Trait-environment relationships in marine fish : An exploration of patterns, drivers and assembly processes shaping marine fish communities

Abstract

Marine ecosystems are exposed to a multitude of environmental and anthropogenic pressures, such as overfishing, habitat loss, invasive species and global warming. Such pressures alter the structure and biodiversity of natural communities, thereby potentially affecting the functions and services that marine ecosystems provide, such as food and coastal protection. The goal of ecosystem-based management is to sustainably make use of marine ecosystems whilst protecting its health and functioning. This requires knowledge on how natural communities respond to the range of environmental and anthropogenic pressures. The trait-based approach may provide such knowledge and understanding. It is based on the traits that species carry, which determine which environments a species can inhabit and how it may respond to disturbances. A trait can be any characteristic that one can measure on an organism, and can be related to its behaviour, morphology, life history or physiology. Moreover, characterizing natural communities in terms of their traits allows for comparing ecosystems with entirely different species compositions, thereby enhancing the possibility of finding general patterns across communities and ecosystems. In this thesis I applied a trait-based approach to marine fish communities with the aim to understand how they are structured in terms of traits, and to use traits to assess how fish communities respond to changes in environment and fishing pressure. The response of a community to a change in environment or an anthropogenic disturbance can be detected by investigating shifts in the mean trait values expressed by the community. The North Sea fish community has a history of intense fishing pressure, as well as climate-driven changes during the last few decades. Furthermore, there are strong spatial gradients in environmental conditions, from the shallow, southern parts to the deeper, northern parts where the North Sea meets the North Atlantic. We therefore explored if the North Sea fish community shifted in its mean trait values over time, as well as in space, and if this could be explained by the environment and fishing effort. The community showed a strong temporal shift in its trait composition from large, slow-growing, late-maturing and long-living mean trait values to smaller, faster-growing, earlier-maturing and shorter-living ones. Although the high historical fishing pressure on large species, such as Atlantic cod, and the consequent decrease of such species likely contributed to these shifts, we found that they could also be explained by the increases in temperature, salinity and phytoplankton biomass. However, the temporal changes over time were not the same throughout the area, with some areas showing no shift in community trait means. Moreover, the trait composition of the community varied strongly in space, following the gradients in temperature, seasonality and depth. Our results demonstrated that traits can be used to assess how marine fish communities are structured in space and respond to environmental changes over time, and emphasize the importance of taking both spatial and temporal dimensions of marine fish communities into account. In order for the trait-based approach to be useful in predictive models, it is important to test the generality of the empirical relationships between traits and environment, and to identify which traits respond most strongly to which environmental variables. We therefore combined data on fish communities from the North Atlantic and Northeast Pacific, including >1,200 species, that cover large latitudinal and environmental gradients. We found that particularly traits related to the fast-slow continuum, i.e. age at maturity, lifespan and the growth coefficient K, varied most strongly with environment. Hence, warm, shallow and seasonal waters were associated with fast-growing, early-maturing and short-living species, whereas cold, deep and stable waters were mostly inhabited by slow-growing, late-maturing and long-living species. When using the observed trait-environment relationships to project the trait composition of marine fish communities worldwide, we found that traits were following both large-scale latitudinal gradients in temperature, as well as local coastal-offshore gradients in depth and seasonality. Given the consistent patterns observed across areas with entirely different species composition our results indicate that these general trait-environment relationships of marine fish may prove useful to predict changes in marine fish communities over time. Besides knowing how communities are structured in terms of traits, it may provide additional insight to identify the underlying community assembly processes that lead to the observed structure and composition of marine fish communities. Potential community assembly processes could be related to both biotic and abiotic factors. In this thesis, we studied community assembly of marine fish for the first time at a large spatial scale by assessing if the variation in traits within communities was different from random. Some communities were found to be strongly shaped by the environment that acted as a filter by only selecting species with a particular set of traits. For instance, high temperatures filtered out species with a large size and a long lifespan. Several communities were also found to be shaped by biotic interactions, as observed from the high variation in fecundity and offspring size, suggesting that multiple reproductive strategies can coexist within the same community. However, the majority of communities seemed to be randomly assembled, suggesting that both processes may act simultaneously, thereby cancelling each other out, or that stochastic processes like dispersal and immigration/emigration are more important. The strength at which the observed assembly processes were operating varied moderately with environment, particularly with temperature and depth, but not with fishing pressure and primary production. Further knowledge on potential other community assembly processes acting on marine fish and the inclusion of intraspecific trait variation are needed to fully understand the underlying mechanisms of marine fish communities are assembled. This thesis demonstrated that the structure of marine fish communities is strongly linked to the environment by acting on the trait composition of communities – both in space and time, as well as at small and large spatial scales. We identified key response traits and trait-environment relationships for marine fish that should be further explored and tested in other types of ecosystems. This thesis will hopefully inspire future research to create predictive models that can explore how marine fish communities may change under global warming, in combination with fishing, and to inform ecosystem-based management on the expected changes to come. This will allow users of marine ecosystems to adapt and to continue to benefit from the services provided by marine ecosystems in the future.