Updated Bottomfish Stock Assessment Used to Project Alternatives for Total Allowable Catch in the Main Hawaiian Islands

During 1988-2007, the Hawaii bottomfish stock, considered across the entire archipelago, was not overfished (biomass as a fraction of BMSY was greater than 0.7), and only in 1989 was overfishing occurring (harvest rate as a fraction of HMSY was greater than 1.0).

PIFSC scientist Jon Brodziak recently updated the stock assessment for Hawaiian bottomfish using catch data through 2007. The Hawaiian bottomfish assessment was recently updated through 2007. The data sources and analytical approaches were similar to or identical to those used in the previous benchmark assessment. The update was conducted using re-audited bottomfish catch and effort data from Hawaii Division of Aquatic Resources (DAR) catch records collected during 1948-2007.

A revised bottomfish catch-per-unit effort (CPUE) data set was constructed for fishing trips using deep handline gear from the re-audited catch and effort data. A generalized linear model was applied to develop a spatially standardized CPUE series for the main Hawaiian Islands (MHI). The standardization model included fishing year, month, and fishing area as predictors of bottomfish CPUE on directed fishing trips (those with > 50% bottomfish catch by weight); all of the predictors had a highly significant (P<0.0001) influence on observed CPUE by trip. The spatially standardized CPUE time series was used for estimating parameters of the bottomfish production model.

A Bayesian production model incorporating the spatially standardized CPUE time series was used to estimate bottomfish biomass and harvest rate time series. The production model was fit to catch and standardized CPUE data for each of the three Hawaiian fishing zones: the MHI zone, the Mau zone, and the Hoomalu zone. The production model fitting incorporated uninformative priors for carrying capacity, process error, observation error, and catchability parameters and an informative prior for intrinsic growth rate.

In most recent years, bottomfish in the main Hawaiian Islands management unit have been overfished (biomass as a fraction of BMSY has been less than 0.7) and in all but one year since 1980 overfishing has been occurring (harvest rate as a fraction of HMSY has been greater than 1.0). Data from 1948 through 2007 were analyzed.

The model was also used to develop stochastic short-term projections of future total allowable catches and associated risks of overfishing. These projections explicitly included uncertainty in the posterior distribution of estimated bottomfish biomass in 2007 and population dynamics parameters. Assessment results for the archipelago-wide bottomfish stock indicated that the stock was not currently experiencing overfishing in 2007. In particular, the mean ratio of harvest rate in 2007 to the harvest rate to produce MSY was H2007/HMSY = 0.62 and the estimated probability of overfishing in 2007 was 0%. Assessment results also indicated that the archipelagic stock was not overfished in 2007. In particular, the mean ratio of biomass in 2007 to the biomass needed to produce MSY was B2007/BMSY = 1.13 and the estimated probability that B2007 exceeded BMSY was 97%. In contrast, results for the main Hawaiian Islands bottomfish management unit indicated that the stock was currently experiencing overfishing (H2007/HMSY=1.11) and was depleted (B2007/BMSY=0.62). In fishing year 2007, there was a 0% probability that MHI biomass exceeded BMSY and an 87% chance that the harvest rate exceeded HMSY.

Bottomfish biomass, catch, and relative biomass were projected for fishing years 2008-2010 under alternative fishing mortality and catch assumptions for the MHI. In these projections, status quo fishing mortality rates were assumed for the Mau and Hoomalu zones in 2008. Several constant catch projections were run to estimate the risk of overfishing in the MHI assuming a constant total allowable catch (TAC) was set and achieved for the MHI scenario in fishing year 2009-2010. Under this scenario, the set of median catches of MHI bottomfish in 2009 that would produce probabilities of overfishing in the MHI ranging from 0% to 100%, by 5% intervals, were calculated. These TACs were then applied to the MHI in 2009-2010 while status quo fishing mortality rates were assumed in the Mau and Hoomalu zones. In this case, projected probabilities of overfishing, relative biomasses, and probabilities of depletion of MHI bottomfish indicated what would likely occur if constant TACs were set and achieved in the MHI during 2009-2010.

For a range of projected total allowable bottomfish catches in the MHI (TACs, column 1), this table shows associated TAC for Deep-7 species (column 2), various probabilities of overfishing (columns 3-5), relative stock biomass in the MHI at the beginning of fishing year 2010 (column 6), and probability that bottomfish in the MHI management unit are depleted in fishing year 2010 (last column).

For more information contact: Jon Brodziak