作　　者： ; ; ; ; ;

机构地区： 中国水产科学研究院淡水渔业研究中心

出　　处： 《江西农业学报》 2010年第9期118-120,共3页

摘　　要： 对5口中华绒螯蟹养殖塘连续4个月测定pH、溶解氧、氨氮、亚硝酸盐、总氮、总磷、COD和酸性高锰酸钾指数,利用克鲁斯卡尔算法进行超度量聚类分析,并进行典型相关分析及典型冗余分析。结果显示:pH值和COD在水质的8个参数中起中心作用,以这两个变量为中心将8个变量分为两组,第一组为pH值、氨氮、溶解氧和总磷;第二组为COD、高锰酸钾指数、亚硝酸盐及总氮;第一典型相关系数为0.7221(P<0.05),相关显著;典型变量V1可以解释24.64%的组内变差,并解释12.84%的另一组的变差,典型变量W1可以解释40.23%的组内变差,并解释20.97%的另一组的变差;W1对pH值有30.49%的预测能力,V1对总氮、亚硝酸盐含量的预测能力分别为32.54%和38.72%。 Water quality indexes were detected in five Eriocheir sinensis ponds for four months,including pH-value,dissolved oxygen,ammonia,nitrite,total nitrogen,total phosphorus,COD and acid K2MnO4 index.Meanwhile,ultra-metric clustering analysis was done by Kruskal algorithm.Then canonical correlation analysis and canonical redundancy analysis were carried out.The results showed that pH-value and COD were in the center of the clustering tree,and the 8 water quality indexes could be divided into two groups,group 1 included pH,ammonia,dissolved oxygen and total phosphorus,the other group consisted of COD,acid K2MnO4 index,nitrite and total nitrogen.The first canonical correlation coefficient was 0.7221(P<0.05).Canonical variable V1 could explain 24.64% within-group variance and 12.84% variance of another group.Canonical variable W1 could explain 40.23% within-group variance and 20.97% variance of another group.W1 could predict 30.49% pH,and V1 could predict 32.54% total nitrogen and 38.72% nitrite content.

关 键 词： 中华绒螯蟹 水质 度量 聚类分析 典型相关分析 亚硝酸盐含量 高锰酸钾指数 组内变差 预测能力 典型变量 总氮 溶解氧 典型相关系数 总磷 中心作用 冗余分析

领　　域： [农业科学]