Spatiotemporal cross-validation of three survey methods for improved abundance indices of red king crab (Paralithodes camtschaticus)

Sustainable fisheries require input data of sufficient quantity and quality to inform stock assessments and management. However, capacity and resources to collect data are limited, underlining the need for methods that can maximize the utility of available fisheries-independent and -dependent data. The goal is to improve estimates of stock size and uncertainty, even from time series with inconsistent coverage and gear changes. This is of particular concern for the many stocks that are defined as datalimited, especially many shellfish stocks that are often not sufficiently covered by regular scientific surveys due to poor catchability in standard bottom trawls or habitat mismatch. This creates the need for methods that improve the coverage of such species by extracting and integrating information from available data. Additionally, the integration of data from different sources could prevent disjointed and potentially contradicting indices, especially when new survey methods are implemented parallel to existing surveys. Furthermore, combining already available data from surveys using differing gear types represents a cost-effective use of data and reduces the need for specialized surveys for target species. However, ahead of any consolidation of data from different surveys or gear types, information from different gears needs to be compared and, if possible, calibrated against each other. The red king crab survey in northern Norway provides an experimental setup with nearby stations to assess the catchability of bottom trawl, pots, and video. How well bottom trawls catch king crab is uncertain, and direct observations from video-based surveys and pots comparable to those used in commercial fishing may be more suitable approaches to assess changes in abundance. Interest in less intrusive survey methods has been heightened by concern for habitats that are vulnerable to damage caused by destructive sampling devices such as trawls. Visual surveys have emerged as a promising method for surveying benthic resources. We will present how observed densities and catch compositions of red king crab compare across survey gear, and how information from different gear types can be combined in an integrated estimation of abundance indices. We find common trends across gear types but also highlight challenges linked to substantial spatial variation in a highly structured fjord ecosystem. The results of this case study contribute novel insights into the integration of different data sources and gear types into coherent estimates of abundance indices and their uncertainty to facilitate implementation of new survey methods.