Recommendations on the design of a multispecies benthic marine invertebrate dive survey program for stock monitoring

Abstract

DFO Science has designed a multispecies dive survey protocol to provide unbiased, coast wide monitoring of benthic invertebrate stocks (as may be required under the updated Fisheries Act) and associated habitat information for a suite of benthic marine invertebrate species (Green (Strongylocentrotus droebachiensis), Purple (S. purpuratus) and Red Sea Urchin (Mesocentrotus franciscanus), Geoduck (Panopea generosa), Giant Red Sea Cucumber (Apostichopus californicus), Northern Abalone (Haliotis kamtschatkana), and Sunflower Sea Star (Pycnopodia helianthoides)). Based on information available at this time, the proposed survey design can provide estimates of coast wide stock status for Red Sea Urchin and Giant Red Sea Cucumber, and relative abundance indices for Geoduck, Green Sea Urchin, Purple Sea Urchin, Northern Abalone and Sunflower Sea Star. The new protocol was tested through a series of pilot surveys conducted on a subset of areas of the BC coast each September from 2016-2021. Design of the pilot surveys was based on previous dive survey data and experience and demonstrated the practical feasibility of the protocol, while also gathering preliminary information to guide the following recommendations about the statistical design of the survey: Use the dive survey protocol described in Appendix 12.1 of the Research Document, including the quadrat skipping sampling scheme that is dependent on transect length, a target depth range of -2 m to 12.2 m relative to chart datum, and a maximum transect length of 125 m. Exclude sections of shoreline with fetch values lower than 20,000 m or higher than 2.52 million m when randomly placing transects to ensure sampling effort is focused on suitable habitat while also avoiding potentially unsafe areas for diving. Ensure surveys occur at the same time of year to avoid introducing seasonal variation to the data. Use the common (across species) coast wide standard deviation-to-mean ratio of density (animals/m2) equal to 1.27 (i.e., σ/μ = 1.27) derived from the pilot surveys in calculations to determine the initial target number of transects to be sampled. The number of transects to be sampled may change if the in situ standard deviation-to-mean ratio of density differs substantially from this coast wide estimate. Conduct at least 240 transects coast wide to estimate stock status to adequately manage the risks of making incorrect stock status determinations. This target number of transects is based on an Acceptance Sampling analysis and uses the current reference points for Red Sea Cucumber (RSU) (Lochead et al. 2019) and Giant Red Sea Cucumbers (Hajas et al. In press), predetermined risk tolerances, and field-derived estimates of variability from the pilot surveys. The target number of transects could change in the future as new data become available or if reference points, observed variability in the data, and/or risk tolerances change. Implement a two-stage, random sampling design that minimizes the time required to cover the entire BC coast and optimizes the efficient use of available resources. Ideally, the entire coast would be covered in a single year (roughly 42 days of ship time required). A realistic alternative would be to divide the coast in two or three sections and rotate through the sections over two or three years (requiring roughly 23 days or 18 days of ship time per year, respectively). Should the available resources (such as ship time, etc.) dictate a rotation longer than two years, then it is recommended that a panel design be considered. Continue to explore pre- or post-stratification variables to improve survey precision, as data become available. Although there is insufficient evidence from the present analyses to suggest that stratification by fetch, Sea Otter occupancy, substrate type, or depth improve survey precision at this time, the dynamic ecosystem occupied by these species implies that these or other variables could become relevant to the survey design in the future. Areas for future work were identified and include: incorporation of relevant information from species distribution and/or habitat suitability models as they become available, more explicit consideration of climate change impacts, feasibility of collecting environmental data (i.e., ocean climate data) on future dive surveys, improved understanding of larval dispersal (source/sink dynamics), Sea Otter range expansion, and inclusion of invertebrate density estimates from areas at non-diveable depths (i.e., areas with depths greater than approximately 18 m).