The ecological foundation for ecosystem-based management of fisheries: mechanistic linkages between the individual-, population-, and community-level dynamics

Lennart Persson 1 Anieke Van Leeuwen 0 Andre M. De Roos 0 0 Institute for Biodiversity and Ecosystem Dynamics, University of Amsterdam , PO Box 94084, Amsterdam NL-1090 GB , The Netherlands 1 Department of Ecology and Environmental Sciences, Umea University , Umea SE 90187 , Sweden Persson, L., Van Leeuwen, A., and De Roos, A. M. The ecological foundation for ecosystem-based management of fisheries: mechanistic linkages between the individual-, population-, and community-level dynamics. - ICES Journal of Marine Science, 71: 2268 - 2280. Food-dependent growth and size-dependent interactions form cornerstones in the dynamics of fish populations. Using two freshwater examples, we illustrate the importance of considering both these cornerstones for understanding system dynamics. Moreover, a proper understanding of the dynamics requires mechanistic linkages between individual-, population-, and community-level processes based on mass conservation principles. In one example, we further find that quantitative predictions of individual-level energy flows are essential for understanding the community dynamics. This mechanistic approach to understanding system dynamics is generally not reflected in fisheries models as an overview shows that only half of them incorporate food-dependent growth, and none fully observe the principles of mass conservation. As a marine example we examine patterns in the Baltic Sea system and show that no relationship between cod growth and sprat biomass is present related to the low size resolution in prey fish. Linking individual cod performance to its resource base is complicated by the many prey types cod uses over its life cycle. We conclude that an ecological perspective including size- and food-dependent processes is vital for ecosystem-based fisheries management making necessary a proper description of the interactive trophic structure as a result of mechanistic linkages between individual, population, and community processes. - Text books in ecology generally assume that the dynamics of populations are the result of mortality and reproduction only (Begon et al., 1996; Turchin, 2003). In this definition, one basic aspect of the individuals life historyontogenetic growth or developmentis ignored, although ontogenetic growth is indeed a vital process, preceding reproduction by ensuring the energy allocation for the production of biomass. Fish provide an obvious example of the importance of individual development as individuals commonly increase in body weight over several orders of magnitude while developing from egg to mature individual (Werner and Gilliam, 1984). Moreover, the rate by which an individual fish grows is generally dependent on food availability, a fact long recognized and described by fish biologists (Beverton and Holt, 1957; Backiel, 1978). The pioneer paper by Hjort (1914) on the dynamics of fish populations also highlighted the changes and variability in individual growth of fish. Fisheries management is mainly concerned with the biomass production of entire populations, which represent a collection of many individuals. Fisheries management approaches generally ignore the resource dependence of this production. Studies inspired by the dynamics of fish populations in freshwater lakes (cf. De Roos and Persson, 2001; Persson et al., 2003, 2004), however, illustrate the potential of gaining a thorough insight in community regulation and functioning from models that consistently link population output (biomass production) to individual energy acquisition (food-intake). With its focus on stock production and its basis in single-species and age-structured models, marine fisheries management has traditionally incorporated these insights only to a limited extent. Following the research experience from trophic cascades in lakes (Carpenter et al., 1987; Carpenter and Kitchell, 1993), increasing evidence now shows that also in marine communities changes at the top of the foodweb influence overall ecosystem dynamics: examples include the Black Sea (Daskalov et al., 2007), the North West Atlantic (Frank et al., 2005), and the Baltic Sea (O sterblom et al., 2006; Mollmann et al., 2008). Understanding such community-wide changes requires an ecosystem-based management perspective that strives after an integrated assessment, considers the functioning of the entire ecological community (FAO, 2003; Christensen and Walters, 2004), and mechanistically accounts for the linkages between the individual, the population, and the community level. Such a general understanding, however, has been lacking so far for two main reasons: first, available data have been under-valued and little-used (e.g. changes in population size distributions are not incorporated in the estimates of reproductive capacity or spawning-stock biomass), and second, there has been a lack of particular and essential information that allows for an appropriate description of the trophic configuration to be studied. Where the first issue is possible to solve by turning around and taking an ecological perspective, the second part calls for a reconsideration of what kind of information is essential for developing an understanding of the dynamics of systems that are heavily influenced by size-structured interactions. To illustrate in more detail the points made above, we first discuss two examples from freshwater systems, which both show that to understand the community dynamics it is essential (i) to consider the size- and food-dependent interactions and (ii) to mechanistically link processes at different levels (individual, population, community) of organization. The second example additionally shows that a proper quantitative handling of individual-level processes by using rather detailed physiologically structured population models (Metz and Diekmann, 1986; De Roos and Persson, 2001) may be necessary to gain an understanding of the dynamics of the system as a whole. Next, we discuss how the models that are commonly used for fish stock assessments and predictions for stock production with a focus on multispecies models account for the basic concept of resource-dependent development and biomass production and contrast this with assumptions of the aforementioned physiologically structured population models. Finally, we focus on the Baltic Sea system as a marine example, where we first consider the extent to which the data that are usually available for marine fish stocks can or cannot provide a sufficient basis for a more thorough ecological understanding. We point out that insufficient time resolution and especially resolution of size frequency in the data on both predatory as well as prey fish tend to prevent establishing a proper link between resource (prey) availability and consumer (predator) performance. Furthermore, we discuss the basic problem of deriving a proper trophic description of the dynamic interactions between different trophic components. We also point out that th (...truncated)