation represents a net source of fixed nitrogen in the global nitrogen cycle. N₂ fixation in turn is controlled by the available phosphorus, iron (a requirement for the enzyme nitrogenase), and oxygen (a potent inhibitor of nitrogenase). This input of new nitrogen drives the biological system from the current nitrogen limitation towards limitation by phosphorus, and potentially affects the magnitude of carbon sequestration from the atmosphere. Since on average only the biogenic material based on new nutrients can be exported from the upper ocean, the export of carbon by the biological pump depends directly on the supply of new nutrients.

Modeling efforts directed towards a better understanding of marine biogeochemical cycles and their response to anthropogenic perturbations have increased considerably in recent years. However, the achievement of long-term goals, like the prediction of carbon sequestration from the atmosphere by the oceans, still requires substantial improvements of the currently available numerical models. One main task is the better representation of functional diversity and multi-elemental cycling to capture major shifts in biogeochemical mechanisms (Doney, 1999). N₂ fixation is a key process that is not yet represented well in present models.

Here we present a simple model that explicitly describes the functional behavior of diazotrophs. The model includes the elemental cycling of nitrogen and phosphorus to allow a differentiation between nitrogen and phosphorus control and simulates the effect of N₂ fixation on the N : P stoichiometry in the euphotic zone. The phytoplankton community is divided into two functional groups, diazotrophs represented by Trichodesmium and other phytoplankton, which are characterized by different stoichiometric ratios. This variable treatment of N : P stoichiometries leads to an uncoupling of the nitrogen and phosphorus cycles. The parameterization of N₂ fixation is based on observed physiological responses of Trichodesmium to environmental conditions. Effects of temperature, irradiance, and wind speed are taken into account. The ecological model is coupled to a one-dimensional physical model of the upper water column at stn. ALOHA (22°45′N, 158°W) in the subtropical North Pacific Ocean. This site represents an excellent test case for a modeling study not only because a comprehensive set of physical and biological data are simultaneously collected during the HOT field program (Karl and Lukas, 1996), but also because dramatic changes in the microbial community structure, in particular massive blooms of Trichodesmium, occurred during the observational program. These blooms alter nutrient cycling mechanisms and stoichiometries of inorganic and organic matter pools (Karl et al., 1997). The present model is intended as a step towards a mechanistic simulation tool that allows a future assessment of the magnitude of marine N₂ fixation and an exploration of hypotheses on the effect of N₂ fixation on the sequestration of atmospheric CO₂.