A size-structured food-web model for the global ocean*

We present a model of diverse phytoplankton and zooplankton populations embedded in a global ocean circulation model. Physiological and ecological traits of the organisms are constrained by relationships with cell size. The model qualitatively reproduces global distributions of nutrients, biomass, and primary productivity, and captures the power-law relationship between cell size and numerical density, which has realistic slopes of between 21.3 and 20.8. We use the model to explore the global structure of marine ecosystems, highlighting the importance of both nutrient and grazer controls. The model suggests that zooplankton : phytoplankton (Z : P) biomass ratios may vary from an order of 0.1 in the oligotrophic gyres to an order of 10 in upwelling and highlatitude regions. Global estimates of the strength of bottom-up and top-down controls within plankton size classes suggest that these large-scale gradients in Z : P ratios are driven by a shift from strong bottom-up, nutrient limitation in the oligotrophic gyres to the dominance of top-down, grazing controls in more productive regions. The size structure of phytoplankton communities is an important determinant of marine ecological function and biogeochemical cycling. Whereas the biomass produced by small phytoplankton is rapidly recycled in the microbial loop at the ocean surface, larger cells sink more rapidly, transporting carbon to the deep ocean and driving the biological carbon pump. Phytoplankton communities that are dominated by large cells are also thought to be associated with short, direct food chains that support large fish populations. These features of marine communities, together with the strong empirical relationships that are found between physiological traits and organism size (Litchman et al. 2007), have motivated the development of marine ecosystem and biogeochemistry models towards the explicit representation of phytoplankton size classes (Moloney and Field 1991; Baird and Suthers 2007; Banas 2011). The biogeochemical and ecological functions of marine ecosystems are also affected by taxonomic diversity. Different taxa are often associated with different elemental composition and biogeochemical roles, and recent models have begun to include many different functional groups, in an attempt to capture the taxonomic and biogeochemical diversity of marine communities (reviewed by Hood et al. 2006). The interaction of these factors gives rise to a clear global biogeography. Low nutrient regions are dominated by small phytoplankton, such as Prochlorococcus, Synechococcus, and picoeukaryotes, whereas more productive regions support not only these small cells, but also an additional abundance of larger species, including the coccolithophores, diatoms, and dinoflagellates. These patterns have been observed at local (Schartau et al. 2010), regional (Raimbault et al. 1988), and global scales (Hirata et al. 2011). The large-scale size distribution is so pronounced that its global signature has been detected from space (Kostadinov et al. 2009). This pattern of size–class superposition occurs because the amount of phytoplankton biomass in each size class appears to be limited (Chisholm 1992), and total biomass is typically distributed fairly evenly among logarithmically spaced size classes (Sheldon et al. 1972; Chisholm 1992). This leads to a power-law relationship between phytoplankton size and numerical abundance that is ubiquitous throughout the global ocean (Kostadinov et al. 2009). The exponent, or slope, of this relationship is typically found to lie between 21.5 and 20.75 (Cavender-Bares et al. 2001; Cermeño et al. 2006), but in general it is thought to become less steep (i.e., less negative) as total biomass increases and larger cells become established (Kostadinov et al. 2009). Figure 1 shows the relationship between phytoplankton size, numerical abundance, and total phytoplankton biovolume in the eastern equatorial Pacific, observed during the IronEx II iron fertilization experiment (Schartau et al. 2010). The dashed and solid lines correspond to measurements taken inside and outside a mesoscale patch of water in which a phytoplankton bloom was stimulated by the addition of dissolved iron. The slope of the powerlaw relationship changes with total phytoplankton abundance, becoming less negative inside the iron-fertilized patch, as the abundance of large cells increases in relation to small cells. What mechanisms dictate this size structuring of marine communities? Empirical observations and theoretical considerations reveal that the physiological rates and ecological interactions of plankton are strongly correlated with organism size. For example, small phytoplankton typically have higher light and nutrient affinities than larger phytoplankton (Finkel 2001; Litchman et al. 2007), and smaller cells are also known to have the largest maximum growth rates, at least within taxonomic groups (Tang 1995). The relationships between phytoplankton size and community structure have previously been explored in numerical models. These range in complexity from zerodimensional models (Laws 1975; Armstrong 1994; Banas 2011), to more complex regional studies (Moloney and Field 1991; Baird and Suthers 2007; Stock et al. 2008). In * Corresponding author: [email protected] 1 Current address: CERES-ERTI, École Normale Supérieure, Paris, France Limnol. Oceanogr., 57(6), 2012, 1877–1891 E 2012, by the Association for the Sciences of Limnology and Oceanography, Inc. doi:10.4319/lo.2012.57.6.1877