【作者】 官文江；

【导师】 李先华；

【作者基本信息】 华东师范大学 ， 地图学与地理信息系统， 2008， 博士

【摘要】 近几十年来,我国近海底层渔业资源出现衰退,但属中上层鱼类的鲐鱼资源较为丰富,其产量维持在40多万吨,占我国海洋捕捞产量的3%左右,在我国海洋捕捞渔业中占有重要地位,而东、黄海鲐鱼产量约占全国鲐鱼产量的78%。我国鲐鱼产量在波动中有所上升,但鲐鱼等中上层鱼类更易受到海洋环境影响,加之受到强大的捕捞压力,鲐鱼资源前景不容乐观,对鲐鱼资源及渔场动态的研究日益得到重视。与此同时,我国海洋遥感技术得到快速发展:我国已发射两颗海洋一号水色卫星,2009年将发射海洋二号动力环境卫星,2012年还将发射海洋三号海洋监视监测卫星,海洋信息获取能力得到极大提高,海洋遥感在海洋渔业的研究与应用日益受到我国海洋遥感与海洋渔业学者的关注。本文以中上层鱼鲐鱼为案例,探讨海洋遥感在近海渔业中的应用潜力及存在的问题,进一步拓展鲐鱼渔场及资源变动研究,为鲐鱼渔业管理提供理论支持。论文分为四个部分,第一部分概述了海洋遥感在海洋渔业的应用与进展,东、黄海鲐鱼的生物学、渔场与资源研究以及中国及周边国家与地区的鲐鱼围网渔业;第二部分介绍了东、黄海海洋环境与东、黄海海洋遥感监测研究及遥感数据源;第三部分为渔业资料的处理与分析,第四部分是遥感在鲐鱼渔场与资源动态变化中的应用。研究的主要结论如下:(1)根据1998-2003年东、黄海国营大型鲐鲹鱼灯光围网生产统计数据,利用一般线性模型对其捕捞效率进行了估算,讨论了线性模型误差结构的选择及数据变换对结果的影响。根据鲐鱼单位捕捞努力量渔获量(CPUE)数据呈正偏,以及CPUE均值与方差在对数尺度下的线性关系,选择了负二项分布、伽马分布与对数正态分布作为误差分布。研究表明,由于CPUE零值的存在,其对误差分布结构有很大影响。当采用最大似然估计时,对数正态分布与伽马分布的CPUE需加一常数(δ),δ取值对结果有较大影响,随δ增大将使估计的捕捞效率对比度得到压缩。为了避免δ取值的影响,同时采用了Delta-GLM方法。通过不同模型的比较,选择了Delta-负二项或Delta-伽马方法。根据Delta-负二项或Delta-伽马方法估算结果,各渔业公司间的捕捞效率存在很大的差异,且具有明显的区域性。32°N以北海区,捕捞效率高低依次为苏渔、辽渔、青渔、舟渔、宁渔、沪渔;台湾东北部海区,捕捞效率高低依次为苏渔、辽渔、舟渔、沪渔、宁渔、青渔。(2)研究了商业捕捞数据CPUE与资源量的关系。对此问题从两个方面进行了探讨:一是构造了一个集成了鱼类资源增长、渔船捕捞能力及鱼群与渔船间相互作用的元胞自动机模型,以探讨鱼群探捕与渔船作业可能对商业性CPUE与资源量之间关系的影响。文中分别模拟了①鱼群集中或随机分布,渔船随机分布;②鱼群由随机逐渐集中,渔船通过记录历史捕捞产量也逐渐集中分布,同时规定每艘船的最大捕捞量;③鱼群由分散到集中再到分散,而渔船集中在鱼群分布概率最大区;④渔船从随机分布到逐渐集中分布,鱼群集中分布不变等四种渔业上客观存在的情形。在渔船随机分布的情形下,不管鱼群如何分布,商业性CPUE与资源量均呈线性关系;在鱼群分布逐渐集中、渔船由于渔民经验积累也随之逐渐集中的情形下,商业性CPUE与资源量能表现出高稳性和高贫化性的特点。由于鱼群的集散或渔船进入渔区的时间长短不一,会造成商业性CPUE与资源量负相关的现象。上述模拟情况说明,在渔业资源评估中需要关注模型应用的前提条件以及模型的完善。文中还探讨了元胞自动机在渔业资源评估中应用的可行性。二是利用理想自由分布理论探讨了鲐鱼捕捞努力量的空间分布特点对CPUE与渔业资源密度关系的影响。从结果可知,在鲐鱼围网渔业中,随着捕捞努力量的集中,渔船之间存在一定的相互干扰,从而使得CPUE不一定随资源量的增加而增加,CPUE有被均匀化的趋势,捕捞努力量可能更能表达资源量的多少。上述结果表明CPUE受到众多因数的影响,如作业方式、努力量的空间分配(捕捞努力量直接指向高密度渔区)、努力量密度等,这使得CPUE与资源密度的关系难以保持线性。(3)根据1999-2003年鲐鱼捕捞生产数据,分析了东、黄海鲐鱼渔场的配置与东、黄海遥感获取的水温、叶绿素等环境要素的关系。结果表明,在北部渔场,黄海暖流对鲐鱼渔场的分布与变动有非常明显的影响,在冬季,鲐鱼主要分布于黄海暖流流轴边缘暖水侧,尽管叶绿素浓度数据混有泥沙等信息,但叶绿素浓度对鲐鱼的空间分布具有限制性作用;在南部渔场,上升流边缘偏暖水侧是鲐鱼渔场的分布区,鲐鱼渔场的分布在台湾暖流控制的区域,叶绿素浓度在该海域与渔场的分布关系不大。利用GAM模型进一步分析了各环境要素与CPUE的定量关系发现:在北渔场,海表水温、海面高度距平、海表温度梯度、海面风速与CPUE的关系密切;而在南渔场,海表水温距平、海面高度距平、海表风速、涡动能与CPUE的关系密切;两渔场叶绿素浓度对CPUE的影响均不显著。鲐鱼渔场的形成具有时空马尔科夫过程特征,利用GAM模型难以分析这种时空配置对渔场形成的影响,如北渔场,CPUE与水温的关系就隐含了特定时间、空间位置等条件,而不能单纯理解为温度单要素的作用,用该模型进行预测有可能会得出错误的结论。为进一步探索海洋环境与鱼群空间分布、演化的关系,本文利用遥感水温、叶绿素浓度与地形数据,构造了一个随机元胞自动机模型模拟黄海鲐鱼的空间分布的动态演化,但该模型需进一步发展。(4)鲐鱼资源的变动同海洋环境有密切的关系,研究海洋环境对鲐鱼资源的影响对制定合理的鲐鱼渔业管理计划至关重要。本文利用东海外海国营大型鲐鳕鱼灯光围网捕捞数据及TRMM/TMI遥感水温数据,分析了鲐鱼资源变动同水温的关系。利用主成分分析方法对遥感水温时间系列影像进行分析,得出鲐鱼资源受东海环境短周期变化的影响而表现出相应资源波动,黑潮及台湾暖流与东海外海鲐鱼资源变动关系密切,产卵场产卵期的温度升高有利于鲐鱼产量的提高,水温影响鲐鱼资源在时空分布上的调整。文章进一步指出,利用商业捕捞数据研究渔业资源变动,CPUE的标准化非常重要,渔业资源评估模型的构造必须考虑温度等环境因素,应加强环境变化引起的管理风险的评估。最后,简单探讨了遥感数据在渔业资源评估中的应用。更多还原

【Abstract】 In the last decades, although most demersal fish stocks have collapsed along the coast of China, the Chub mackerel (Scomber japonicus) still supports a large pelagic fishery in the East China Sea and Yellow Sea. The yield of chub mackerel in Chinese offshore fishery is about 0.4 million ton which is about 3 percent of the total output of marine capture fisheries in China. The catch from the East China Sea and Yellow Sea consists of 78 percent of this total yield. However, the future of this important fisheries resource is unclear with high fishing intensity and large changes in the ecosystem. Current research is focused on improved understanding of its population dynamics and spatial dynamics of fishing ground. The improved technology of ocean remote sensing in our country has resulted in increased capacity in monitoring environments of the fishery on a large spatial scale. The two ocean color remote sensing satellites (HY-1A and HY-1B) were launched in May 2002 and July 2007, and the ocean dynamic satellite series (HY-2) and ocean watch & monitor satellite series (HY-3) is scheduled to be launched, respectively, in 2009 and 2012. These satellites greatly enhance our abilities of collecting the information of oceanic environment in China Sea. Chub mackerel, a pelagic fish species, is used as an example to show the potential use of remote sensing in improving modeling the dynamics of marine fisheries stocks.This dissertation consists of four parts. Part one reviews (1) the development of ocean remote sensing in marine fisheries; (2) chub mackerel biology; (3)fishing ground and resource assessment; and (4) the Chub mackerel purse sense fisheries in China and related countries and area. Part two reviews the environments of East China Sea and Yellow Sea, the ocean remote sensing research, and the corresponding remote sensing data. Part three is focused on the methods of estimating relative fishing efficiencies and analyzing the relationship between CPUE and biomass based on the fisheries-dependent data. The last part tries to incorporate the remote sensing data into a production model in assessing fish stock dynamics. The main findings of this study are as follow:(1) Reliable estimation of effective fishing effort, which is proportional to fishing mortality, can provide information critical to the assessment and management of fisheries stocks. To estimate effective fishing effort, we need to understand fishing efficiency and factors that may influence it. In this study fishing efficiency was estimated for mackerel purse seine fisheries using a generalized linear model. This was done for different companies involved in the fisheries. Different choices of error structures were considered in the estimation and their impacts on the estimation were evaluated. The negative binomial distribution, gamma distribution, and log-normal distribution were chosen as error distributions according to log-linear regression of variance versus log-mean of CPUE (catch per unit effort). Zero CPUE values in data were found to have great impacts on the assumed error structure and adding a constant (5) to CPUE was needed for the gamma distribution and log-normal distribution in maximum likelihood estimation. As 5 increased, the contrast of estimated fishing efficiency was reduced greatly. Delta approaches were also chosen as an alternative way to deal with zero CPUE values in this study. Comparing the results of different models, we considered Delta-negative binomial and Delta-gamma as most appropriate error distributions for this study. The results showed that the fishing efficiency differed greatly among fisheries companies and among different areas.(2) CPUE is an abundance index commonly used in fisheries. It is often assumed that CPUE is proportional to fish abundance. However, the relationship between CPUE and fish abundance derived from the fisheries dependent data may be influenced by behaviors of fish and fishermen, making the proportional assumption invalid.The paper presents a cellular automata model which simulates the reproduction and spatial movement of individual fish schools and the corresponding commercial catch and movement of individual fishing boat. The model was applied to evaluate the relationship between commercial CPUE and stock abundance in different distributional patterns of fishing boats and fish schools. The study considers four scenarios such as:①the distribution of fishing boat is random but fish schools may be either random or not,②the distribution of fish schools is random initially, but become aggregated and the movement of fishing boat is corresponding to such changes in the spatial distribution of fish schools,③the distribution of fish schools is the same as that in②, but spatial distribution of fishing boats is always aggregated; and④the distribution of fishing boat is random initially, and become aggregated, and fish school is aggregated. I found that when the distribution of fishing vessels was random the commercial CPUE was proportional to abundance otherwise there existed nonproportionality between CPUE and stock abundance. For instance, the abundance is fluctuant in reserve with CPUE as in scenarios③and in scenarios④, I find different relationship between CPUE and abundance in different exploitive stages as fishermen get more and more experience in catching. The result also shows the cellular automata model is useful to explore or analyze some theory relationship in fishery stock assessment and management.We analyze the relationship between the proportion of catch and effort allocated to different locations based on the ideal free distribution theory. The study shows that the prediction of ideal free distribution was approximately supported in north fishing ground, i.e. to a certain extent, the CPUE was equalized and was independent of the fish abundance, and to the contrary, the effort may be in proportion to the abundance. Although this prediction was not supported in south fishing ground, the CPUE decreased with the standardized effort as its value was larger than 26, which implied that the interference competition was present and the relationship between CPUE and fish abundance may be weakened, even break down. Finally, we suggest we need to keep an eye on CPUE equalized when using the fish dependent data to study the dynamic of fish stocks.(3) I laid each position of catch from 1999 to 2003 on maps of SST (sea surface temperature) and chlorophyll-a generated from remote sensing. I find that in north fishing ground, Yellow Sea warm current has an important influence on the location of chub mackerel fishing ground. Although the concentration of chlorophyll-a may not be estimated reliably by remote sensing because of the sediment, the chlorophyll-a doubtlessly limit the distribution of the fish. On the southern fishing ground, the map shows upwelling plays an important role in determining the position of the catch. Generally, the fishing ground is at the warm edge of front and the Taiwan warm current almost controls the range of distribution of chub mackerel. However, the influence of chlorophyll-a on the fish distribution is unclear. A GAM model was used to analyze the quantity connection between CPUE and ocean environmental elements estimated from remote sensing. The result shows that on the northern fishing ground the effects of SST, SST grads, sea wind and mean sea level anomaly are significant. On the southern fishing ground, the effects of SST anomaly, mean sea level anomaly, sea winds and eddies kinetic energy are significant. The effect of chlorophyll-a is not significant on both fishing ground. But the formation of fishing grounds has a tight connection with the fish migrations and the spatial structure of environmental elements. So we should take the necessary cautions in using the model for prediction. In order to explore the connection of the distribution and its corresponding evolvement of fishing ground of chub mackerel with the change of environment, we construct a cell automata model to simulate the distribution of chub mackerel by using SST, chlorophyll-a and sea depth data. The result shows that the model performs reasonably well, but need to be fine-tuned and improved in future.(4) The yield of mackerel is strongly affected by environmental variability. It is important to identify the influence of the ocean environment on these small pelagic fishes for developing a fishery management plan for sustainable development of fisheries. We used offshore catch data in the East China Sea from the large purse sense of China and sea surface temperature derived from TRMM/TMI to explore the relationship between them. The result of principal component analysis of the time-series images showed that the abundance of small pelagic fishes had a short period of fluctuation, positively corresponding to the same period of environmental change and the intensity of the Kuroshio current and Taiwan warm current. The sea surface temperature of spawning ground from March to April also had significant and positive effects on CPUE from the corresponding fishing ground. The spatial and temporal distribution of the fish changed as a part of response to the environment change. Finally, we make suggestions for fisheries stock assessment. First the CPUE should be standardized and adjusted according to different spatial and temporal distributions of fishing effort. The environmental variables must be incorporated in models for fisheries stock assessment and risk evaluation must be made allowing for dynamic resource induced by environmental factors.In this dissertation, I also discussed the framework for expanding the dynamic production model by incorporating remote sensing data in the estimation of fisheries stock dynamics.更多还原