Research On The Habitat Distribution Model Of Albacore (Thunnus Alalunga) In The South Pacific

Albacore(Thunnus alalunga)is one of the main targets in tuna longline fishery and is also a highly migratory pelagic species whose spatio-temporal distribution is vulnerable to environmental variables.Better understanding of the effects of environment variables on potential habitat of albacore is able to find the optimum habitat for it in the South Pacific which can reduce the cost for finding out the fishing ground,and is of great scientific significance for understanding the distribution of albacore.In recent years,the albacore in the South Pacific is mainly caught by China fleets and Chinese-Taipei fleets.Based on the logbook data collected by China during 2015-2017 in the South Pacific,combining with environmental factors of sea temperature at different depth,sea salinity at different depth,chlorophyll-a concentration,sea surface wind,sea surface height,mixed layer depth,we analyzed the relationship between catch rate and environmental factors so as to examine the effects of environmental factors at different depth by establishing generalized additive model(GAM)and maximum entropy model(Max Ent).In addition,the correlation coefficient of each environmental factor had been established.In addition,Those factors with close correlation are grouped and modeled by correlation analysis.The potential habitat of albacore in the main fishing season in 2017 was explored,and the predication accuracy of the two models were compared by superimposing the actual catch data.The results are obtained as follows:(1)The relationship between the catch rate and environmental factors was analyzed by GAM.Sea surface temperature and sea temperature at depth of 120 m,sea surface temperature and sea surface height,sea temperature at depth of 120 m and sea surface height,sea temperature and sea salinity at depth of 300 m were highly-correlated factors.However,sea surface salinity,chlorophyll a concentration and northward sea surface wind had no significant correlation with the other environmental factors.The explained cumulative deviance was 30%?40%;the environmental factors sorted by importance are listed as following: sea temperature at depth of 120 m,sea surface temperature,sea temperature at depth of 300 m,sea salinity at depth of 120 m,sea surface height,sea salinity at depth of 300 m,sea surface salinity,mixed layer depth,northward sea surface wind,eastward sea surface wind and chlorophyll a concentration.The sea temperature at depth of 120 m was negatively correlated with CPUE(Catch per unit effort)at 15?30?.The trend of sea surface temperature was similar to the sea temperature at depth of 120 m,with a positive correlation at 25?28?.The sea temperature at depth of 300 m and CPUE showed a significant positive relationship at 10?18?.(2)The effect of environmental factors on catch rate in the main fishing season(from May to August)was analyzed by Max Ent.The average AUC value obtained by Max Ent reached 0.893,and the AUC value in 2016 was higher than that in 2015.The optimal range of environmental factors were 18.9?29.9? of sea surface temperature,18.3?28.4? of sea temperature at depth of 120 m,13.5?17.4? of sea temperature at depth of 300 m,35.9?36.7‰ of sea surface salinity,0.2?0.7m of sea surface height,64.2?116.9m of mixed layer depth,-2.2?3.4m/s of eastward sea surface wind,-2.7?4.9m/s of northward sea surface wind.The environmental factors sorted by importance were listed as following: sea temperature at depth of 300m(33.8%);sea surface temperature(33.7%);sea temperature at depth of 120m(27.8%);sea surface salinity(22.3%);eastward sea surface wind(15.3%);sea surface height(13.9%);northward sea surface wind(9.7%);mixed layer depth(6.1%).There were also differences in the contribution rate of environmental factors in different years.The environmental factors with high contribution rate in 2015 were sea surface temperature,sea temperature at depth of 300 m,sea temperature at depth of 120 m.The contribution rate of sea temperature at depth of 300 m,sea surface salinity,sea temperature at depth of 120 m were high in 2016.(3)The main area of operation for Chinese albacore longline fishery fleets situated at 150°E?130°W,0°N?40°S in the South Pacific,mostly concentrated at 140°E?180°E,20°S?40°S and 175°W?130°W,25°S?50°S.Longitudinally CPUE in the east is higher than that in the west,the south is better than the north on the latitude.The main fishing season appeared from May to August.Albacore was mainly distributed at 10°S?40°S,150°E?130°W in the south Pacific in two models.Using latitude line 25°S as a reference boundary,the distribution of fishing ground in south of 25°S is better than north of 25°S.In the future,there is the possibility of new fishing ground to be explored in south of 25°S.The predication accuracy was 20%?90%.The predication accuracy of medium potential habitat was relatively high,but both the high potential habitat and the low potential habitat were low.Specifically,the ability of GAM to predict the high potential habitat was higher than that of Max Ent.When predicting the low potential habitat,Max Ent was superior to GAM.In the future,GAM can be used to predict the high and medium potential habitat of the species,and Max Ent can be used to predict both the medium and low potential habitat of the species.