Population‐Level Habitat Breadth Varies With Richness in Reef Fishes

ABSTRACT Aim It has been hypothesised that niche breadth decreases with richness due to interactions, such as competition, forcing species to specialise. This hypothesis has been tested at the community‐level using species‐level niche breadth estimates. However, evidence for changes in niche‐breath among populations of the same species are scant. Our aim was to examine the niche breadth to richness relationship within species, which is crucial for understanding the role of interactions, as opposed to large‐scale climate, in altering realised niche breadth. Location The Pacific Ocean. Time Period 1988–2015. Major Taxa Studied Fishes. Methods We focus on reef fishes along a large‐scale richness gradient not accompanied by marked environmental changes. Fishes were surveyed in four distinct habitats, which allow to estimate habitat‐breadth for each population. We calculated habitat‐breadth across multiple populations of 154 species, and tested how habitat‐breadth varied with richness. We further tested the effect of traits and trait‐distinctiveness on the richness‐sensitivity of habitat‐breadth. Results Habitat‐breadth varied with species traits, with larger species more commonly habitat generalists while schooling and planktivorous species more commonly habitat specialists. Importantly, habitat‐breadth was negatively correlated with richness for 109 out of the 154 species, and, across all species, the relationship was highly significant. We found some support that species with distinct traits displayed less sensitivity of habitat‐breadth to richness, but the relationship was dependent on the type of trait‐distinctiveness index used. Conclusions This is the first large‐scale evidence that population‐level habitat‐breadth changes with richness. Results suggest that the realised niche is population‐specific and that niche breadth is reduced in high‐diversity settings where more intense interactions, such as competition, are expected. This implies that populations, specifically in species rich areas, do not use their entire fundamental niche. Therefore, the ability to predict habitat preferences response to global changes based on current habitat associations, without accounting for species interactions, may be limited.