中国水产科学Issue(4)：832-851,20.DOI:10.3724/SP.J.1118.2014.00832
基于分位数回归的库克群岛海域长鳍金枪鱼栖息环境综合指数
An integrated habitat index for albacore tuna (Thunnus alalunga) in the waters near the Cook Islands based on the quantile regression method
摘要
Abstract
We developed an“Integrated Habitat Index (IHI)”model based on the quantile regression method using sur-vey data collected at 43 sites in waters near the Cook Islands from September, 2012 through November, 2012. The model variables included vertical profile data for temperature, salinity, chlorophyll-a, horizontal current, vertical current and catch per unit effort (CPUE) of albacore tuna(Thunnus alalunga), and the interactions among these variables. Mod-els were developed for five 40 m water strata between 40 m and 240 m and the entire water column to predict the spatial distribution of albacore tuna. The environmental variables measured at modeling sites were used as inputs to the IHI models to predict the IHI value of the 5 strata and the entire water column. We tested for a significant difference be-tween the observed CPUE and predicted CPUE within the 5 water strata and the entire water column using a Wilcoxon test. The Spearman correlation coefficients were assumed to indicate the predictive power of the IHI model. The trend line of the arithmetic average about the predicted IHI for the 5 strata was compared with the CPUEs at the specific depth stratum. The environmental variables at validation sites were used to validate the model’s predictive power. These data were input into the CPUE models to predict the CPUE of the 5 water strata and the entire water column. We used a Wilcoxon test to compare between the predicted and observed CPUEs within the 5 water strata and the entire water column to validate the IHI results. The CPUE for albacore tuna in the 5 strata and the entire water column exhibited a skewed normal distribution with a longer left tail. There was no significant difference between the nominal CPUEs and predictive CPUEs of albacore in the 5 water strata and the entire water column at the modelling sites or the validation sites. The IHI models had good predictive power, and were able to accurately predict the distribution of albacore tuna in the 40-80 m and 160-200 m strata and in the entire water column. The key environmental parameters in the IHI models differed among the depth strata. In the shallow water (40-80 m), there was a close relationship between the CPUE and water color. Conversely, water temperature, sea currents, and bait and water temperature were the most significant vari-ables in the 80-120 m, 120-160 m, and 160-240 m strata, respectively. Albacore tuna prefer to feed at around 160-240 m, and generally between 160 and 200 m. The CPUE was 20.31 ind/(1000 hooks) in this water strata. The IHIs for al-bacore tuna in the area between 13°S-15°S, 162°W-167°W and 11°S-12°S, 161°W-167°W were relatively high. When fishing in these areas, our results suggest that fishing gear should be deployed at a depth of 160-240 m to increase CPUE and reduce by-catch.关键词
长鳍金枪鱼/栖息环境综合指数/分位数回归/库克群岛/捕捞效率Key words
Thunnus alalunga/integrated habitat index/quantile regression/Cook Islands/fishing efficiency/CPUE分类
农业科技引用本文复制引用
杨嘉樑,黄洪亮,宋利明,饶欣,吴越,齐广瑞..基于分位数回归的库克群岛海域长鳍金枪鱼栖息环境综合指数[J].中国水产科学,2014,(4):832-851,20.基金项目
国家高技术研究发展计划项目(2012AA092302) （2012AA092302）
农业部远洋渔业探捕项目(D8006128005) （D8006128005）
上海市教育委员会科研创新项目(12ZZ168) （12ZZ168）
高等学校博士学科点专项科研基金联合项目(20113104110004) （20113104110004）
