论文部分内容阅读
使用广义线性模型(GLM)和广义可加模型(GAM)对印度洋中国大眼金枪鱼渔业的单位捕捞努力量渔获量(CPUE)进行标准化。在CPUE标准化模型中,考虑了空间(经度与纬度)、时间(年与月)和环境(包括各深度温度、各深度盐度和海平面高度)等变量。结果表明,标准化CPUE和名义CPUE在时空分布上呈相似的趋势。年CPUE随时间呈现下降的趋势,高CPUE经常出现在42°E～60°E、85°E～90°E、15°S～5°S和10°N～15°N的区域内。GLM和GAM分析都显示出经度是影响CPUE最重要的变量,可分别解释17.3%和23.81%的变异;纬度、经度和纬度的交互效应、年份、381 m水层温度、317 m水层温度对CPUE的影响也是明显的。此研究中GLM模型比GAM模型更合适。
The catch per unit effort (CPUE) of the oceans-sized tuna fishery in the Indian Ocean was normalized using the Generalized Linear Model (GLM) and the Generalized Additive Model (GAM). In the CPUE standardized model, variables such as space (longitude and latitude), time (year and month), and environment (including depths, depths of salinity and sea level) are taken into account. The results show that the standardized CPUE and the nominal CPUE show a similar trend in space-time distribution. The annual CPUE tends to decrease with time. High CPUEs often occur in the region of 42 ° E ~ 60 ° E, 85 ° E ~ 90 ° E, 15 ° S ~ 5 ° S and 10 ° N ~ 15 ° N. Both GLM and GAM analyzes show that longitude is the most important variable affecting CPUE and can account for 17.3% and 23.81% of variance respectively; interaction effects of latitude, longitude and latitude, year, 381 m water layer temperature, 317 m water layer temperature pair The impact of CPUE is also obvious. The GLM model is more suitable than the GAM model in this study.