Study On The Assessment Model Of Chilean Jack Mackerel Resources In The Southeast Pacific Ocean Under The Background Of Climate Change

For a long time,the low-frequency climate change phenomena?such as El Ni?o-Southern Oscillation,Pacific interdecadal oscillation,North Atlantic Oscillation,etc.?caused by the impact of climate change have affected the marine fish through the direct effect on the fish individuals or the indirect effect on the food chain transmission of the ecosystem.Research shows that more than 60%of the increased net energy in the climate system has been stored in the upper ocean in the past half century,and the pelagic fish are the first to suffer from the severe impact of climate change.Chilean jack mackerel?Trachurus murphyi?is an important economic fish in the Southeast Pacific Ocean.It is the target fish of purse seine fishery in Peru and Chile,and also the target fish of large-scale trawler processing ships in China and other countries.Taking Chilean jack mackerel in the Southeast Pacific as an example,studying the resource assessment model under the background of climate change is helpful to establish an effective fishery resource management method with climate change adaptability and provide scientific reference for the sustainable development of fishery resources under the background of climate change.In this study,the data of catch,Catch per unit effort?CPUE?and the low-frequency climate change phenomena such as El Ni?o,Southern Oscillation?SO?,Pacific interdecadal oscillation?PDO?,Atlantic Oscillation?NAO?and the data of global sea and air temperature anomaly index?d T?are collected,models for predicting the catches of Chilean jack mackerel in the Southeast Pacific Ocean are established by using correlation analysis and neural network model.And factor sensitivity analysis method is used to identify the effects of characterization factors of climate change on the catches of Chilean jack mackerel.Then,based on the residual yield model,climate fusion models are established to assess the current situation of Chilean jack mackerel resources in the Southeast Pacific Ocean under the background of climate change,and to construct management strategies for sustainable development of fishery resources.The main conclusions are as follows:?1?Through correlation analysis and cross-correlation analysis,the set of relatively independent characterization factors--{Nino,NAO,SOI,d T,PDO,NPI}and the lag order of each factor are determined.Taking these factors as the input variable and Chilean jack mackerel catch as the output variable,a three-layer neural network model with an implicit layer is established.Set the number of hidden layer nodes form 3 to 13,eleven sets of model operation schemes are established.Run these models to get the model results.Evaluating each scheme by 5 evaluation indexes such as prediction standard deviation and efficiency coefficient,Akaike information criterion the optimal fitting model is obtained.The optimal model fitting results show that the catch fitting value of the model has basically the same change trend with the observation value,and the scatter points in the scatter diagram of catch observation value and fitting value are basically clustered around the straight line:y?28?1.071x,so the fitting effect of model is good.Finally,factor sensitivity analysis of the characterization factors of climate change in the neural network model is carried out.The results showed that the six input variables of the model affected the catches of Chilean jack mackerel from high to low are:NAO,PDO,NPI,Nino,d T and SOI.?2?Based on the data of catches and CPUE of Chilean jack mackerel in the South Pacific Fisheries Organization from 1986 to 2016,an evaluation model of Chilean jack mackerel fishery is established.Setting the intrinsic natural growth rates of the model parameters r are uniform distribution,normal distribution and lognormal distribution,environmental capacity K and catching coefficient q are the non-information distribution--uniform distribution,and three simulation schemes are obtained.Run the model to get the evaluation results.The model results show that the posterior distribution of parameters under the three schemes is greatly different from its prior probability distribution,which shows that the data provides enough information for the posteriori probability distribution of parameters.The deviation information criterion values?DIC?of the three schemes show that the fitting effect of the uniform distribution scheme is slightly better.Under the three schemes,the biological reference points:B_MSY?B_MSY/4?B₂₀₁₆,F_0.1?F_MSY?F₂₀₁₆are close to each other.The current resource analysis shows that under the three schemes,Chilean jack mackerel fishery resources are lower than the target reference point but higher than and very close to the limit reference point,which indicates that Chilean jack mackerel fishery resources have been overfished.However,because the death coefficient of Chilean jack mackerel is less than the target reference point of death coefficient,the fishing intensity of Chilean jack mackerel fishery has not yet reached the category of overfishing.?3?Added six characterization factors of climate change?NAO,PDO,NPI,Nino,DT and SOI?to the Surplus production model of resource assessment one by one,so obtained six climate fusion models.Setting the prior distribution of parameters of six models is to be uniform,Used Bayesian analysis method to solve each model.Two models with better fitting effect are selected from six models by using the deviation information criterion value?DIC?:the climate fusion model including factor NAO and including factor PDO.The model results show that the parameters and the posterior distribution of the two models are very different from their prior probability distribution,which shows that the data provides enough information for the posterior probability distribution of the parameters.The current resource status shows that:under the two models,the current resource is between the target reference point and the limited reference point,and the resource status needs to be closely monitored;however,because the death coefficient of the Chilean jack mackerel is less than the target reference point of the death coefficient,the fishing intensity of Chilean jack mackerel fishery has not yet reached the category of overfishing in terms of fishing intensity.?4?Finally,a comparative analysis on the evaluation results and decision analysis results of model o?a general resource evaluation model with uniformly distributed parameters?,model I?including factor NAO?and model II?including factor PDO?is made.By comparing the error between the CPUE fitted value and the real value of the three models,and comparing the deviation information criterion value?DIC?of the three models,we can get the conclusion that the climate fusion model with NAO has the better fitting effect.The calculation shows that there are some differences in the index values of each biological reference point of the three models.Generally speaking,the values of B_MSY?B_MSY/4?B₂₀₁₆ of the two climate fusion models are close to each other,and there is a certain gap between the two models and the general evaluation model.From the current resource status,the three models show that:Chilean jack mackerel fishery resources are lower than the target reference point but higher than the limit reference point,indicating that the resources have been overfished;but the climate fusion model shows a slightly better resource status than the general resource evaluation model.In terms of fishing intensity,all three models show that Chilean jack mackerel fishery is not in the state of overfishing.Set the implementation period of simulation assessment management strategy as2017-2026,set risk assessment indicators,and analyze and compare the resource management strategies of Chilean jack mackerel under the three models.The results show that the harvest rates of the three models are 0.16,0.22 and 0.18 respectively when the highest catch is obtained.There are some differences in the maximum catches that can be obtained by the three models,the highest is the climate fusion model with PDO,which is 3561.6 kilotons,and the lowest is 2499.1 kilotons of the general resource assessment model.When the three models reach the maximum catcheable capacity,the recovery of resources and the probability of resources recovering to health level are relatively low after the implementation of management measures;however,the risk of resource collapse of the three models is different,among which the general resource assessment model is the largest and nearly 50%,while the climate fusion model is relatively small.According to the principle of "maximizing the expected return and minimizing the risk",taking into account the risk of resource recovery to health level and resource collapse after the implementation of management measures,the"best management strategy"of the three models is adjusted.It can be found that:compared with the general resource assessment model,the harvest rate of the resource assessment climate change fusion model is slightly higher,that is 0.17 or 0.18,and the catches are relatively high;and the risk indicators of the general resource assessment model are quite different from the climate change fusion model.From the perspective of risk indicators,the climate fusion model is relatively optimistic than the general model,and the risk of resource collapse is smaller.