The European Research Training Network on Fisheries-induced Adaptive Changes in Exploited Stocks (FishACE) is set up to investigate the prevalence and consequences of fisheries-induced adaptive changes in exploited aquatic systems in European waters. This objective necessitates the development and application of novel methodological tools for investigating empirical data, together with the careful construction of theoretical models suitable for complementing empirical analyses and evaluating managerial options. At the same time, the network will provide advance training for a new generation of scientists who will be educated to tackle the challenges posed by evolutionary changes in exploited resources.

Today, fishing is the dominant source of mortality in most commercially exploited fish stocks. According to the United Nation’s Food and Agricultural Organization, world capture fisheries have reached a ceiling, with three stocks out of four being maximally exploited or overexploited. Since all fish species were genetically adapted to the environmental conditions experienced prior to intensive exploitation, the current, drastically altered conditions cannot possibly leave their life-history patterns unaffected. In other words, fishing not only decreases the abundance of fish, but also changes their genetic composition. This evolutionary dimension of fisheries has been overlooked or downplayed for decades, so that fisheries scientists and managers are just now awakening to the formidable risks posed by further unmanaged fisheries-induced evolution.
This awakening appears to have been facilitated by two independent developments. First, over the past decade the formerly narrow focus in fisheries management has been widened, from the traditional goal of avoiding over-fishing and securing maximum yield, to an ecosystem perspective recognizing a much wider range of potential threats to marine biodiversity and its constituents. Consequently, in compliance with the FAO Code of Conduct for Responsible Fisheries, most major fishing nations in the North Atlantic have recently agreed to respect a precautionary approach to “conservation, management and exploitation of living aquatic resources in order to protect them and preserve the aquatic environment, taking account of the best scientific evidence available”.
As a second factor, a suit of scientific advances have suggested that fisheries-induced evolution is ubiquitous and that, if unmanaged, such processes may pose serious threats to exploited stocks and to their value as resources for humankind:

• There is growing recognition that fishing causes severe changes in the demographic properties of exploited stocks. This applies, in particular, to maturation: in most stocks, fish mature earlier today than they used to only a few decades ago.


• At the same time it has been shown that earlier maturation may have adverse implications for the reproductive potential of fish stocks, not only because large females produce more offspring per unit of body weight than smaller ones, but also because the size of females and the quality of their offspring tend to be positively correlated.


• Furthermore, thanks to recently developed improved statistical methods, it is becoming increasingly clear that most of the documented changes in maturation are indeed evolutionary responses, and not mere effects of phenotypic plasticity.


• Corroborating theoretical expectations, it has recently been demonstrated experimentally that size-selective fishing can cause genetic reductions in growth that result in a decline of harvestable biomass.

The scientific agenda proposed for this Research Training Network will be based on two major lines of analyses: analyses of empirical data and analyses of theoretical models. The notion of reaction norms provides the conceptual framework for most of the empirical studies suggested in this proposal. Reaction norms describe the different phenotypes produced by a single genotype under different environmental conditions. Reaction norms themselves are genetically determined traits that can evolve under natural selection. Estimation of reaction norms thus provides means to overcome confounding effects of environmental changes in phenotypic data.

The empirical analyses are based on three main data sources:

• Individual-level biological data. Extensive time series of measurements of length, body weight, age, and maturity stage are available through scientific surveys and sampling of commercial catches, designed to give a representative picture on the structure of populations and for the basis of their rational management.


• Population-level biological data. Individual-level data are supplemented by data aggregated to population level estimates. These are often available from the reports of the expert groups working under the auspices of the International Council for Exploration of the Sea (ICES).


• Environmental data. In addition to data on fish populations in focus here, various types of environmental data are also available, including abundance estimates of other fish species, phytoplankton and zooplankton biomasses, temperature, salinity, and movements and inflows of water masses.
  

Empirical data will be analysed using an array of statistical methods. These include the established statistical toolbox, comprising both classic methods and modern recent developments such generalized linear models, mixed models and geostatistics. These methods will be complemented by more specialised, recently developed tools for estimating reaction norms for age and size at maturation from various types of often incomplete fisheries data, partly relying on computation-intensive methods such as bootstrapping and randomisation. Analyses of artificial data generated with simple population models will be used to validate analyses tailored for particular data.
For modelling the evolution of reaction norms, three approaches will be used, each emphasizing different aspects ecological and evolutionary dynamics:

• Quantitative genetic models. This time-honoured framework has been developed for the study of quantitative traits influenced by many genes of small additive effects, and originates from the context of animal breeding.


• Adaptive dynamics models of reaction norms represented as function-valued traits. Similarly to certain advanced quantitative genetics models, these models represent reaction norms as infinite-dimensional functions, and thereby allow any shape to evolve. However, in contrast to quantitative genetics models, a realistic description of the ecological setting is emphasised. In order to achieve this, genetic detail is sacrificed.


• Eco-genetic models. This novel approach aims at combining a realistic description of the ecological setting as well as population structure with a description of genetic detail at a level that still allows predicting the rate of evolutionary change.

More specific evolutionary repercussions of exploitation will be studied based on two further model classes:

• Evolutionary food web models. The type and strength of interactions between species pairs, as well as the distribution of interactions in the food web as a whole influences the evolutionary effects of fishing. Food web models are characterized by their topology and distribution of interaction strengths among the nodes in the web.


• Evolutionary energy allocation models. Energetic trade-offs constrain evolution of all key life-history traits. Energy allocation models consider the ecological and evolutionary dynamics of energy allocations to growth, reproduction, and maintenance. In addition, behavioural sub-models for describing foraging behaviour can be included.