The global oceans currently absorb ~30% of anthropogenic CO2 emissions. The carbon cycle that regulates this ocean-atmosphere CO2 exchange, and the associated vertical distribution of dissolved carbon and alkalinity that influences the ocean's absorption capacity, depends on several processes. These are described as a series of interacting "pumps": a physical/chemical solubility pump; a biological 'soft tissue' pump; and a calcium carbonate pump. Understanding these three pumps, how they interact, and their atmospheric CO2 feedbacks is especially critical for accurate predictions of how the marine carbon cycle and global climate will change in the future. Calcium carbonate is a white, chalky mineral produced by a range of marine organisms. Importantly, when it dissolves it increases the alkalinity of seawater, which can reduce the seawater CO2 concentration below atmospheric CO2 levels and 'suck' anthropogenic CO2 from the atmosphere. Knowing exactly where it dissolves (how near the ocean surface) is therefore key to understanding the role this calcium carbonate pump plays in regulating ocean chemistry and atmospheric CO2. The operation of the calcium carbonate pump not only depends on the production rate but also the types of carbonate minerals that are produced by marine organisms, the rate at which they sink, and how rapidly these carbonate minerals then dissolve. Most ocean carbon cycle models make the assumption that carbonate production is dominated by the plankton and coccolithophores (microscopic algae). However, we now know that very large amounts of carbonate are excreted by marine bony fish (teleosts). This carbonate, which we now also know is mineralogically diverse depending on the fish species, is continuously produced in the intestines of fish and excreted as waste. The potential significance of this process to the marine CaCO3 pump was recognised in an initial modelling exercise led by PI Wilson (Science, 2009) which conservatively suggested that fish may account for at least 3-15% of total marine CaCO3 production globally, and realistically as much as 45%. Since that first modelling exercise the science behind this process has advanced hugely. As a group (and through the work of others) we now know that fish produce a hugely diverse range of carbonate mineral types, which existing knowledge would suggest should dissolve at very different rates. As a result, the assumptions in the first modelling efforts that fish produce uniform and relatively soluble carbonate types are no longer valid. Whilst we can already address some of the knowledge gaps, there is little or no data for fish from families that comprise ~94% of global fish biomass - including almost no data for mesopelagic fish that alone account for at least 60% of fish biomass. The daily vertical migration of their immense biomass is hypothesised to drive a novel "upward alkalinity pump", which may provide an important offset to the downward transport of alkalinity driven by other established processes. Also, we now have good evidence to show that production rates by fish vary with metabolic rate (which is greatest in the globally significant active epipelagic fishes), and importantly also depending upon feeding and diet (especially the calcium content of the diet). Thus, again, necessary assumptions in early models that all fish produce carbonate at the same rate are no longer realistic to use for modelling. Over and above these issues we also have little to no data on the rates at which these carbonates sink in the oceans or dissolve. The aim of this project is therefore to deliver new empirical data on fish carbonate production, mineralogies, solubilities and sinking rates to inform the first spatially- and mineralogically-resolved global production estimates, thus enabling us to parameterise models assessing fish contributions to the marine carbon cycle both under present day conditions, and for climate change scenarios in the future.

Aquatic ecosystems provide a variety of critical goods and services, including fisheries, energy and coastal protection. However, these fragile ecosystems are increasingly threatened by interacting stressors such as climate change, pollutants and overfishing. To protect key ecosystem services and ensure food security in an increasingly unpredictable climate, there is a pressing need to improve our understanding of fish habitat requirements and their vulnerability to different stressors. In this programme, we are using an integrated systems approach to address these knowledge gaps, providing novel solutions for the sustainable management of aquatic ecosystems in the 21st Century. To deliver sustainable fisheries management we need to understand fish connectivity patterns across species and life stages. Our knowledge of fish movements and habitat use has been significantly enhanced by the advent of electronic archival tags, but sample sizes can be limited by high costs, and the tags are typically restricted to larger-bodied species and life stages. Accordingly, we often know very little about juvenile life stages and the nursery habitats supporting our commercial fisheries. Luckily, all animals are equipped with their own intrinsic 'sensors' that record a wealth of information about the internal and external environment as they grow. By interrogating biogeochemical tracers in incrementally-grown tissues such as fish ear stones and eye lenses, we will read their 'life stories' and gain unique insights into their past health and habitat use. These tissues often exhibit growth bands ('biochronologies'), analogous to tree rings, which allow us to look back in time to reconstruct the fish's age, growth and movement timings, providing exciting opportunities to explore latent and cumulative effects of stressors on their physiology and health. Despite clear opportunities for archival tags, chemical tracers and biochronologies to cross-validate and augment each other, they are rarely used in combination. This programme aims to demonstrate how to integrate and scale these tools to support effective ecosystem management. This global programme involves a series of case studies that integrate emerging technologies (e.g. electronic archival tags and machine learning), novel chemical tracers and modelling to quantify and predict the movements and fitness of key fish species over a broad range of global change scenarios. We are using the North Sea as a model system to explore how integrated fish health and connectivity monitoring could enhance marine spatial planning. By pairing otolith chemistry and archival tag data from the same fish, we will determine optimal methods for reconstructing individual movement patterns using otolith tracers. To shed light on the mechanisms driving interannual variability in fisheries performance, we will develop and validate new tools for reconstructing fish health and contaminant exposure history. To reveal the critical habitats, dietary sources and fine-scale movement patterns of vulnerable salmonids we will use isotopic maps ('isoscapes') and tracers combined with novel machine learning methods. To quantify the latent and cumulative effects of hypoxia on fish size and fitness we will combine archival tag records and chemical tracers, then predict the impact of this growing environmental issue on fisheries productivity and stability. The overall synthesis of these case studies will provide new insights into fish habitat needs and their vulnerabilities to interacting stressors, improving our ability to predict fishery responses to differing global change and management scenarios. Finally, to promote multidisciplinary innovation and the integration of these emerging tools into mainstream resource management, we will establish an International Consortium dedicated to the EXploration, TRanslation and Application of Chemical records in fish Tissues (EXTRACT).