2011～2012年东南太平洋智利竹鱼CPUE的时空变化及其与捕捞因子关系
|
摘要
根据2011和2012年东南太平洋智利竹鱼(Trachurus murphyi)的渔业生产统计数据,利用广义可加模型(GAM)分析了智利竹鱼资源的时空变动及其与捕捞因子的关系。结果表明:GAM模型对单位捕捞努力量渔获量(CPUE)总偏差解释率为24.173%,其中贡献最大的是拖网时间,贡献率为13.758%。东南太平洋智利竹鱼作业渔场主要集中在35.5°~46°S、80°~91°W和25°~28.2°S、74°~76.5°W范围内;最佳捕捞时间为4月、8~11月;最佳拖网时间为8~10 h;最佳的网位为0~40 m;最佳的网口高度为72~95 m。逐步建模法结果显示,影响智利竹鱼CPUE的因子按重要性程度从高到低依次为拖网时间、网位、网口高度、月份、纬度、经度、曳纲长度、网口水平扩张。
The Chilean jack mackerel,Trachurus murphyi,is a pelagic species widely distributed in southeast Pacific region,including waters of continental shelf and the high seas of Ecuador,Peru and Chile,extending from middle latitude to New Zealand waters. It is an important component in Pacific fishery resources and fishing production. Thus it is significant to study the relationship between spatiotemporal distribution of Trachurus murphyi and marine environmental factors,which will lay a foundation for resource change and fishery forecast. A lot of researches have been done,including relationship between Trachurus murphyi and a single environmental factor( such as Surface Sea Temperature,SST). The relationship between fishery and environmental factors is nonlinear and cannot be added. It is the generalized additive model( GAMs) that makes it possible to resolve the complicated relationship between fishery and many factors,and many references have explored it. However,it does not involve the impact of fishing on catch per unit effort( CPUE). Fishing factors such as fishing time( e. g. net time),fishing power and capacity( e. g. characteristics of fleet,fishing gear and equipment),spatial and temporal factors,can influence CPUE. So the CPUE needs to be standardized with impacts of relevant factors removed or minimized in fishery production. For these reasons,based on fishing data of Trachurus murphyi from 2011 to 2012,provided by Shanghai Deepsea Fisheries Corporation whose vessels fished beyond exclusive economic zone of Chile,this paper analyzed the relationship between spatiotemporal distribution of Trachurus murphyi and fishing factors. The daily data included vessel name,related start time and end time of each net,speed,net position,related start and end location in latitude and longitude,and tons of catch per net. For GAM analysis,temporal( month),spatial( longitude and latitude) and fishing factors( net time,net position,towing speed,vertical opening,warp length,horizontal spreading) were explanatory variables,and CPUE was response variable. F test was used to evaluate significance of factors,while Akaike Information Criterionis( AIC) was used to test the fitting degree of the model by adding factor step by step. Results showed that the rate of deviance explained about CPUE was 24. 173 %. The factor providing the largest contribution was net time( 13. 758%),followed by net position( 2. 668%),vertical opening( 2. 539%) and month( 2. 110%). F test showed that in addition to towing speed and horizontal spreading,other factors were significant( P < 0. 05). AIC test showed that residual deviance and AIC were significantly reduced with explanatory variables added to the GAM model one by one. But when warp length was added to the model,AIC increased. AIC was significantly reduced when horizontal spreading was added to the model. So two variables contributed to the model. However,towing speed had no contribution to the model with increasing AIC when it was added to the model. So we can conclude that fishing grounds of Trachurus murphyi in the ocean were mainly located in waters of 25°- 28. 2° S、74°- 76. 5° W and 35. 5°- 46° S、80°- 91° W. The optimal fishing months were April,August,September,October and November; the optimal net duration was 8- 10 hours; the optimal location of net was 0- 40 m; and the optimal vertical opening was 72- 95 m. The order of 8 variables affecting the CPUE of Trachurus murphyi in sequence of importance was net time,net position,vertical opening,month,latitude,longitude,warp length and horizontal spreading,respectively. Every factor had different influences on resource distribution of Trachurus murphyi in southeast Pacific. The impact of fishing factors on CPUE was more important than spatiotemporal factors.
The Chilean jack mackerel,Trachurus murphyi,is a pelagic species widely distributed in southeast Pacific region,including waters of continental shelf and the high seas of Ecuador,Peru and Chile,extending from middle latitude to New Zealand waters. It is an important component in Pacific fishery resources and fishing production. Thus it is significant to study the relationship between spatiotemporal distribution of Trachurus murphyi and marine environmental factors,which will lay a foundation for resource change and fishery forecast. A lot of researches have been done,including relationship between Trachurus murphyi and a single environmental factor( such as Surface Sea Temperature,SST). The relationship between fishery and environmental factors is nonlinear and cannot be added. It is the generalized additive model( GAMs) that makes it possible to resolve the complicated relationship between fishery and many factors,and many references have explored it. However,it does not involve the impact of fishing on catch per unit effort( CPUE). Fishing factors such as fishing time( e. g. net time),fishing power and capacity( e. g. characteristics of fleet,fishing gear and equipment),spatial and temporal factors,can influence CPUE. So the CPUE needs to be standardized with impacts of relevant factors removed or minimized in fishery production. For these reasons,based on fishing data of Trachurus murphyi from 2011 to 2012,provided by Shanghai Deepsea Fisheries Corporation whose vessels fished beyond exclusive economic zone of Chile,this paper analyzed the relationship between spatiotemporal distribution of Trachurus murphyi and fishing factors. The daily data included vessel name,related start time and end time of each net,speed,net position,related start and end location in latitude and longitude,and tons of catch per net. For GAM analysis,temporal( month),spatial( longitude and latitude) and fishing factors( net time,net position,towing speed,vertical opening,warp length,horizontal spreading) were explanatory variables,and CPUE was response variable. F test was used to evaluate significance of factors,while Akaike Information Criterionis( AIC) was used to test the fitting degree of the model by adding factor step by step. Results showed that the rate of deviance explained about CPUE was 24. 173 %. The factor providing the largest contribution was net time( 13. 758%),followed by net position( 2. 668%),vertical opening( 2. 539%) and month( 2. 110%). F test showed that in addition to towing speed and horizontal spreading,other factors were significant( P < 0. 05). AIC test showed that residual deviance and AIC were significantly reduced with explanatory variables added to the GAM model one by one. But when warp length was added to the model,AIC increased. AIC was significantly reduced when horizontal spreading was added to the model. So two variables contributed to the model. However,towing speed had no contribution to the model with increasing AIC when it was added to the model. So we can conclude that fishing grounds of Trachurus murphyi in the ocean were mainly located in waters of 25°- 28. 2° S、74°- 76. 5° W and 35. 5°- 46° S、80°- 91° W. The optimal fishing months were April,August,September,October and November; the optimal net duration was 8- 10 hours; the optimal location of net was 0- 40 m; and the optimal vertical opening was 72- 95 m. The order of 8 variables affecting the CPUE of Trachurus murphyi in sequence of importance was net time,net position,vertical opening,month,latitude,longitude,warp length and horizontal spreading,respectively. Every factor had different influences on resource distribution of Trachurus murphyi in southeast Pacific. The impact of fishing factors on CPUE was more important than spatiotemporal factors.
引文
[1]张衡,张胜龙,王雪辉,等.东南太平洋秋冬季智利竹鱼摄食习性初步分析[J].海洋渔业,2013,35(2):161-167.
[2]FAO Fisheries Department.Review of the state of world fishery resources:marines fisheries[R].Rome:FAO Fisheries circular No.920,1997:58-65.
[3]邵帼瑛,张敏.东南太平洋智利竹鱼中心渔场分布及其与海表温关系的研究[J].上海海洋大学学报,2006,15(4):468-472.
[4]牛明香,李显森,徐玉成.智利外海竹鱼中心渔场时空变动的初步研究[J].海洋科学,2009,33(11):105-109.
[5]ARCOS D F,CUBILLOS L A,NUNEZ S P.The jack mackerel fishery and EL Nino 1997-98 effects off Chile[J].Progress in Oceanography,2001(49):597-617.
[6]牛明香,李显森,徐玉成.基于广义可加模型的时空和环境因子对东南太平洋智利竹鱼渔场的影响[J].应用生态学报,2010,21(4):1049-1055.
[7]LI G,ZOU X R,CHEN X J,et al.Standardization of CPUE for Chilean Jack Mackerel(Trachurus murphyi)from Chinese trawl fleets in high seas of the southeast Pacific ocean[J].J.Ocean Univ.China(Oceanic and Coastal Sea Research),2013,12(3):441-451.
[8]STENSETH N C,MYSTERUD A,OTTERSEN G,et al.Ecological effects of climate fluctuations[J].Science,2002,297(5585):1292-1296.
[9]DINGSTOR G E,CIANNELLI L,CHAN K S,et al.Density dependence and density independence during the early life stages of four marine fish stocks[J].Ecology,2007(88):625-634.
[10]CIANNELLI L,CHAN K S,BAILEY K M,et al.Nonadditive effects of the environment on the survival of a large marine fish population[J].Ecology,2004(85):3418-3427.
[11]詹秉义.渔业资源评估[M].北京:中国农业出版社,1995:60-64.
[12]GUISAN A,EDWARDS T C,HASTIE T J.Generalized linear and generalized additive models in studies of species distributions:setting the scene[J].Ecological Modelling,2002(157):89-100.
[13]BURNHAM K P,ANDERSON D R.Model Selection and multimodel Inference:a practical informationtheoretic approach[M].2ndEd.New York:Springer,2002.
[14]邵帼瑛.基于神经网络的东南太平洋公海智利竹鱼渔情预报的研究[D].上海:上海海洋大学,2007.
[15]刘慧.南太平洋海洋环境因子研究及对智利竹鱼渔场的应用分析[D].上海:上海海洋大学,2013.
[16]化成君,张衡,樊伟.东南太平洋智利竹鱼资源和渔场的时空变化[J].生态学报,2011,31(19):5676-5681.
[17]邹晓荣.东南太平洋智利竹鱼资源、渔场和捕捞技术的研究[D].上海:上海海洋大学,2003.
[18]张衡,张胜茂.东南太平洋智利竹鱼渔场及单位捕捞努力量的时空分布[J].生态学杂志,2011,30(6):1142-1146.
[19]黄洪亮,陈雪忠.东南太平洋智利竹鱼中层拖网捕捞技术[J].中国水产科学,2005,12(1):99-103.
[20]曹铭昌,周广胜,翁恩生.广义模型及分类回归树在物种分布模拟中的应用与比较[J].生态学报,2005,25(8):2031-2040.
[21]温仲明,赫晓慧,焦峰,等.延河流域本氏针茅(Stipa bungeana)分布预测——广义相加模型及其应用[J].生态学报,2008(1):192-201.
[2]FAO Fisheries Department.Review of the state of world fishery resources:marines fisheries[R].Rome:FAO Fisheries circular No.920,1997:58-65.
[3]邵帼瑛,张敏.东南太平洋智利竹鱼中心渔场分布及其与海表温关系的研究[J].上海海洋大学学报,2006,15(4):468-472.
[4]牛明香,李显森,徐玉成.智利外海竹鱼中心渔场时空变动的初步研究[J].海洋科学,2009,33(11):105-109.
[5]ARCOS D F,CUBILLOS L A,NUNEZ S P.The jack mackerel fishery and EL Nino 1997-98 effects off Chile[J].Progress in Oceanography,2001(49):597-617.
[6]牛明香,李显森,徐玉成.基于广义可加模型的时空和环境因子对东南太平洋智利竹鱼渔场的影响[J].应用生态学报,2010,21(4):1049-1055.
[7]LI G,ZOU X R,CHEN X J,et al.Standardization of CPUE for Chilean Jack Mackerel(Trachurus murphyi)from Chinese trawl fleets in high seas of the southeast Pacific ocean[J].J.Ocean Univ.China(Oceanic and Coastal Sea Research),2013,12(3):441-451.
[8]STENSETH N C,MYSTERUD A,OTTERSEN G,et al.Ecological effects of climate fluctuations[J].Science,2002,297(5585):1292-1296.
[9]DINGSTOR G E,CIANNELLI L,CHAN K S,et al.Density dependence and density independence during the early life stages of four marine fish stocks[J].Ecology,2007(88):625-634.
[10]CIANNELLI L,CHAN K S,BAILEY K M,et al.Nonadditive effects of the environment on the survival of a large marine fish population[J].Ecology,2004(85):3418-3427.
[11]詹秉义.渔业资源评估[M].北京:中国农业出版社,1995:60-64.
[12]GUISAN A,EDWARDS T C,HASTIE T J.Generalized linear and generalized additive models in studies of species distributions:setting the scene[J].Ecological Modelling,2002(157):89-100.
[13]BURNHAM K P,ANDERSON D R.Model Selection and multimodel Inference:a practical informationtheoretic approach[M].2ndEd.New York:Springer,2002.
[14]邵帼瑛.基于神经网络的东南太平洋公海智利竹鱼渔情预报的研究[D].上海:上海海洋大学,2007.
[15]刘慧.南太平洋海洋环境因子研究及对智利竹鱼渔场的应用分析[D].上海:上海海洋大学,2013.
[16]化成君,张衡,樊伟.东南太平洋智利竹鱼资源和渔场的时空变化[J].生态学报,2011,31(19):5676-5681.
[17]邹晓荣.东南太平洋智利竹鱼资源、渔场和捕捞技术的研究[D].上海:上海海洋大学,2003.
[18]张衡,张胜茂.东南太平洋智利竹鱼渔场及单位捕捞努力量的时空分布[J].生态学杂志,2011,30(6):1142-1146.
[19]黄洪亮,陈雪忠.东南太平洋智利竹鱼中层拖网捕捞技术[J].中国水产科学,2005,12(1):99-103.
[20]曹铭昌,周广胜,翁恩生.广义模型及分类回归树在物种分布模拟中的应用与比较[J].生态学报,2005,25(8):2031-2040.
[21]温仲明,赫晓慧,焦峰,等.延河流域本氏针茅(Stipa bungeana)分布预测——广义相加模型及其应用[J].生态学报,2008(1):192-201.