基于调查数据的东海小黄鱼资源变化模式及评价

为筛选表征小黄鱼(Larimichthys polyactis)种群时间变化的主要特征,并分析其变化模式,根据2000―2015年渔业资源调查数据,利用最小/最大自相关因子方法(min/max autocorrelation factors,MAF)将生物学、捕捞强度、空间分布等19个指标进行融合分析,探讨了小黄鱼种群的年际变化模式和资源现状。通过一阶变异值计算各指标与时间的相关性,并以1为时间相关性分组标准,共筛选出7个指标,分别为生殖群体的性比组成(Sexratio S)、纬度分布重心(YCG)、异速生长系数(BS)、铺展面积(SA)、资源密度(CPUES)、条件因子(Condition S)和75%体长。将筛选后的指标用于MAF分析,根据MAF主分量值,两个主分量的一阶滞变异值为0.16、0.19,表现出相对较弱的时间连续性。从MAF得分的位相年际变化上,2000―2015年期间主分量MAF1可划分为3个位相,分别为2000―2002年、2003―2012年、2013―2015年;从趋势变化上,可分为2个趋势,分别为2000―2007年和2010―2015年。主分量MAF2可以划分2个位相,分别为2000―2007年和2010—2014年;从趋势上,可划分为4个趋势,分别为2000―2003年、2004―2006年、2007―2012年和2013—2015年。对第一主分量贡献度最高的指标是生殖群体的纬度分布重心(–0.756)和异速生长系数(–0.609);对第二分量贡献度最高的指标有生殖群体的性比组成(0.590)、异速生长系数(0.539)和扩散面积(–0.606)。主分量MAF1得分的年际变化与纬度分布重心和异速生长系数负相关;MAF2得分与性比、异速生长系数正相关,而与铺展面积负相关。由上可知,MAF1和MAF2在时间尺度上的变化模式反映了各指标的共同变化方向。研究结果为进一步甄选识别种群变化的主要指标及评估资源变化趋势提供科学依据。

Evaluation of temporal changes of small yellow croaker stock status in East China Sea using trawl survey indices

The indicator-based approach to fish stock assessment uses many indicators that characterize different attributes of a fish stock in order to assess its status. We considered 19 biological indicators to characterize location, dispersion, traits, fishing and abundance for the small yellow croaker, the indicators were derived from the 16-years (2000-2015) series of bottom trawl surveys over the East China Sea. The one-lag variogram for each indicator was computed, scaled to the indicator variance and ranked, the indicators with highest continuity at lag-one were selected. Min/max autocorrelation factors (MAFs) were calculated for the period 2000-2015 to summarize the multiple time series, detect changes and identity which indicators were responsible for the detected change. According to the variogram results, seven of the 19 indicators exhibited a marked time correlation at the first lag of the variogram below one, including four biological traits (sex ratio, allometric growth coefficient, condition factor, and third quartile of fish length), two spatial indicators (gravity in latitude and spreading area), and one abundance indicator (biomass index). Then the seven selected indicators were used to calculate MAFs during 2000-2015. The first two MAFs had low one-lag variogram values, 0.16 and 0.19, respectively, which represented lower time continuity. The continuity index was also calculated for each of the seven indicators on the first two MAFs, and the four indicators (YCG, BS, SexratioS, and SA) with the highest continuity index were selected to represent the history of the stock. The observed trends of the multivariate time series are described through the MAFs scores. MAF1 divided 16 years into three regimes (2000-2002, 2003-2012, and 2013-2015), and two trends were observed. MAF1 increased from 2000 until 2007 and then decreased until 2015. Whereas MAF2, which was not monotonic and had very small discontinuities, was detailed in two regimes (2000-2012 and 2010-2014) with four trends. From 2000 to 2003, it was close to being flat. From 2003 to 2006, it decreased, and it increased from 2006 to 2012, after which it decreased until 2015. The indicators that contributed the most to MAF1 were YCG (-0.756) and BS (-0.609), and the indicators that contributed the most to MAF2 were SexratioS (0.590), BS (0.539), and SA (-0.606). MAF1 was negatively correlated to YCG and BS, whereas MAF2 was positively correlated to SexratioS and BS but negatively correlated to SA. Due to the different inter-annual variation of biological traits and spatial indicators, the MAFs also exhibited different temporal change patterns at the different time scales.