| 基于GAM和GWR模型分析环境因子对鱼类分布的影响 |
| |
| 引用本文： | 简盈,张云雷,宋业晖,张崇良,纪毓鹏,任一平.基于GAM和GWR模型分析环境因子对鱼类分布的影响[J].海洋学报,2022,44(7):103-111. |
| |
| 作者姓名： | 简盈 张云雷 宋业晖 张崇良 纪毓鹏 任一平 |
| |
| 作者单位： | 1.中国海洋大学 水产学院，山东 青岛 266003 |
| |
| 基金项目： | 国家重点研发计划（2019YFD0901204） |
| |
| 摘 要： | 多鳞鱚（Sillago sihama）是山东近海重要的渔业种类之一。本研究根据2016年秋季（10月）在山东近海开展渔业资源底拖网调查取得的数据，分析该海域多鳞鱚的空间分布特征，并运用广义可加模型（GAM）和地理加权回归（GWR）模型探究影响其分布的因素及其与环境因子的非线性和空间非平稳性关系。GAM拟合结果显示，影响秋季多鳞鱚分布的环境因子主要有水深、底层水温和底层盐度，水深的偏差解释率最大，为23.50%。GWR模型拟合结果显示，多鳞鱚分布与水深和底层水温之间存在空间非平稳性关系。水深与多鳞鱚相对资源量呈负相关关系，底层水温与多鳞鱚相对资源量呈正相关关系。赤池信息准则和决定系数（R2）指标对比结果显示，GWR模型的表现优于GAM，在渔业生态数据分析中表现出较好的发展潜力。本研究为今后开展渔业生物空间分布提供了一种新的方法。
|
| 关 键 词： | 多鳞鱚 地理加权回归模型 广义可加模型 空间非平稳性 空间分布 |
| 收稿时间： | 2021-12-21 |
Effect of environmental factors on fish distribution based on GAM and GWR model : A case study of Sillago sihama in the Shandong coastal waters |
| |
| Institution: | 1.Fisheries College, Ocean University of China, Qingdao 266003, China2.Laboratory for Marine Fisheries Science and Food Production Processes, Pilot National Laboratory for Marine Science and Technology (Qingdao), Qingdao 266237, China3.Field Observation and Research Station of Haizhou Bay Fishery Ecosystem, Ministry of Education, Qingdao 266003, China |
| |
| Abstract: | Sillago sihama is an important fishery species in China and plays an important role in the marine ecosystem of the Yellow Sea. Species distribution models can be used to predict its distribution by establishing the relationships between its abundance and environmental factors. However, due to high mobility of the marine animals, the relationship between their distribution and environmental factors is often nonlinear and variable with spatial locations. Based on data collected from bottom trawl survey in the Shandong coastal waters in autumn of 2016, both generalized additive model (GAM) and geographically weighted regression (GWR) model were used to analyze nonlinear and spatial nonstationary relationships between distribution of the species and environmental factors, and results from the two models were compared. Results from the GAM indicated that the main environmental factors were depth, sea bottom temperature and salinity, and depth had the largest deviance explained (23.50%). GWR model results showed that there were spatial non-stationary relationships between distribution of the species and depth and sea bottom temperature. GWR model results indicated a negative correlation between depth and biomass of the species, and a positive correlation between sea bottom temperature and biomass of species. Regarding performance of the models, GWR model showed advantages over GAM in identifying influencing factors and predicting distribution, and GWR model was recommended for use in similar applications. |
| |
| Keywords: | |
|
| 点击此处可从《海洋学报》浏览原始摘要信息 |
| 点击此处可从《海洋学报》下载免费的PDF全文 |
|
