# Weak effects of local prey density and spatial overlap on predation intensity in a temperate marine ecosystem

## TL;DR

This study finds weak links between prey availability and predation by cod in the Baltic Sea, challenging assumptions in marine ecosystem models.

## Contribution

The study uses spatiotemporal models and long-term data to assess predator-prey interactions in a temperate marine ecosystem.

## Key findings

- Predator-prey overlap has decreased over three decades in the Baltic Sea.
- Only Saduria shows a clear link between prey availability and predation by cod.
- Pelagic prey like herring and sprat show weak connections to cod predation patterns.

## Abstract

Quantifying the impact of lower trophic level species abundance on higher trophic level predators (and vice versa) is critical for understanding marine ecosystem dynamics and for implementing ecosystem‐based management. Trophic ecosystem models generally predict a tight coupling between prey and fish predators, such that higher abundance of lower trophic species increases the abundance of higher trophic level predators. This assumes that predator feeding rates are limited by prey availability to some degree. Despite being a key component of predator–prey interactions and multispecies fisheries management, relatively few studies have assessed the impacts of prey availability on predation patterns of mobile, generalist fish predators using spatiotemporal models and local‐scale stomach content, predator, and prey data. In this study, we explore the association between local density of key prey and predator stomach contents, and predator–prey spatiotemporal overlap and predation indices, using the Baltic Sea as a case study. We use three decades of spatially resolved biomass and stomach content data on Atlantic cod (Gadus morhua), and biomass data on three of its key prey: herring (Clupea harengus), the isopod Saduria entomon, and sprat (Sprattus sprattus). Using geostatistical generalized linear mixed‐effects models fitted to relative biomass density and prey‐mass‐per‐predator‐mass, we estimate spatiotemporal trends and annual indices of biomass‐ weighted and area‐expanded per‐capita and population‐level predation, predator–prey overlap, and the correlation between these indices. Range shifts have resulted in reduced predator–prey overlap over time, which is now the lowest in three decades. For Saduria, we find an association between prey availability and stomach contents, but not for herring or sprat. Similarly, only in Saduria do we find a positive correlation between population‐level predation indices and the spatiotemporal overlap. Although behavioral interactions with pelagic prey are challenging to infer from stomach content and acoustic data due to high mobility leading to fine‐scale spatiotemporal mismatch, the weak connection with local‐scale availability, and low correlation between population‐level predation and spatial overlap, could imply weaker coupling between pelagic prey and cod than previously thought. These findings provide key information on the strength of species interactions, which is crucial for the continued development of multispecies models and ecosystem‐based fisheries management.