Abstract
Vertical walls of submarine canyons represent features of high conservation value that can provide natural areas of protection for vulnerable marine ecosystems under increasing anthropogenic pressure from deep-sea trawling. Wall assemblages are spatially heterogeneous, attributed to the high environmental heterogeneity over short spatial scales that is a typical feature of canyons. Effective management and conservation of these assemblages requires a deeper understanding of the processes that affect faunal distribution patterns. Canyons are recognised as sites of intensified hydrodynamic regimes, with focused internal tides enhancing near-bed currents, turbulent mixing and nepheloid layer production, which influence faunal distribution patterns. Faunal patterns also respond to broad-scale hydrodynamics and gradients in water mass properties (e.g. temperature, salinity, dissolved oxygen concentration). Oscillating internal tidal currents can advect such gradients, both vertically and horizontally along a canyon's walls. Here we take an interdisciplinary approach using biological, hydrodynamic and bathymetry-derived datasets to undertake a high-resolution analysis of a subset of wall assemblages within Whittard Canyon, North-East Atlantic. We investigate if, and to what extent, patterns in diversity and epibenthic assemblages on deep-sea canyon walls can be explained by spatial and temporal variability induced by internal tides. Vertical displacement of water mass properties by the internal tide was calculated from autonomous ocean glider and shipboard CTD observations. Spatial patterns in faunal assemblage structure were determined by cluster analysis and non-metric Multi-Dimensional Scaling plots. Canonical Redundancy Analysis and Generalised Linear Models were then used to explore relationships between faunal diversity and assemblage structure and a variety of environmental variables. Our results support the hypothesis that internal tides influence spatial heterogeneity in wall faunal diversity and assemblages by generating both spatial and temporal gradients in hydrodynamic properties and consequently likely food supply.