Characters.—From study of a large series of specimens of Anas oustaleti, Yamashina (1948) described two types of plumages: one type resembles that of A. platyrhynchos and another type resembles that of A. poecilorhyncha. He based his conclusions on both a study of prepared skins and observations of the molt of living specimens as reported by Kuroda (1941-1942). The following descriptions are quoted from Yamashina (1948:122).
Adult male in nuptial plumage of A. platyrhynchos type: "Whole head is dark green, except at the sides where buff feathers are plentifully intermingled, a dark brown streak through the eye, and faint white ring on the lower neck. Feathers on scapulars and sides of body are as those of Anas poecilorhyncha. Sides of body are vermiculated but some brown feathers are found even in the full nuptial plumage. Upper breast is dark reddish chestnut with dusky spots. Upper and under tail-coverts are as in Anas platyrhynchos. Speculum is as that of Anas platyrhynchos, but the tips of the greater coverts are buff instead of white. Central tail feathers are more or less curled upward. Base of bill is black, tip is olive color. Iris is dark brown. Feet, reddish-orange, webs darker." Eclipse plumage of adult male resembles that of A. platyrhynchos.
Adult male in nuptial plumage of A. poecilorhyncha type: "Resembles Anas poecilorhyncha pelewensis from the Palau Islands and Truk Island, but sides of head are browner, superciliary stripes and ground color of cheeks are more buffy. Feathers on upper breast and sides of body are more broadly edged with brown. Speculum is usually violet-purple as in the platyrhynchos type, but in two specimens from Saipan and Tinian, respectively, it is dark green as in Anas poecilorhyncha pelewensis. Tips of the secondaries are usually white, but sometimes very faint as in Anas poecilorhyncha pelewensis, and in one specimen from Saipan they are buffy. Bill is olive color with a black spot in the center of the upper mandible. Iris, dark brown. Feet, dark orange, darker in joints and webs." Eclipse plumage of adult male resembles the nuptial plumage.
Measurements.—Measurements of nine ducks from Guam and Saipan are: wing, 238-266 (252); tail, 75-84 (81); exposed culmen, 49-53 (51); tarsus, 41-43 (42).
Specimens examined.—Total number, 9 (5 males, 2 females, 2 unsexed), as follows: Mariana Islands, USNM—Saipan, 2 (Oct. 2, 3)—Guam, 1 (June 6); AMNH—Saipan, 2 (Aug. 7, 11)—Guam, 4 (Jan. 10, April 6, Dec. 11, 16).
Nesting.—At Guam, Seale (1901:25) found nests of the Marianas Mallard "among the reedy swamps and streams of the island." He obtained two downy young in June. Kuroda (1941-1942) reports nesting at Lake Challankanoa, Saipan, in July. He writes that nests contained 7 to 12 eggs. Ducklings and incubated eggs were obtained in June and July, but he is of the opinion that the breeding season may be longer. He notes that adults exhibit both nuptial plumage and eclipse plumage at the same time, suggesting that breeding may occur at various times in the year. A nest with seven eggs taken on July 4, 1941, at Hagoi Lake, Tinian, is described by Kuroda as having been found among rushes and constructed of dead leaves, stems, and roots and lined with down. He describes the eggs as being grayish-white with a pale greenish tinge, and measuring 61.6 by 38.9. Marshall (1949:202) saw a family of ducklings in April.
Remarks.—The Marianas Mallard is rare; probably it never has been very abundant in the small chain of islands to which it is restricted, because fresh water marshes and swamps are not extensive. The bird was first recorded by Bonaparte as Anas boschas a. Freycineti in 1865. This name was a nomen nudum and later the same specimen in the Paris Museum was named by Salvadori (1894) as Anas oustaleti. In 1888, Marche obtained six specimens at Guam; these were reported on by Oustalet (1896:49). Later collecting showed that the duck inhabited also the islands of Saipan and Tinian. There have been no records of this duck in the more northern islands of the Marianas. According to Yamashina (1948:121) in the period from 1931 to 1940, the Japanese obtained 38 specimens of the Marianas Mallard at Tinian and Saipan. In 1940, four birds from Tinian were shipped alive to Japan and kept in an aviary by Kuroda. At Tinian in 1940, one of the collectors observed two flocks of A. oustaleti, each containing 50 or 60 individuals. The Japanese took specimens at a lagoon area and at fresh water lakes. Yamashina describes one of the localities, Lake Hagoi on Tinian, as "a small body of fresh water surrounded by about 40 acres of marsh." During the war, servicemen reported the presence of the Marianas Mallard at both Saipan and Tinian. Moran (1946:261) counted twelve ducks at Saipan. Stott (1947:525) saw seven birds at Lake Susupe on Saipan in December, 1945. He writes that the birds were gentle and easily approached and that they preferred winding channels in reed beds to open water. Marshall obtained two ducks at Lake Susupe in early October, 1945. These specimens are in the United States National Museum. He (1949:202) found ducks at both Saipan and Tinian; twelve was the greatest number seen at any one time. Gleise (1945:220) estimated that there were twelve birds on Tinian in 1945, remarking that their habitat was swamp area.
At Guam and Rota, the NAMRU2 party failed to obtain any specimens but received reports of the presence of ducks on both islands. At Guam, reports were obtained of ducks of unknown species at a fallow rice paddy in August, 1944, and in a marsh near Agat on June 13, 1945. The presence of Japanese soldiers in the interior of Guam made it inadvisable to investigate marshes and swamps of the interior and the upper courses of streams. H. G. Hornbostel, as quoted by Phillips (1923:54), reported that ducks were found at Guam only in the Tolofofo River Valley. The NAMRU2 field parties investigated the lower reaches of this valley and found no evidence of the ducks. The upper part of this valley was used as an artillery range in 1945. Probably the firing of field guns was a disturbing influence to any birds that might have been there. If the ducks were on Guam at that time, they must have been secretive and restricted in their movements. At Rota, two ducks which might have been A. oustaleti were seen by the NAMRU2 party on October 20, 1945, in a cultivated field.
These recent reports indicate that the Marianas Mallard is secure for the present on the islands of Saipan and Tinian, but thoughtful conservation practices need to be placed in operation to insure its survival in the future.