应用渔业体长分析方法ELEFAN和SLCA估算鱼类生长参数的研究

A STUDY OF ELEFAN AND SLCA FOR ESTIMATING GROWTH PARAMETERS

本文应用蒙特卡罗模拟方法,通过假设不同的生长参数、样本大小和白噪音水平,模拟出不同假设条件下的鱼类种群时间序列的体长频数分布,再应用体长分析方法ELEFAN(e-lectronic length-frequency analysis)与SLCA(Shepherd's length-composition analysis)估算模拟数据的生长参数K和L∞。用总生长特征φ估算求得生长参数的可靠性。模拟结果表明在大多数情况下ELEFAN比SLCA更能准确稳定地估计生长参数。生长速率和极限体长越大,φ估计值的精确度越高。当体长的白噪音水平小于20%时,估计结果较准确。当样本小于60时,样本越大估计结果越准确;当样本大于60时,样本增大对结果影响不大。应用本文的方法估计西非科特迪瓦的法荷水库的长臂虾(Macrobrachium vollenhovenii)真实渔业数据也得到了较好的结果。

Monte Carlo simulation was used to generate length-frequency data under different scenarios, such as different values of the growth rate parameter （K）, asymptotic length （L∞）, sample size and CV level. Two length-based methods ELEFAN and SLCA were applied to estimate K and L∞. The growth performance index（φ） was calculated for comparing different fits of K and L∞. Monte Carlo Simulation analysis showed that in most cases, ELEFAN performed better than SLCA. The higher the K and L∞ are, the more reliable the estimated φ is. When the white noise level is less than about 20%, the estimated parameters are more reliable. When the sample size is large（〉60）, the estimated result is more accurates a further increased sample size has little effect on the result. This method was also applied to the published data of fresh-water shrimp（Macrobrachium vollenhovenii） in west A frlca and reliable results were obtained.