A MISSING LINK between continental shelves and the deep sea: Addressing the overlooked role of land-detached submarine canyons

This ambitious project will enable a step change in understanding of the sporadic but large flows of sediment, climatically-important organic carbon, and pollutants through submarine canyons, which connect continental shelves worldwide to the deep-sea.

>9000 large submarine canyons occur on all the world's submerged margins, often dwarfing river systems in scale. Such canyons can transfer large quantities of natural sediments, organic carbon and nutrients that sustain important ecosystems, and are increasingly recognised as hotspots for seafloor pollution that threatens the biodiversity they host. The sediment flows that travel along canyons can be fast and dense, breaking cables that underpin global communications. It is therefore important to understand when and how such flows are triggered, the amount of material that is transported, and crucially, how these vary between types of canyon.

Monitoring of turbidity currents has focused on 'land-attached' canyons fed by rivers or long-shore drift, where powerful turbidity currents have been shown to effectively transport sediment and carbon over 1000s km. Despite accounting for >70% of canyons worldwide, land-detached canyons (that lie far from shore) remain un-monitored, exposing a major gap in our understanding of global particulate transport. This bias results from a long-held view that land-detached canyons are disconnected from sediment inputs during present day sea levels. New measurements in Whittard Canyon (in the Celtic Sea, 250 km from shore) challenge this paradigm, revealing that land-detached canyons can feature turbidity currents of similar frequency and power to major land-attached canyons. These surprising new results raise the following questions, and motivate our project, which aims to determine the mechanisms and fluxes of particulate transfer via land-detached submarine canyons to the deep-sea for the first time.

How can frequent turbidity currents occur if a canyon head lies far from present day sediment supplies? We will deploy an array of sensors on the continental shelf and within the Whittard Canyon head to measure the conditions before and coincident with turbidity currents, and will repeatedly map the seafloor to identify how and where sediment is transported to the canyon head. We will then make the first source to sink measurements along a land-detached canyon, and the second of any major deep-sea canyon worldwide, hence in itself this will represent a significant scientific milestone.

What is the nature, concentration and burial efficiency of organic carbon or pollutants, and how does this compare to land-attached canyons? We will analyse seafloor and sediment trap samples to determine what quantities of organic carbon and pollutants are transported along the canyon, and to what extent they remain effectively locked up in the seafloor as a result of burial. Phytoplankton blooms occur at the head of Whittard Canyon during spring and summer, when turbidity currents are most frequent, providing fresh (marine) organic carbon in a similar manner to how river floods convey fresh carbon to land-attached canyons. We also observe mobile litter accumulations so will test to what extent turbidity currents transport pollutants as well as organic carbon and how its distribution relates to seafloor biodiversity hotspots. As well as posing an ecological threat, pollutants such as microplastics may effectively act as 'tracers', evidencing contemporary canyon flows.

What volumes of natural and anthropogenic material are transferred via land-detached canyons? Global budgets exist for particulate transport to and across the ocean, but none include land-detached canyons. We will provide a first order calculation to assess the global significance of land-detached canyons, first assessing the contribution to deep sea transport across the Celtic Margin, and then up-scaling our results to determine what is missing from existing global budgets.