Central Coast Monitoring Framework -
First Nations in Canada play a large and growing role in monitoring and stewardship of natural resources. On the Central Coast of BC, major gaps in monitoring for spawner escapement, fisheries catches, and climate currently hinder co-management. There was therefore an outstanding need for the development of a strategic and collaborative framework for monitoring salmon populations, fisheries, and the environmental conditions that drive inter-annual variability in salmon returns.
In 2019, the Pacific Salmon Foundation convened the four Central Coast Nations, Fisheries and Oceans Canada, and regional salmon experts to develop the Central Coast Monitoring Framework, a strategic plan for First Nations-led monitoring grounded in the values and priorities of the Central Coast Nations. The monitoring actions identified through this process will be implemented over the coming years, building foundations for First Nations leadership of monitoring and decision making under newly ratified Fisheries Reconciliation Agreements.
Tagging sockeye at the Koeye weir
Computer vision tools for in-season management -
There is a timely opportunity to collaboratively develop and apply new tools towards effective salmon management. Terminal fisheries conducted in-river or at a river mouth offer the advantage of targeting a single population, eliminating risks associated with mixed-stock fisheries, and enabling in-season management to reconcile harvest goals and conservation objectives. Given the well-documented complexities of current mixed-stock management regimes, there have been increasing calls for a revival of traditional management systems, and terminal fisheries as a pathway towards greater food security and resilient fisheries-based livelihoods in coastal communities. Through the Reconciliation process some commercial fishing licenses are being transitioned to First Nations communities through a voluntary buyback program. At the same time, significant breakthroughs in the fields of deep learning, real-time processing and computer vision have led to the streamlining of many industrial processes. This historical inflection point provides an unprecedented opportunity to reimagine how salmon fisheries operate, leveraging these breakthrough technological and computing tools to a novel purpose, supporting the revitalization of locally-managed terminal fisheries.
Bridging the fields of computer science, fisheries management and population monitoring, we are working with the Heiltsuk Integrated Resource Management Department (HIRMD), Kitasoo Xai’Xais Stewardship Authority (KXSA), and Haida Fisheries to revive locally-managed terminal fisheries, creating computing tools that automate salmon enumeration and provide in-season estimates of escapement. Our computer science team will build a computer neural network-based deep-learning process to automate identification and enumeration of salmon from video and sonar data. By building off-grid infrastructure we will power camera-based enumeration at two locations on the Central Coast, and work with the Haida Nation to automate counts from an existing Adaptive Resolution Imaging Sonar (ARIS) unit being run on the Yakoun River. Data will be stored, processed and compressed on site, and uploaded on a daily or weekly basis to a shared data repository for processing on the SFU computing cluster, thereby providing in-season estimates of salmon-abundance to support adaptive co-management and conservation. These escapement data is being incorporated into community-driven fisheries management through a collaborative process that engages with community fishers and DFO to understand socio-cultural contexts, values and needs, and applies tools in quantitative stock-assessment and adaptive management to minimize risk and increase sustainable fishing opportunities.
Working with Indigenous community partners, we will co-create novel tools for local fishery management. With support from the BC Salmon Recovery and Innovation Fund this project will apply computer deep learning to power the resurgence of management systems that supported sustainable fisheries for millennia. By developing infrastructure that builds upon existing capacity and addresses community needs we will strengthen scientific foundations for sustainable salmon fisheries and power a transformation that can benefit fishing dependent communities across the province.
Koeye River salmon ecosystem study -
Since 2012 ,this collaborative project by QQs Projects Society, the Hakai Institute, the Heiltsuk Integrated Resource Management Department (HIRMD), and Salmon Watersheds lab at Simon Fraser University (SFU), has worked to build long-term monitoring of sockeye and coho salmon in the Koeye River, a major salmon bearing watershed on the Central Coast.
Using a weir, smolt trap, and a network of RFID antennas we are estimating the number of sockeye and coho smolts that go to sea each year, as well as the number of returning adult sockeye. Each year we trap and tag smolts with uniquely coded PIT tags which are redetected on in-river antennas when they return to spawn. These estimates of smolt-to-adult survival are shedding new light on the survival of wild salmon during their lengthy marine migrations. Marine survival is a key driver of variability in salmon returns, and has been previously unknown for wild sockeye and coho salmon on the Central Coast.
Every June we install a weir in the lower reaches of Koeye River, where we capture, tag and sample migrating adult sockeye salmon. These tagged fish make their way beyond the weir and are detected at upstream RFID antennas and resighted during repeated fall counts. Using redection and count data we are understanding the effect of climate on survival during spawning migrations, and estimating the escapement of adult sockeye.
These data are supporting co-management and the development of a Heiltsuk sockeye management plan. Our paper here
First Nations-led catch monitoring and mixed-stock fishery sampling
Indigenous Peoples of the Northern Pacific Rim have harvested salmon for subsistence and livelihoods for more than 10,000 years. Over the last 150 years, colonization and industrialization have radically transformed salmon fishing, displacing culturally-managed in-river fisheries and moving most harvest into coastal waters. Among the most profound transformations in management brought on by colonization was the shift to mixed-stock ocean fisheries, which gradually replaced Indigenous in-river salmon fisheries as the primary method of harvest. Managers have long recognized that mixed-stock harvest can undermine the sustainability of salmon fisheries, if smaller or less productive populations are harvested in fisheries targeting abundant stocks. In these instances, overharvest can threaten the long-term viability of wild salmon populations and fisheries. Management measures to limit the impacts of these mixed-stock fisheries may also be harmful to fishing communities if they reduce fishing opportunity for locally abundant populations.
On British Columbia's Central Coast, the total number of salmon harvested and the stock-composition of harvest is currently unquantified in many areas. The absence of reliable stock-specific estimates of catch in marine fisheries creates pervasive uncertainty and risks for wild salmon. Working with the Central Coast Nations, the Central Coast Indigenous Resource Alliance (CCIRA), Fisheries and Oceans Canada, Simon Fraser University, Coastal Rivers Conservancy, and the Pacific Salmon Foundation, we aim to dramatically improve genetic baselines used to quantify population-level harvest in mixed-stock fisheries, and develop First Nations-led monitoring and catch sampling programs that provide accurate annual estimates of stock-specific harvest in mixed-stock fisheries. These efforts will inform adaptive co-management that minimizes harvest impacts on at-risk stocks while providing fishing opportunities for healthy or enhanced populations.
Estimating conservation targets in data-poor sockeye populations -
The management of data-limited fisheries is a key challenge in British Columbia and around the world. Bayesian and hierarchical population model provide a powerful tool for integrating multiple data sources and borrowing information across populations with varying data quantity and quality. Since sockeye are primarily lake rearing, we used estimates of lake productivity as Bayesian priors on population carrying capacity to estimate stock-recruit relationships for 70 populations of sockeye on the North and Central Coast. This work leverages decades of lake and population monitoring by DFO in sockeye systems on the North and Central Coast to estimate preliminary conservation and management targets for sockeye populations that support subsistence fisheries for remote coastal communities.