The Study On The Performances Of Length-frequency Analysis Methods On The Simulated And Real Fishery Data Sets

ELEFAN (electronic length-frequency analysis) and SLCA (Shepherd's length-composition analysis) are the most popular length-frequency analysis methods in fisheries stock assessment. Monte Carlo technique is used to evaluate the two methods under simulated and real fishery scenarios. Because the growth rate parameter (K) and asymptotic length (L∞) tend to be highly and negatively correlated, the growth performance index (φ) is calculated for comparing different fits of K and L∞.The length frequency plots under 12 different simulated scenarios with different biological and fishing characteristic indicate that, the higher the K and L∞are, the lower the overlap index is. If there is a wide separation of the age groups, the length frequency analysis can give a good estimate.Monte Carlo simulation is used to generate length-frequency data under 12 different simulated scenarios. ELEFAN and SLCA are applied to estimate K and L∞of VBGF(von Bertalanffy growth function) from the simulated frequency data. Monte Carlo simulation analysis show that the higher the K and L∞are, the more reliable the estimatedφis. This result is consistent with the simulated length frequency plot. When white noises of data are less than about 30%, the estimated parameters are mostly reliable. When the sample size is large (>60), the estimated result is more...