Risk Analysis of Plausible Incidental Exploitation Rates for the Pacific Sleeper Shark, a Data-Poor Species in the Gulf of Alaska

Abstract

Monte Carlo simulation was used to investigate the sustainability of incidental exploitation rates (U) for Pacific Sleeper Sharks Somniosus pacificus in the Gulf of Alaska (GOA) under status quo management. Monte Carlo simulations were implemented with a standard, length-based, age-structured model that was evaluated with forward projection. Given the paucity of relevant data, we investigated the sensitivity of simulation results to a range of assumptions about key model parameters by using 24 alternative model configurations, each simulated 1,000 times. The risk analysis results were most sensitive to Pacific Sleeper Shark U-values. The aggregate fraction of simulations ending in an overfished condition increased from 0% under the low-U scenario to 59% under the high-U scenario. Risk analysis results were also sensitive to the assumed shape of the length-based selectivity curve (asymptotic or dome shaped) but were less sensitive to the range of assumptions about other key model parameters, including maximum age and stock productivity. These results indicate that a priority for Pacific Sleeper Shark management is to reduce the uncertainty in U. This major uncertainty will be decreased by an observer program that is now in place to monitor the historically unobserved GOA Pacific Halibut Hippoglossus stenolepis fishery, which incidentally catches Pacific Sleeper Sharks.

Received March 19, 2015; accepted December 7, 2015 Published online May 16, 2016

Acknowledgments

This research was conceptualized during a National Oceanic and Atmospheric Administration (NOAA) Advanced Studies Program (D.L.C.) in collaboration with the NMFS-AFSC (M.F.S.) and the School of Fisheries and Ocean Sciences, University of Alaska Fairbanks (M.D.A.). Funding (D.L.C.) was provided by the NMFS Auke Bay Laboratory, the NMFS Pacific Islands Fisheries Science Center, and the NMFS Southeast Fisheries Science Center’s Panama City Laboratory. The implementation of steepness in the Beverton–Holt stock–recruitment relationship benefited from conversations with Liz Brooks, Jon Brodziak, and Enric Cortés (NOAA Fisheries). The implementation of catch at age and length benefited from conversations with Felipe Hurtado-Ferro (School of Aquatic and Fishery Sciences, University of Washington, Seattle) and from feedback obtained based on preliminary presentations of this work during a Center for the Advancement of Population Assessment Methodology selectivity workshop (La Jolla, California; March 2013). Beth Mathews (University of Alaska Southeast, Juneau) contributed to data obtained from Southeast Alaska. The manuscript benefited from comments made by two anonymous reviewers.