Inferring associations between marine foundation species and associated fauna: a trait-matching approach

Paul Sourice¹, Thomas Benoit¹, Aurélien Boyé¹, Mathieu Chevalier¹, Martin Marzloff¹

¹Ifremer, France

Marine foundation species play a crucial role in supporting coastal biodiversity by buffering environmental stress and providing biogenic habitats for a wide range of species. However, relationships between foundation species and associated species are often poorly understood, as they involve a myriad of ecological processes. Trait-based approaches can provide valuable mechanistic insights on how biogenic habitats can drive benthic biodiversity. Yet, except for a few well-studied habitats such as coral reefs or seagrass meadows, habitat formers’ influence on marine biodiversity, and the role of their traits in shaping benthic communities are rarely quantified. Using monitoring data from the Reef Life Survey global dataset, which includes diver-based visual censuses that surveyed the occurrence of marine fauna (including fish, and invertebrates) across diverse biogenic habitats, we aim to characterise associations between species traits and biogenic habitat characteristics in a two-step manner: first, we reconstructed bipartite networks between benthic habitats and associated fauna, using a range of alternative inference techniques; and then, we interpreted these co-occurrence networks between foundation species and associated fauna in the light of trait databases (related to both foundation and associated species) to investigate potential trait-matching patterns. This framework will help understand and predict the cascading effects resulting from habitat degradation, and identify species that are most sensitive to habitat changes. Additionally, we aim to discriminate between the traits of habitat-specialist taxa and those of generalist species. By relating faunal traits favoured by specific habitat characteristics, we hope to disentangle the key mechanisms underpinning variability in reef species communities. Importantly, this work constitutes a first step towards an original modelling framework of coastal benthic ecosystems, which will capture trait-mediated ecological feedback between habitat-formers and associated fauna.