Four bitterling species, Tanakia lanceolata, Acheilognathus typus, A. melanogaster and A. tabira erythropterus, are native to Lake Kasumigaura, Japan. Although their local extinction is a matter of concern, the reasons for recent population declines and their underlying mechanisms have been poorly understood. In this study, we conducted field research on the population dynamics of these species from 1999 to 2011, to determine the cause of the population decline. Special attention was paid to the ecological impact of non-native bitterlings and the availability of unionid mussels as their spawning hosts. During the study period, T. lanceolata and A. typus were not collected at all around the lake basin. Populations of A. tabira erythropterus and A. melanogaster gradually disappeared by the year 2010. In contrast, a non-native bitterling, A. macropterus, which newly invaded around the year 2000, showed steady increase in its population after 2005. Of the other non-native bitterlings, Rhodeus ocellatus ocellatus showed a population decline, while A. rhombeus was rarely found throughout the study period. Generalized additive mixed model (GAMM) analysis revealed that the non-native bitterling, A. macropterus, has no impact on native bitterling populations. Since there was a remarkable decrease in the abundance of unionids by 2006, the population decline of the native bitterlings can be attributed to the decreased availability of their spawning hosts. In spite of the unionid mussels' absence, the non-native A. macropterus population expanded. It is likely that A. macropterus can utilize as its spawning host a cultivated freshwater pearl mussel, which is a hybrid between a Japanese Hyriopsis schlegeli and a Chinese H. cumingii, suggesting the possibility that its population is supported by pearl culture.
Sediment oxygen demand (SOD) and chemical composition of the profundal zone (≥ 80 m water depth) of the northern part of the North Basin of Lake Biwa, Japan were analysed to identify the mechanism of formation of large-scale oxygen-deficient waters. Sediment core samples for SOD measurements were taken from five stations (depth of 40-95 m). In addition, surface sediment samples for investigation of elemental composition (carbon, nitrogen, and phosphorus) and origin (autochthonous or allochthonous) of organic matter were collected from the entire area (65 stations). The horizontal distributions of organic matter content (g m-2) was almost uniform at a depth of ≥ 40 m, except for a limited zone. The percentages of organic matter (%) that predominantly consisted of autochthonous organic matter was higher in the profundal zone. The horizontal distribution of SOD at each measuring time was nearly uniform for all stations at a depth of ≥ 40 m, though SOD varied seasonally from 192 to 271 mgO2 m-2 d-1. The plausible reason for the uniform distribution of SOD at a depth of ≥ 40 m was the uniformity in organic matter composition and hypolimnion water temperature in this region. These findings suggest that oxygen dynamics in the water column above the bottom sediment greatly affects the mechanism of formation of large-scale oxygen-deficient waters.