Nan-Jay Su

Sex-structured population dynamics of blue marlin Makaira nigricans in the Pacific Ocean

The population dynamics of the blue marlin Makaira nigricans stock in the Pacific Ocean were estimated for 1971–2011 using a fully integrated length-based, age-, and sex-structured model. Fishery-specific catch, size composition, and catch-per-unit of effort were used in the modeling as likelihood components. Estimated dynamics were consistent with a stock that is fully exploited and stable over the last several years. No significant trends in recruitment were noted; however, female blue marlin were estimated to make up a majority of the catch, and historical exploitation has disproportionately changed the age structure of females relative to males. This result is due to differences in assumed life history and estimated selectivity. Changes to important life history parameters that are responsible for the productivity of the stock would potentially change the interpretation of current stock status.

Incorporating habitat preference into the stock assessment and management of blue marlin (Makaira nigricans) in the Pacific Ocean

Stock assessments that include a spatial component or relate population dynamics to environmental conditions can be considered one way of implementing an ecosystem approach to fisheries. A spatially-structured population dynamics model that takes account of habitat preference is developed and then applied to Pacific blue marlin (Makaira nigricans), as they prefer certain habitats and migrate seasonally. The model is fitted to fishery catch-rate and size data, along with information on the relative density of the population over space derived from a habitat preference model fitted to oceanographic and biological variables. Results show that blue marlin are more abundant in tropical waters, and females account for most of the biomass. Assessments that allow for environmental factors, movement dynamics and sexual dimorphism indicate that this population is in an over-exploited state, with current spawning stock biomass below the level corresponding to maximum sustainable yield (SMSY) and current fishing mortality exceeding that needed to achieve MSY (FMSY). A risk analysis based on samples from a Bayesian posterior distribution suggests that the population will remain above SMSY after 20 years if exploitation rates are below the level corresponding to FMSY.

LENGTH-BASED ESTIMATES OF GROWTH AND NATURAL MORTALITY FOR BLUE MARLIN (Makaira nigricans) IN THE NORTHWEST PACIFIC OCEAN

Blue marlin is an important resource for commercial and recreational fisheries. However, several life history parameters needed for stock assessments of this species are poorly determined. Sex-specific catch-at-size data (eye fork length, EFL) for blue marlin were collected from the Taiwanese offshore longline fishery in the northwest Pacific Ocean, and analyzed using the MULTIFAN, a length-frequency analysis, to derive growth parameters. The von Bertalanffy growth parameters of blue marlin were estimated to be different between sexes. The best model for the females included 11 age-classes in the length-frequency data sets (asymptotic length L∞ = 312.5 cm EFL, growth coefficient k = 0.111 yr^(-1), and theoretical age at zero length t_0 = -2.42 yr), while 9 age-classes were identified for the males (L∞ = 232.8 cm EFL, k = 0.131 yr^(-1), and t_0 = -3.58 yr). Natural mortality rates, based on empirical equations, were estimated to be 0.258 yr^(-1) for the males and 0.213 yr^(-1) for the females, respectively. The estimates of growth parameters and mortality rates derived from this study could be used in stock assessments for blue marlin, and contribute towards the fisheries management of the species.

Standardized Catch Per Unit Effort of Albacore Tuna (Thunnus alalunga) for the Taiwanese Distant-water Longline Fishery in the North Atlantic Ocean, 1967-2011

The trends in nominal catch per unit effort (CPUE) can be influenced by many factors in addition to stock abundance, including the choice of fishing location, operating strategy, and target species. Consequently, CPUE data need to be ‘standardized’ to remove the impact of such factors, and then could be used as an index of relative abundance in fisheries stock assessments. The catch and effort data for albacore tuna caught by the Taiwanese distant-water longline fishery in the North Atlantic Ocean from 1967 to 2011 were standardized using alternative approaches, i.e., generalized linear models (GLMs) and the delta approach. Sensitivity was explored to whether targeting effects of other species are included in the analyses. Year and area explained the most of the deviance (78~86%, depending on the model configuration), and both of them were identified consistently among methods. However, the trends in standardized CPUE of albacore tuna were found to be relatively insensitive to the approach used for catch and effort standardization, including whether the targeting effects were included in the analyses. With a drop before 1970, the trend in standardized CPUE of albacore in the North Atlantic Ocean was relatively stable during the 1970s to the late 1980s, but generally decreased thereafter and increased slightly from 2003 until present. The standardized CPUE of albacore from GLM for this fishery could be considered the best available results to reflect the relative abundance of the albacore stock in the North Atlantic Ocean.

Errors in Coefficient or Expected Value? Effects of Different Methods on Simulated Values for a Linear Model and the Nonlinear Von Bertalanffy Growth Model

Monte Carlo simulation is widely applied to incorporate the uncertainties in fisheries assessment models. However, even modeling the same kind of uncertainty, different practices often occur among studies, which may lead to erroneous results. We demonstrate how simulation results differed among methods of incorporating uncertainty into simulation in different ways: adding random errors to model either coefficients or expected values. Using life history parameter data from sailfish (Istiophorus platypterus), natural mortality from Paulys empirical equation, and lengths-at-age from the von Bertalanffy growth model (VBGM) were simulated using different methods. Different simulation methods did not affect the averages of simulated values from Paulys empirical equation and had only slight effects on the simulated lengths-at-age from the VBGM. For both linear Paulys equation and nonlinear VBGM, the variances of the simulated values from the errors-in-coefficients methods were under-estimated, being approximately 1 to 7% or 40 to 95% of those from errors-in-expected-values methods, depending on whether the correlation among coefficients was included or not. Therefore, adding random errors with either an additive or multiplicative error structure to the expected values is preferred over the errors-incoefficient methods for fully representing uncertainty in the data.