Marine fish diversity patterns, drivers and underlying processes: Present status and predictions under climate change

Biodiversity is the variety of life forms, from intraspecific gene diversity, to different species, communities and entire ecosystems. The uniqueness of all life forms grants them an intrinsic value. But, biodiversity also contributes to human wellbeing by providing key goods and services that human societies rely on, such as food production and climate regulation. Nonetheless, biodiversity is declining at an unprecedented rate due to anthropogenic pressures such as climate change and overexploitation. Consequently, we are now facing what could be the sixth mass extinction on Earth since life began over 3.5 billion years ago. Anticipating the effects of climate change (and other anthropogenic activities) on natural ecosystems is therefore essential for taking the necessary management and conservation decisions that seek to preserve biodiversity, and thus human well-being. Part of the decisions that need to be taken shall address the suitable placement of Marine Protected Areas (MPAs) that guarantee the adequate management and conservation of “areas of particular importance for biodiversity” in order to reach the objective of protecting 30% of marine areas by 2030.

In this thesis, I use modeling tools to explore the underlying drivers shaping marine fish communities, train models capable of estimating patterns and trends in fish biodiversity, and assess the efficiency of current MPAs in safeguarding this biodiversity.

I begin by studying the underlying drivers of marine fish communities at two different spatial scales, i.e., regional and continental. At the regional scale, I work with the case study of the North Sea, using the survey data from the North Sea International Bottom Trawl Survey to explore the fish community by modelling the responses of 67 species to environmental drivers and fishing pressure. For the continental scale, I compile data from 13 different surveys from the Northeast Atlantic Ocean, modelling the distribution and specific responses of 151 fish species to environmental change. At both spatial scales, temperature and productivity stand out as the main environmental drivers of fish community composition. Moreover, I identify spatially-structured processes that contribute in explaining the community composition. Based on the continental-scale model, I then explore the time horizon at which different facets of marine fish biodiversity can be reliably predicted. The results highlight that, although there are some differences across metrics, the models allow reliable forecasting of fish biodiversity for (at least) 10-20 years ahead, which is a relevant and reasonable timeframe to consider from a management and conservation perspective. Consequently, I use the validated models to predict different biodiversity indices and explore how they are expected to change during the coming decades. The findings highlight that current MPA network is far from the goal of effectively protecting 30% of marina areas, and that the proportion of protected areas with high biodiversity is low in the Northeast Atlantic Ocean. Lastly, I set up a management and conservation scenario (i.e., MPA optimization) where the aim is to protect 30% of the high biodiversity areas while minimizing the impacts on current fisheries (i.e., fishing effort). The findings indicate that current MPA network is protecting a small percentage of high biodiversity areas, due to the small surface covered by MPAs and a mismatch with high biodiversity areas. The MPA optimization scenario shows a high efficiency towards protecting current and future (2030, 2050) biodiversity, with relatively low and very localized imapcts on fishing effort. The findings highlight that there is room for improving current MPA network, and that maximizing their efficiency will be achieved by a more active engagement of stakeholders.

The presented work significantly enhances our understanding of marine ecology by advancing knowledge on community assembly processes. Additionally, it demonstrates the utility of modeling tools in anticipating the effects of climate change and informing management and conservation actions.