Molecular analysis of stomach contents from Antarctic toothfish (Dissostichus mawsoni) collected from Area 58 and 88 from 2016 to 2018

Antarctic toothfish, Dissostichus mawsoni is the ecologically important Nototheniidae, which is native to the Southern Ocean. Although it plays an important role in the Antarctic ecosystem, the ecology of D. mawsoni is still fragmental mainly due to the limited accessibility for the sample collection. Diet study provides the basic information of feeding relationship among the different trophic levels. Since 2016, we have been studying the stomach contents of D. mawsoni by the combination of morphological analysis and the molecular identification using next generation sequencing (NGS) platform. We here report about the prey item of D. mawsoni collected from the two research areas (58 and 88) from 2016 to 2018 using metabarcoding analysis of 1,329 its stomach contents. Total 71 haplotypes (8 orders, 17 families, and 32 genera) were identified by COI universal primers, which included 60 fish and 8 cephalopod species. As reported previously, proportions of fish taxa occupied 98 % indicating the major prey items of D. mawsoni are fish species. Higher prey species numbers were obtained by the metabarcoding analysis compared to the morphological analysis especially it was useful to identify the species with similar morphological characters including fish in Family Channichthyidae. Among the prey objects, two fish species, Macrourus whitsoni and Chionobathyscus dewitti were the most important prey items of D. mawsoni and ratio between two species appear to be related to the size of D. mawsoni. Many of fish taxa in the stomach of D. mawsoni showed the multiple haplotypes by NGS analysis suggesting that those haplotypes would provide more useful data to understand the ecology of D. mawsoni. However, the length of the amplified PCR products (~600 bp) in this study may not have enough size for the clear analysis of haplotypes. Further study should be made to obtain the longer DNA marker for the identification of each species and its haplotypes. PacBio sequencing system would be the good candidate for the purpose, which can produce the much longer reads than the MiSeq sequencing system.