| To understand the ecological differences in the feeding of Thunnus obesus and Thunnus albacares in the tropical Western and Central Pacific Ocean(WCPO),and to further understand their population distribution and their response to changes in the marine environment,a sampling survey was conducted during May 2018-February2019 in the tropical WCPO(2°03′S~11°17′S,163°14′E~173°35′E),sampled using tuna longlines and analysed for fork length,sex,catch depth,feeding intensity,stomach contents composition,and gonadal index of captured bigeye and yellowfin tuna.Analysis of variance(ANOVA)was used to investigate the effects of biological factors(fork length,sex,gonadosomatic index)and spatio-temporal factors(lunar phase and depth of catch)on the Shannon-weiner diversity index H′and Pielou homogeneity index J of the stomach contents of bigeye and yellowfin tuna by sex.The relationships between the effects of different biological and spatio-temporal factors on the feeding intensity of bigeye and yellowfin tuna were also analysed using an ordered multiple logistic regression model with feeding intensity class as the response variable,individual fork length,sex and gonadosomatic index(GSI)as biological factors,operational water depth as a spatial factor and moon phase as a temporal factor.The lunar phase was analysed using circular statistics by converting the lunar date into a circular variable.The results show that:(1)A total of 1123 bigeye tuna and 919 yellowfin tuna were sampled,with a male to female ratio of approximately 1:1.85 for bigeye tuna.Fork lengths ranged from 81to 195 cm,with the dominant fork length group being 120-150 cm and the mean fork length measuring 129.2 cm(±23.0 cm).For yellowfin tuna,the male to female ratio was 1:1.32,and fork lengths ranged from 87 to 163 cm,with a mean fork length of 119cm(±32 cm).Both male and female dominant fork length groups were 110-130 cm.(2)An ANOVA analysis of the fork length distribution of bigeye and yellowfin tuna by sex indicated highly significant differences in the fork length distribution of both sexes(bigeye tuna:F=8.26,P<0.001;yellowfin tuna:F=26.23,P<0.001).Similarly,an ANOVA analysis on fork length between species showed significant differences in the fork length distribution between bigeye and yellowfin tuna(F=129.3,P<0.001).(3)Out of the 682 bigeye tuna that had identifiable stomach contents,25.37%were empty.Among the non-empty individuals,analysis of stomach contents showed that gunny squid spp.,shrimp,and sardine spp.were more frequently present,with frequencies of 40.3%,39.7%,and 30.1%,respectively.Sardine spp.was also used as bait for longline fishing.In contrast,among the non-empty stomach individuals of yellowfin tuna,Sardine sp.was the most frequently occurring stomach contents composition,at 52.60%,followed by Gunnera spp.and Miscellaneous fish,at 35.34%and 33.70%,respectively.A comparison of the stomach contents composition of the two species showed that both bigeye and yellowfin tuna fed on cephalopods,fishes,and crustaceans,but there were differences(Bray-Curtis=0.525>0.500),with bigeye tuna feeding more extensively and containing more pelagic other fishes.(4)The ANOVA results indicated that catch depth did not have a significant effect on the Shannon-Wiener diversity index H’for bigeye tuna(P=0.200),but it did have a significant effect on the Pielou evenness index J(P=0.024).In contrast,both fork length and lunar phase had significant effects on H’and J for yellowfin tuna.The results of the ecotone width and overlap index showed significant overlap between the two tuna species(Q=0.802>0.6),indicating that bigeye and yellowfin tuna feeding on bait had mostly overlapping organisms with a few differences present.(5)The analysis of feeding intensity distribution by fork length revealed that male and female bigeye tuna had a higher feeding rank than female yellowfin tuna at the same fork length,but the difference was not significant.On the other hand,male yellowfin tuna had a significantly higher feeding class than females at the same fork length.Furthermore,the distribution of feeding intensity differed significantly between bigeye and yellowfin tuna at the same fork length,with yellowfin tuna having a higher proportion of empty stomachs and a lower proportion of high feeding classes compared to bigeye tuna.(6)To summarize,the distribution of feeding intensity of bigeye and yellowfin tuna was affected by water depth,with a higher percentage of empty stomachs in the water depth range of 50-200 m and a gradual decrease with increasing water depth.The proportion of high feeding classes increased with increasing water depth,and this trend was more obvious in male bigeye tuna.However,both male and female yellowfin tuna showed a higher proportion of high feeding classes in the 200-250 m water layer compared to other water layers,which was different from the preferred water depth of male bigeye tuna.(7)The feeding intensity of bigeye tuna and yellowfin tuna varied with lunar phase.The fitting results of the multivariate ordered logistic model showed that lunar phase had a significant impact on the feeding intensity of both species,with the proportion of individuals with empty stomachs being lower during the full moon phase,and the proportion of individuals with high feeding levels(level 3 and 4)being higher.Additionally,the feeding intensity of both species was mainly influenced by the cosine of lunar phase,with feeding intensity being higher around the full moon phase compared to other phases.When compared to each other,yellowfin tuna had a higher proportion of individuals with empty stomachs and a lower proportion of individuals with high feeding levels at the same lunar phase.(8)The distribution of feeding intensity of bigeye tuna and yellowfin tuna varied with month.There was a significant difference in feeding intensity among months for bigeye tuna(P=0.0017<0.01),while the significance of the effect of month on feeding intensity of yellowfin tuna was not high(P=0.124>0.05),indicating that the feeding of bigeye tuna was more susceptible to the influence of month.The results of the multivariate ordered logistic model showed that GSI had a significant impact on the feeding intensity of bigeye tuna(P=0.046),and the interaction between GSI and month also had a significant impact on feeding intensity(5=-0.229,P=0.028).In contrast to bigeye tuna,only the gonad index had a marginally significant impact on feeding intensity(3=-0.502,P=0.094<0.1,marginally significant)for yellowfin tuna. |