We analysed the mitochondrial control region sequences of 150 Ryukyu robins (Erithacus komadori) from five migratory and three sedentary populations. E. komadori is endemic to the subtropical Asian islands, is well differentiated among populations in morphology and migratory behaviour within its narrow habitat range, and is ideal for examining the effects of altered migratory habits on population differentiation. Maximum-likelihood analysis among the haplotypes resulted in the generation of a single tree in which two phylogenetic clades corresponding to the two subspecies were evident. Within the northern lineage, three distinctive subgroups of populations (one migratory and two sedentary groups) were observed by population genetic analysis. The migration-related wing morphology and molecular data were then compared among groups. The wing shape of the northern lineage was significantly more pointed than that of the southern sedentary lineage, and they were exclusively discriminated from each other. The difference was not so apparent between the sedentary and migratory groups within the northern lineage, suggesting that the present northern subspecies is primarily a migratory lineage. The different migratory behaviours of the robin played an apparent role in maintaining the genetic structure at two different levels--one between the southern and northern lineages and the other between the sedentary and migratory groups within the northern lineage. While gene flow had long been maintained among the islands occupied by migratory individuals, migrants have been unable to contribute to the gene pool of the sedentary populations despite passing through the breeding range of such populations twice each year.

The distribution of the Ryukyu Robin Erithacus komadori was investigated from field observations, museum skins, references and personal communications. The species is endemic to southwestern Japan, with a single specimen, a male, from Korea and one sight record from Taiwan. Many individuals have been observed in the northern Tokara Islands in the breeding season, but few remain in winter, while on the other hand, there are few records of the species in the southern, Yayeyama, islands, and all are from nonbreeding season. It seems that some of its northern populations would have migrated south in winter. The measurements and plumage characters described by Kuroda (1923) for his new form, subrufus, are not distinctive enough, in our comparison, from those of the race komadori. Furthermore, the type specimen of E. k. subrufus was collected in Yonakunijima in the migratory season, October. It is, therefore, condluded that the race subrufus is a synonym of komadori. The sporadic present distribution of the Ryukyu Robin reveals its recent relic status from its former more extensive distribution. Materials and Methods We made field observations in Tairajima and Nakanoshima Islands, Tokara Islands

A ground nest of the Eastern Turtle Dove Streptopelia orientalis was found under a dense bamboo grass cover in burned mature forest in western Hokkaido, northern Japan. Two nestlings successfully fledged from the nest. Artificial nest experiments revealed that arboreal nests suffered more predation than ground nests. The predation rates of both ground and arboreal nests were not significantly related to egg colors (white painted or natural quail egg color) or vegetational concealments (e.g. foliage coverage or height and density of the undergrowth).The main predator of arboreal nests appeared to be the Jungle Crow Corvus macrorhynchos which bred near the study area. The foliage coverage above the arboreal nest was insufficient to prevent the attack of crows, but bamboo grass cover as the undergrowth prevented them from coming down to the ground. Moreover, bamboo grass cover prevented large carnivorous mammals such as the red fox Vulpes vulpes from invading the nesting area.

Breeding ecology of the Short-tailed Bush Warbler Cettia squarneiceps was investigated in western Hokkaido during five seasons (1990)(1991)(1992)(1993)(1994). The first arrival and first egg-laying dates were similar among years. Clutch sizes varied between 4-6 eggs, but they decreased as the breeding season progressed. The mean incubation and nestling period was 12.9±0.5 days (SD, n=18) and 10.0+0.4 days (SD, n=17), respectively. Only females incubated eggs and brooded nestlings for a long period of time. Most of the feeding of the young was conducted by males. Different (beta or gamma) males from the resident (alpha) male appeared and sometimes fed the young in 48.1% of twenty five nests observed. Some males were captured at locations far from their own nests, while there were only few cases in which a female was captured at a place distant from her mate's song area. As it was often observed that a beta male was the neighboring territory holder and his own nest was in the incubation period, it is hypothesized that some males who are emancipated from the incubation, might visit other territories to seek extra-pair copulations (EPC). We have, however, not observed mating between a female and a beta male except in the case of a widow. Within the breeding season, the females usually selected the same partner for the second brood. Main reason why the divorce rate within the season is low for the warbler in our study area is probably the high success rate for the first brood.

No abstract