Vertical Structure of Ocean Ecosystems: A Particle and Dissolved Organic View

The notion of chaotic behavior in ecosystems has produced many studies including the classic food chain model of Hastings and Powell [14]. Chaotic systems exhibit properties that seem at odds with the overall structure of the ocean such as paradox of enrichment and extinctions. Olson and Hood [15] for example show that a H-P model of the Gulf Stream produces an unreasonable north-south productivity gradient. To avoid chaos Steele and Henderson [16] suggest using quadratic mortality, such as μ P² in the model above. While this inclusion does make models more stable and can be rationalized [17], it does not eliminate the transition to chaos. Extension of an H-P type model to include age-structured Z and a microbial web by the author and Glenn Flierl also produces enhanced stability but still exhibits chaos in a full parameter space. The question becomes one of understanding the dynamics of the ocean becomes even more daunting when one considers the large number of species that play a role. The complexity of modeling the 300 or so species described by Venrick [18] in Hawaiian water columns is beyond current capabilities.

The question is whether this system is stabilized by this complexity and the coupling to microbial and a grazing community also consisting of a very high number of players. Steele’s rationale for suppressing chaos in models is that ecosystems should evolve to minimize chaos and its tendency to reduce net production and drive extinction. Independent of the question of chaos, the evidence of relationships such as Redfield ratios imply that there is strong adaptive forces at the ecosystem scale. It also seems that trying to increase the number of players is not feasible. Instead, an explicit treatment of particle size, i.e. P(ψ )where ψ is 2π / l and l are the particle size, could provide further insight. The goal is then to describe the spectrum of particle sizes produced in the photic zone. It is also important to include the spatial-temporal distributions of N and P, the horizontal flows with their associated turbulence and the temporal variations on time scales from the diurnal cycle, weather, seasons, and climate change at various scales.

Ocean physics ranging from fronts and mesoscale eddies to small-scale turbulence produces strong time-dependent patchiness in the ocean [19]. These physical factors can mask intrinsic or even stabilize inherent time-dependent biological interactions. Temporal issues such as the Gran-Sverdrup mismatch hypothesis of blooms [20,21] also lead to variable production and particle loss to depth. Various works are now embedding these biological models into physical models to understand dynamics of upper ocean ecosystems for biogeochemical and fisheries studies. The coupling in the vertical of the photic zone to the deeper aphotic layers, however, demands more investigation. At this point the hypothesis is that strong temporal variability in photic zone P leads to a spectrum of particles that contribute additional carbon to the deep ocean carbon pool added by the coupled biophysical dissolved pump. Moving from the photic zone with its net-production of organic matter into the deep ocean, again consider the balance of particulate material (living or dead) and dissolved components.

These components in the very large-scale ocean circulation make up the physical and biological pumping of carbon into the undersaturated deep sea. The physical transport of surface layer carbon involves currents that carry dissolved inorganic carbon (DIC; CO2, HCO3, CO3) and dissolved organic matter (DOM; proteins, lipids…) into the waters below the photic zone. The biological pump involves both the contribution to the DOC by DOM carried into the deep ocean by the physical pump and the sinking of particle matter (POC, living and dead). One might speculate that the correlation between DOM and flow into the world’s thermoclines and deep waters dominates. Understanding each component of carbon flux needs to be re-examined both in relationship to the carbon problem and in the understanding of the time scale of the global thermohaline circulation. The time scale of the global “conveyorbelt” circulation first described by Defant and Wüst is currently measured by 14C in DIC. The estimated thousand-year time scale [22,23] assumes an imprint of pre-bomb radiocarbon in waters off Greenland without any modifications other than mixing during their transport to the deep Pacific.

Recent measurements of DOC 14C by [2,3] raise significant questions both about the carbon pump and the time scale of the thermohaline circulation. These observations show that DOC is much older than the DIC in the deep ocean (2-6 X 103 vs 103 years). Their explanation involves “dead” carbon (14C = 0) from mid-ocean ridge (MOR) bio-production. The flux of carbon from MOR requires more investigation, but it seems that MOR probably produce more dead DIC than DOM. As an alternative think about the interaction between the deep-sea particle field, DIC, and DOC. Here again specific treatment of a particle size spectrum and the associated variation in particle sinking rate [24]. Measurements of sinking are provided by analysis of thorium and other insoluble elements [25,26]. Again, consider a particle-dissolved fraction model of vertical fluxes in the deep Pacific as shown in Figure 2. Without writing the equations, the model suggests a balance between upwelling of deep Pacific water that carry DIC and DOM upward and bacteria incorporation of DOM into POC.

The surface-initiated particle flux is sinking, moving against the slowly upwelling waters. This counter-flow of carbon components is coupled with the action of microbes that exchange carbon between pools of DIC, DOC, and POC. The model can produce the age differences, but at the same time modifies the understanding of the upward water flux that is measured by ¹⁴C among other means. The exchanges involved with M-M uptakes are also essentially unknown. The model can be formulated with or without an explicit bacteria component (B). In the simple models a broad range of possible parameters allows the model to achieve the age difference observed by Druffel, but the model is very poorly constrained in terms of its parameters. One means of closing this problem is to introduce other tracers. An obvious choice of ¹³C does not seem to provide a constraint. A coupling of the model and the photic zone model to include multiple currencies, i.e. nitrogen, carbon…., is suggested. This allows an explicit consideration of the Redfield ratio and its variation [27,28]. Finally, the eventual contribution of organic matter (P or N) to DIC suggests that the age estimates of the deep Pacific and timescale of the conveyor belt should be re-examined. The conclusion is that vertical particle fluxes have a bearing not only on the biological pump, but on the biophysical portion of it and the global deep circulation itself.