| In 1995, 64% of cod caught in the Baltic International Trawl survey were above 35cm. By 2019, that number had declined to 25%. At 35cm, sprat becomes the main prey of cod, however, due to this reduction in size, fewer cod now predate on sprat. Consequently, cod has effectively lost its role as top predator in the Baltic Sea ecosystem, with potential cascading effects (positive) effects on phytoplankton (via sprat and zooplankton).
Whilst historic fishing caused the crash of the Baltic cod population, it did not necessarily directly cause the decline in growth—the main cause for the small cod. In fact, despite the low body growth being one of the main existential threats to Baltic cod alongside high natural mortality (their population biomass is predicted to decline even in the absence of fishing), we know very little about why cod do not grow. Studies have linked the low growth to food availability, but there is uncertainty in how the quantity of food in cod stomachs has changed, and whether growth or body condition (the weight of cod relative to its length) has declined in direct response to changes in food availability.
We address these critical knowledge gaps by applying spatiotemporal statistical modelling techniques to a newly updated stomach content database of more than 110.000 individual cod stomachs, sampled over 6 decades. We will answer the following questions: (1) How has the amount of food changed (total and for the key prey) in cod stomachs over time and space? (2) How well do stomach contents correlate with growth trends? (3) Are fluctuations in stomach content of specific prey correlated with their abundance and/or the body condition of cod?
The project will be conducted in 2024–2025 (12 months, starting in March). The data are freely available on a database hosted by ICES (the International Council for the Exploration of the Sea). It is currently being updated with a planned re-release in November 2023, as part of a CINEA project (The European Climate, Infrastructure and Environment Executive Agency), where M. Lindmark and collaborator C. Griffiths have leading roles. Spatiotemporal data (sampled over time and space) come with statistical challenges, such as autocorrelation, which can invalidate conclusions. As a remedy, data are often aggregated, at the cost of information. To avoid this, we will apply novel spatiotemporal modelling techniques that directly account for autocorrelation.
This project is central for developing effective conservation management strategies. Based on very limited knowledge, studies have already suggested that fishing for flounder could improve feeding opportunities for cod (by reducing competition), and that we should shift fishing effort to ensure high densities of pelagic species where cod are. However, critical knowledge gaps remain and must be addressed before such management strategies can be considered. For instance, how much has the food in cod stomachs declined over time? Is cod growth even limited by food availability? Without answers to these key questions, management strategies will be ill-informed. Thus, this work will help guide management strategies for the conservation of Baltic cod and help maintain its role in structuring the Baltic Sea ecosystem across trophic levels. |