高橋 鉄美

We conducted an investigation into ghost fishing, a phenomenon in which abandoned fishing equipment remains operational for extended durations, in the southern region of Lake Tanganyika, Republic of Zambia. The perils associated with ghost fishing have achieved international recognition; however, there exists a conspicuous dearth of scholarly investigations in particular regions, notably inland aquatic environments and developing countries. Employing scuba diving for field observations, we discovered that ghost fishing was solely evident in monofilament nylon nets, despite the presence of four different net types in the littoral area of the lake. Notably, endemic crabs Platythelphusa armata and P. maculata from Lake Tanganyika constituted 65.2% of the organisms captured by abandoned nets, with the remainder comprising various fish species. The number of captured organisms exhibited a dependence on the interplay between the length and mesh size of the abandoned nets, indicating that the risk of ghost fishing is heightened when both net and mesh size are substantial. Furthermore, we encountered predatory Spiny Eels Mastacembelus moorii ensnared in nets around their jaws and teeth. Intriguingly, the digestive tract of one of the Spiny Eels contained a partially digested cichlid species entangled in the same nets, strongly suggesting the occurrence of secondary ghost fishing.

Rhodeus ocellatus kurumeus is a small cyprinid fish endemic to the Japanese archipelago. This fish frequently hybridizes with an exotic subspecies Rhodeus ocellatus ocellatus, and non-introgressed populations of R. o. kurumeus have greatly decreased. Previous studies based on mitochondrial and nuclear DNA identified introgressed populations and inferred phylogenetic relationships, but these approaches may lead to underestimates of introgression or ambiguous results owing to the small number of genetic markers used. The present study applied double digest restriction site-associated DNA (ddRAD) sequencing to assignment tests and phylogenetic inferences. Data matrices can vary in size and content depending on the strategies used to process ddRAD sequences; therefore, we prepared 25 data matrices with various processing strategies. The assignment tests based on the 25 data matrices resulted in similar assignment patterns and provided evidence for introgression in some populations that had been identified as non-introgressed populations in previous studies. The maximum-likelihood (ML) phylogenetic trees differed in robustness and topology among the 25 matrices; interestingly, ML trees with low statistical support reflected geographical distributions of fish better than ML trees with high statistical support. These results indicate that ddRAD sequencing can detect introgression with greater sensitivity than conventional DNA markers, that ddRAD sequencing is useful for phylogenetic inference among closely related populations within a subspecies, and that statistically robust ML trees do not necessarily reflect the true phylogeny. Application of ddRAD data to conservation genetics and evolutionary history of this subspecies are also discussed.