摘要
为了利用简化的二值型渔业数据研究物种空间分布与环境因子之间的关系。基于2017年四个季度山东半岛荣成天鹅湖的菲律宾蛤仔(Ruditapes philippinarum)实地调查数据,比较了广义加性模型(GAM)和广义线性模型(GLM)在处理二值型菲律宾蛤仔渔业数据的表现,并利用GAM模型研究了天鹅湖的菲律宾蛤仔空间分布与环境因子之间的关系。研究显示:GAM模型在偏差解释率、模型评估效果和预测性能方面均优于GLM模型;水深、总有机质含量和叶绿素a浓度显著影响菲律宾蛤仔的空间分布(P<0.05)。菲律宾蛤仔的出现概率随叶绿素a浓度增加逐渐上升,随总有机质含量和水深的增加逐渐下降。研究结果表明,二值型简化数据能够有效地揭示菲律宾蛤仔空间分布与环境因子的关系。
To make full use of the simplified binary fishery data in determining the relationship between species distribution and environmental factors,we used the presence-absence binary data of Ruditapes philippinarum collected from Swan Lake,Rongcheng,Shandong Peninsula,in four seasons(spring,summer,autumn and winter)of 2017,compared the performance of the generalized additive model(GAM)and generalized linear model(GLM)for the binary data of R.philippinarum,and determined the relationship between the spatial distribution of R.philippinarum and environmental factors with GAM model.The GAM model performed better than the GLM model in terms of deviation explanation,model evaluation and prediction accuracy.Water depth,total organic matter(TOM)and chlorophyll-a(Chl a)showed significant effects on the spatial distribution of R.philippinarum in Swan Lake(P<0.05).The occurrence probability of R.philippinarum increased with the increase of Chl a and decreased with the increases of TOM and water depth.These results indicated that the simplified binary fishery data can effectively reveal the relationship between the spatial distribution of R.philippinarum and environmental factors.
作者
董建宇
胡成业
王学锋
杨晓龙
张秀梅
Dong Jianyu;Hu Chengye;Wang Xuefeng;Yang Xiaolong;Zhang Xiumei(Fisheries College,Guangdong Ocean University,Zhanjiang 524088,China;Fisheries College,Zhejiang Ocean University,Zhoushan 316022,China)
出处
《中国海洋大学学报（自然科学版）》
CAS
CSCD
北大核心
2024年第4期51-59,共9页
Periodical of Ocean University of China
基金
国家重点研究发展计划“蓝色粮仓科技创新”重点专项(2019YFD0901303)资助。
