| 多变量分位数回归构建印度洋大眼金枪鱼栖息地指数 |
| |
| 引用本文： | 冯波,陈新军,许柳雄.多变量分位数回归构建印度洋大眼金枪鱼栖息地指数[J].广东海洋大学学报,2009,29(3):48-52. |
| |
| 作者姓名： | 冯波 陈新军 许柳雄 |
| |
| 作者单位： | 1. 上海海洋大学海洋科学学院,上海201306;广东海洋大学水产学院,湛江524025
2. 上海海洋大学海洋科学学院,上海201306;大洋生物资源开发和利用上海市高校重点实验室,上海201306;大洋渔业资源可持续开发省部共建教育部重点实验室,上海201306 |
| |
| 基金项目： | 国家科技支撑计划，国家高技术研究发展计划(863计划)，上海市捕捞学重点学科项目 |
| |
| 摘 要： | 以0～300m水层加权平均水温、50～150m水层的温差和氧差及其交互变量为影响因子,运用分位数回归法,寻找出环境变量与大眼金枪鱼(Thunnus obesus)延绳钓钓获率的最佳上界分位数回归方程,计算出栖息地指数(HSI),并应用地理信息系统(GIS)软件绘制各月HSI空间分布图。研究表明:大眼金枪鱼延绳钓钓获率(HR)依加权平均水温(x)、温差(y)、氧差(z)与的最佳上界分位数回归方程为HR0.70=-15.596+2.124x-0.003x3+0.033xyz-0.036y2z+0.107yz2-0.337z3;HSI空间分布为:16°S—10°N印度洋海域HSI高于0.7,HSI>0.8的海域随季节发生显著变化,马达加斯加外海至100°E、16°S—26°S海域常年存在一片HSI<0.4的区域,26°S—40°S海域的HSI介于0.4～0.5,40°S以南海域HSI<0.4,东非外海季节性地出现一片HSI<0.6的海域。利用多个环境变量的栖息地指数模型来预测分析大洋金枪鱼资源分布效果较好。
|
| 关 键 词： | 大眼金枪鱼 栖息地指数 分位数回归 多变量 印度洋 |
Multivariate Quantile Regression on Habitat Suitability Index of Thunnus obesus in the Indian Ocean |
| |
| FENG Bo,CHEN Xin-jun,XU Liu-xiong.Multivariate Quantile Regression on Habitat Suitability Index of Thunnus obesus in the Indian Ocean[J].Journal of Zhanjiang Ocean University,2009,29(3):48-52. |
| |
| Authors: | FENG Bo CHEN Xin-jun XU Liu-xiong |
| |
| Abstract: | Some important variables are selected to carry out quantile regression including weighted average water temperature in the 0～300 m water layer,temperature difference and dissolved oxygen difference between 50m and 150m water layer and their interactions. The optimum upper boundary equation is regressed for these variables against bigeye tuna Thunnus obesus longline hooking rate to compute habitat suitability index (HSI). The computation results are visualized by geographical information system. It shows that the optimum upper boundary equation of weighted average water temperature(x),temperature difference(y),dissolved oxygen difference (z) against bigeye tuna longline hooking rate(HR) is as follows:HR0.70=-15.596+2.124x-0.003x3+0.033xyz-0.036y2z +0.107yz2-0.337z3. The GIS maps illustrate that HSI is greater than 0.7 within 16°S—10°N in the Indian Ocean. Area significantly varies seasonally where HSI is greater than 0.8. HSI ranges from 0.4～0.5 between 26°S～40°S. There is an area of HSI less than 0.4 that exists within Madagascar to 100°E and 16°S～ 26°S throughout the year. Area of HSI less than 0.6 seasonally appears away off East Africa. It is found that the better results of resources distribution could be predicted by use of multi-environmental variables in the estimation of HSI. |
| |
| Keywords: | Thunnus obesus habitat suitability index quantile regression multivariate Indian Ocean |
| 本文献已被 维普 万方数据 等数据库收录！ |
|
