三疣梭子蟹4个野生群体形态差异分析

Analysis of morphological variations among four wild populations of Portunus trituberculatus

通过对鸭绿江口、莱州湾、海州湾、舟山4个野生群体的三疣梭子蟹(Portunus trituberculatus)14项生物学外部形态性状的测量,采用多元分析方法对其进行比较分析。聚类分析结果表明,海州湾、舟山2个群体形态最为接近,鸭绿江口和莱州湾2个群体形态较为接近。对样本所属群体进行判别分析,4个群体雌蟹的综合判别率为87.0%,其中海州湾最低,为80.0%;莱州湾最高为92%。4个群体雄蟹的综合判别率为80.8%,其中海州湾最低,为72.0%;鸭绿江口最高,为96%。并建立4个群体各自的判别函数公式。通过对16项形态比例参数进行单因素方差分析,找到了4个群体间存在显著差异的形态比例参数。3种分析结果均认为,4个群体三疣梭子蟹在形态上已产生一定程度的差异,但所有的差异均未达到亚种水平,需要多项参数综合判别才能将4个群体的三疣梭子蟹区分开。

Portunus triuberbuculatus is an important economical species in China. In recent years, the wild stocks of P. triuberbuculatus along the coast of China have suffered fatal decrease in their resources due to overfishing and cultivation. So it is very important to analyze the morphological diversity of P. triuberbucu- latus. In October 2005,50 individuals were collected individually from four wild geographic populations distributed in the Yalu River estuary, Laizhou Bay, Zhoushan, and Haizhou Bay. In this research, cluster analysis,discriminant analysis and One-Way ANOVA analysis were applied in the analytical process. Based on 14 morphological characters of each population of P. trituberculatus from four locations, cluster analysis and discriminant analysis were conducted. Crabs from Haizhou Bay and Zhoushan were pooled into one group while crabs from Yalu River estuary and Laizhou Bay were pooled into another group by cluster analysis. In discriminant analysis, total accuracy rate of discriminant on female crabs from four locations was 87.0 %. The accuracy rate of Haizhou Bay population was the lowest（80.0 % ）, while that of Laizhou Bay population was the highest （92%）. Total accuracy rate of discriminant on male crabs from four locations was 80.8%. Haizhou Bay population was the lowest at 72.0 %, while Yalu River estuary population was the highest at 96 %. The discfiminant functions of the populations from four locations were established, respectively. OneWay ANOVA was conducted on 16 proportional parameters. The results showed that some proportional parameters were significantly different among the four populations. In this study all results supported that there were some morphological differences among the four populations, but variations of all the morphological traits were not characterized subspecies. Some integrative analysis was needed to distinguish accurately all geographic populations.