To study the moult of remiges and rectrices of the Chestnut-cheeked Starling Agropsar philippensis, we collected 2,056 shed feathers at communal roosts in Kadoma City, Osaka Prefecture in 2011 and 2013, and in Kawanishi City, Hyogo Prefecture in 2015. The collected feathers were identified as primaries (P1 to P9), secondaries (S1-2, S3-4, S5-6), tertiaries (Te1 to Te3), and rectrices (R1, R2-4, R5, R6), and were classified as adult male, adult female, and juvenile feathers. The feathers were collected from early July to early October. Chestnut-cheeked Starlings have not been recorded breeding or wintering in the study areas, so they moulted at a stopover site during migration. Almost all parts of feathers were collected, but the number of feathers collected was not uniform. Based on the sequence of moulting, few feathers were collected from the first part of the moult ― the inner primaries and R1, and the last part ― the inner secondaries and P9. These observations indicate that they moult these parts in other locations and times. Three migration patterns and their corresponding divided moult of Chestnut-cheeked Starlings were proposed: (a) After breeding, the bird stays at the breeding site, leaves there in early September, stays at Stopover Site 2 for about one month, and then migrates to the wintering site in early October. The first and main parts of the feathers are moulted at the breeding site, and the last part at Stopover Site 2. (b) After breeding, the bird leaves the breeding site in early July, stays at Stopover Site 1 for about two months, moves to Stopover Site 2 in early September, stays there for about one month, and then migrates to the wintering site in early October. The first part is moulted at the breeding site, the main part at Stopover Site 1, and the last part at Stopover Site 2. (c) After breeding, the bird leaves the breeding site in early July, stays at Stopover Site 2 for about three months, and then migrates to the wintering site in early October. The first part is moulted at the breeding site, and the main and last part at Stopover Site 2. The proportions of adult females and juveniles were significantly lower than that of adult males for all feather parts collected. This suggests that adult females and juveniles tend to stay in the breeding grounds longer, resulting in a higher proportion of adult males in the study area, which corresponds to Stopover Site 1 in migration pattern (b). The importance of protecting the communal roosts at migration stopover sites for the conservation of the Chestnut-cheeked Starling was discussed.
View full abstract