An assessment of sampling biases across studies of diel activity patterns in marine ray-finned fishes (Actinopterygii)

Abstract

Understanding the promotion and regulation of circadian rhythms in marine fishes is important for studies spanning conservation, evolutionary biology, and physiology. Given numerous challenges inherent to quantifying behavioral activity across the full spectrum of marine environments and fish biodiversity, case studies offer a tractable means of gaining insights or forecasting broad patterns of diel activity. As these studies continue to accumulate, assessing whether, and to what extent, the cumulatively collected data are biased in terms of geography, habitat, or taxa represents a fundamentally important step in the development of a broad overview of circadian rhythms in marine fish. As such investigations require a phylogenetic framework, general trends in the phylogenetic sampling of marine fishes should be simultaneously assessed for biases in the sampling of taxa and trait data. Here, we compile diel activity data for more than 800 marine species from more than five decades of scientific studies to assess general patterns of bias. We found significant geographic biases that largely reflect a preference toward sampling warm tropical waters. Additionally, taxonomic biases likewise reflect a tendency toward conspicuous reef associated clades. Placing these data into a phylogenetic framework that includes all known marine fishes revealed significant under-dispersion of behavioral data and taxon sampling across the whole tree, with a few subclades exhibiting significant over-dispersion. In total, our study illuminates substantial gaps in our understanding of diel activity patterns and highlights significant sampling biases that have the potential to mislead evolutionary or ecological analyses.