Quantifying food competition between two demersal fish species from spatiotemporal stomach content data

Abstract

Competition is challenging to quantify in nature and inference is often made on indirect patterns of potential competition (e.g., trends in population trajectories and overlap in spatiotemporal distribution and resource use). However, these indicators are not direct measures of fitness, nor do they say if the contested resource is limited in supply, which are key features of competition. We combine stomach content and biomass density data from scientific bottom trawl surveys to evaluate if food competition is occurring between two dominant demersal fish species in the southern Baltic Sea: Atlantic cod (Gadus morhua) and flounder (Platichthys spp.). We use multivariate generalized linear latent variable models (GLLVMs) to quantify diet similarities across the domain, diet overlap indices to test if predator density drives diet overlap at intermediate spatial scales, and spatiotemporal generalized linear mixed models (GLMMs) fit to individual-level prey weights to evaluate effects of local-scale covariates. We find clear dietary clusters by species and size. The latter is especially pronounced in cod, which shift from benthic to pelagic prey at around 30 cm. The dietary overlap is low and does not decline with predator density. However, signs of resource partitioning to reduce interspecific food competition are evident in the local-scale analysis. As flounder densities increase, small and large cod tend to feed less on the isopod Saduria entomon — an important prey species. However, the benthic prey weight in small cod, and benthic and total prey weight in large cod, are not affected by flounder densities, suggesting that interspecific food competition is not limiting cod feeding but affects their diet composition. We find some support for intraspecific food competition in large cod and flounder. Our study illustrates the importance of local-scale processes when inferring food competition from stomach content data, and such results are important for quantifying ecological interactions.

Competing Interest Statement

The authors have declared no competing interest.

Footnotes

* Changed to weighted regression and minos bug fixes

* https://github.com/maxlindmark/cod-interactions/tree/main