The role of zooplankton for carbon export, nutrient recycling and phytoplankton bloom phenology in an ocean biogeochemical model

Abstract

Marine zooplankton, i.e., heterotrophic marine plankton, serve as trophic links between primary producers and higher trophic levels, and as recyclers for nutrients and carbon in the pelagic ecosystem. In addition, they play a major role for the carbon export flux due to fecal pellet production and fragmentation of particles. They are distributed all over the ocean and constitute a large variety of organisms. Because of large uncertainties in the estimation of parameters and the forms of equations, zooplankton are often parameterized in strongly simplified forms in ocean biogeochemical models. Nowadays, however, increasing data availability from experiments and observations makes it possible to implement different zooplankton functional types in models. This thesis presents the implementation of new zooplankton functional types into an ocean biogeochemical model. Subsequently, the sensitivity of net primary production, carbon export and nutrients to the implementation of these new zooplankton functional types was analyzed. In my thesis, I use a global setup of the biogeochemical model Regulated Ocean Ecosystem Model (REcoM) coupled with the Finite Element Sea-Ice Ocean Model (FESOM). I implemented an explicit parametrization of micro-, meso-, and polar macrozooplankton based on process rates and biomass observations from the literature, as well as a representation of fast-sinking detritus. This extended version of REcoM was used to analyze the role of zooplankton for carbon export, nutrient recycling, and phytoplankton bloom phenology. In a second step, a new sinking routine that considers the roles of mineral ballasting and seawater viscosity on the particle sinking speed and the effect of oxygen on remineralization rates was added to the model. This set-up was used to assess the role of each factor (ballast minerals, seawater viscosity, and oxygen concentration) for the export and transfer efficiencies of carbon, i.e. the amount of particulate organic carbon that is exported across the euphotic depth and reaches the deep ocean. The implementation of the new zooplankton groups changes the carbon transfer efficiency and net primary production in the model. Publication I and III highlight the influence of zooplankton on the transfer efficiency of carbon. Publication I shows that the transfer efficiency of carbon reaches up to 50% due to the high biomass of polar macrozooplankton in the Southern Ocean. Similarly, it was illustrated in Publication III that the high mesozooplankton biomass increases the transfer efficiency of carbon to 80% in the Equatorial Pacific. In addition, the model results presented in Publication I and II show the stimulation of net primary production due to the fast recycling of nutrients. After the parametrization of three zooplankton functional types, the new state of the model leads to a 25% increase in annual mean net primary production. In addition to the effects on annual mean bulk fluxes, the more complex representation of zooplankton also affects the timing of phytoplankton blooms and biogeochemical fluxes. Zooplankton fecal pellets constitute an important share of sinking particulate organic carbon depending on the season in the Southern Ocean. In Publication I, it is shown that the typical shift from a dominance of phytodetrital aggregates in spring to zooplankton fecal pellets later in the year is now reasonably reproduced by the model after the implementation of polar macrozooplankton. Zooplankton grazing can play a decisive role in phytoplankton bloom phenology since it is a loss mechanism for phytoplankton. In Publication II, it is shown that the increased loss rates of phytoplankton due to stronger zooplankton grazing lead to the later start of the spring bloom. In addition, nutrient recycling by zooplankton prevents the fast exhaustion of nutrients by phytoplankton and consequently leads to a later end date of the bloom. In the end, the more complex parametrization of zooplankton provides a modeled phytoplankton bloom phenology closer to observations. The results also indicate that the explaining mechanism behind the bloom phenology changes. While the start of the spring bloom is explained better with the ’Critical Depth Hypothesis’ in the low grazing scenario, the system aligns with the ’Dilution-Recoupling Hypothesis’ in the high grazing loss simulation. Finally, the global spatial distribution of export and transfer efficiencies are analyzed in Publication III. In particular, I examined the impact of ballast minerals, seawater viscosity, and oxygen-dependent remineralization on export and transfer efficiencies. These three processes are often not considered in biogeochemical models. My results show that the global mean of export efficiency across the euphotic zone stays similar ( 13%) when the effects of mineral ballasting, seawater viscosity, and oxygen-dependent remineralization are added to the model. However, the global mean carbon transfer efficiency is more sensitive to these processes and varies between 25% and 32% in different simulations dependent on the representation of these processes. The magnitude of the effect of each process varies spatially. While the effect of ballast minerals can increase the transfer efficiency by a factor of nine in high latitudes and subtropical gyres, including oxygen-dependent remineralization can increase the transfer efficiency by 28% in low latitudes. The influence of seawater viscosity on the transfer efficiency is smaller compared to the other effects, and it increases the transfer efficiency 8% in subtropical gyres. The thesis highlights that the zooplankton compartment in biogeochemical models should not only be treated as a closure term, and zooplankton functional types should be implemented in the global ocean biogeochemical models by using available datasets from the literature. It further underscores that missing out process representations of mechanisms that underlie carbon export has considerable effects on estimated carbon transfer efficiencies in biogeochemical models. Thus, further attention should be paid on the representation of missing processes related to particle formation and sinking.