The Birds of Yap, Western Caroline Islands

HARVEY 1. FISHER 1

THE YAP GROUP of islands, which lies about was made to collect large series of any spe 450 miles southwest of Guam and 125 miles cies; rather an attempt was made to secure northeast of the Palaus, is composed of four representatives of all species present. Records main islands: Rumung, Map, Gagil-Tomil, of the reptiles and amphibians taken have and Yap. There are several islets in the la­ been published (Fisher, 1948). goon between Yap and Tamil islands. The University of Hawaii sponsored this The highest elevation (585 feet ) is in the trip. I wish 'to acknowledge the aid of the north-central part of Yap Island (Fig. 1). United States Navy in furnishing rransporra ­ Except for this hilly area, the rest of Yap tion and other facilities. In particular, I wish Island is below 200 feet. All of Gagil-Tomil to thank Captain O. M. Murphy, United is below the 200-foot elevation, and the high­ States Navy Commander of the Yap Area, est parts of Map and Rumung are from 250 for use of various facilities at Yaptown and to 300 feet above sea level. The greatest for his interest in rhe project. I also want to length of the group is about 16 miles and thank Mr. EveniLevi of Tutuila, American the greatest width is 6.5 miles. Samoa, who accompanied me from Hono­ Ulithi atoll is the nearest land mass, lying lulu; he helped greatly in many ways. Dr. about 90 miles to the northeast. To the south­ Ernst Mayr has checked the identification of west are the Pala·us. Thus, Yap lies in a chain all species discussed and has made critical of islands from the Palaus in the south to taxonomic comments, many of which are the Marianas in the north. Nove of the is­ included in the species accounts. lands in the chain is widely separated from It is no longer possible for me to continue the others. Furthermore, to the east is the these studies. Hence, it seems worth 'while to whole group of Caroline Islands. Because put on record the observations of this brief these islands seem to form a perfect series of survey. "stepping sto~es" for the movement of plants As indicated above, the islands are rela- and animals from the southwest Pacific to o tively low. They are almost completely cov­ the north Pacific through the Marianas and ered with vegetation. Yap and Rumung show eastward through the Carolines to the central a similar zonation of vegetation . On the Pacific Ocean, it was deemed desirable to inner side (toward the lagoon ) these islands study birds of Yap, especially since the exten­ have extensive swampy areas, and mangroves sive work of the Whitney-South Seas Expe­ of two kinds form dense thickets at the ditions did not include Yap, and no modern ~ater 's edge. On the outer side the islands collections of birds from this area are known. have large areas of shallow water inside the Preliminary investigation and collections coral reef. For the most part, these have coral were made on Yap from July 28 to August and rock substrata with little mud. At the 24, 1946. A total of 148 verrebrarespeci­ upper ends of the tidal swamps on the lagoon mens were taken : 17 amphibians, 28 rep­ side are semi-open mudflats. Above these tiles, 9 mammals, and 94 birds. No attempt swamps or mudflats, or above the narrow beach on the sea side, the dense jungle starts , 1 Department of Zoology, University of Illinois, Urbana. Manu script received April 28, 1949. immediately ' and continues upward to an [55 ] 56 PACIFIC SCIENCE, Vol. IV, January, 1950

elevation of about 200 to 250 feet on the Thalasseus bergii, Crested Tern. *-Nine southern end of Yap Island; on the northern individuals were observed on tops of pilings end of Yap Island the hilly regions have of an abandoned fish trap off Orila, Map dense stands of trees, but here and there are Island. open areas of grassland with scattered Pall- . Gygis alba candida, Fairy Tern.*-This danus trees. Most of the rolling region to species was abundant on all the islands in the south is this same type of savanna. The those places where there were tall trees some central part of Rumung is much the same. distance apart. It was not seen in the deep Perhaps because of their lower elevation, jungle growth or in the savanna areas and Map and Gagil-Tomil islands do not have was most frequently observed in the tall these open areas. The usual swamps and breadfruit, banyan, and coconut trees and in mangroves are present as a fringe about the tall , isolated mangrove trees in the swamps. circumference, but tall trees cover most of I did not find it in low shrubs such as are used the rest of the surface. by G.a. rothschildi in the western Hawaiian islands. · . In the discussion of the species, the habitats will be mentioned as mangroves, swamps, Also in contrast to rothschildi, candida was jungle, or savanna , referrin g to the various very noisy at night. This constant calling situations described above. An asterisk indi­ was observed on three occasions, at 9:00, cates that specimens were taken . 10:30 P.M., and 12: 15 A.M. th~ Puffinus pacificus, Wedge-tailed Shear­ The fairy tern , like man-e-war bird. water.*-Five were observed flying over the is frequently used as an ornament of dress. south entrance to Tamil Harbor. Three natives had outlines of this tern tat­ tooed on their shoulders. Puffinus l'herminieri, Dusky Shearwater. Anous stolidus pileatus, Common Nod­ -Only one was seen, at the north end of dy.*-The noddy was found only on Tarang Rumung Island. Island in Tamil Harbor. Here, an estimated Phaethon lepturus, White-tailed Tropic 250 were seen, usually in the tops of pan­ Bird.-Twos and singles were observed sev­ danus, breadfruit, and chestnut trees. Old eral times over Map Island. nests were found in the top whorls of pan· Fregata ariel, Least Man-o'-War Bird.­ danus trees, 12 to 18 feet from the ground. Observed at close range at Yaptown (one No nests were observed in higher or lower bird ) and at Balabat (three), August 6 and sites. Egg shells were found on the ground, 13, respectively. Nine flew over Balabat 01: indicating that nesting had occurred not long August 15. N atives reported that the species before. Some nests were apparently under sometimes nested on Yap Island. The bird is construction, for they contained in their walls closely associated with native traditions ant! seaweed that was still green. Furthermore. customs, such as ceremonial dances, in which two birds carrying strands of seaweed were its feathers are used as ornaments. Identifica­ observed flying into nests. tion was made of one roughed-out skin used Pluvialis dominica fulva, Pacific Golden as a headdress. Plover.*-In late July an occasional single Sterna fuscata, Sooty Tern.-On July 30. or pair was observed. By August 15, they six birds were resting on the beach at the were quite numerous, 15 to 25 birds in 150 north end of Rumung Island. yards of beach, along the beaches, tidal flats, Sterna lunata, Gray-backed Tern. *- One and marshes, but they did not fly or feed in collected from a flock of nine Tbalasseus flocks. They were also abundant in open bergii off Orila, Map Island. areas of savanna and around the Nif airfield.

. '. ~ \. , Birds of Yap-FISHER 57

N

YAP GROUP rROM JAPANESE CHART NQ;ZI07 U.S. HYDROGRAPHIC orncc

NAUT ICA MIL ES 4- 5 II /0

FIG . 1. Th e Yap Group of islands. Charadrius mongolus, Mongolian Dot­ Numenius tahitiensis, Bristle-thighed Cur­ terel. '*' -One was collected on the mud of a lew. '*' - Only two singles were observed, both small taro patch in the jungle ~ mile west on rocky beaches. of Balabat, Tringa glareola, W ood Sandpiper. '*' -One Numenius phaeopus variegatus, Whim­ was collected on a mudflat at Moloai, Yap brei. '*' -Singles, pairs, and groups of three Island, August 10. No others were seen. were commonly found on mudflats where Heteroscelus incanus brevipes, W ander­ there was no vegetation. Three were observed ing Tatler.'*' - As would be expected, several feeding on the grass-covered airstrip at Nif. individuals were seen on rocky stretches of No change in numbers was observed during beach, but they were just as numerous Oil our .stay. small mudflats and around small patches of 58 PACIFIC SCIENCE, Vol. IV, January, 1950

Anas (superciliosa ?).-The "dark, female mallard with white line above its eye," ob­ served by Levi and me probably was super ciliosa. It was on the saltwater lagoon just off Yaptown. Captain O. M. Murphy said that the birds were 'frequently seen around the craters in the Nif airfield and that they were present throughout the year, but I did not find them there. Gallus gallus, Domestic Chicken.-A few semi-domestic fowl were found in the jung le near the native villages. I was told that be­ ~_; FIG. 2.Dense jungle grow th, 150 feet eleva­ fore the war such feral chickens nested in ..' tion, Dugar, Yap Island. Habitat of Rhipidura, . many different places. During the na va J M onarcha, and Myzomela. blockade in the latter part of the war the taro in the dense jungle. One was observed Japanese military forces hunted down most in an open area of the savanna region. of them for use as food. Arenaria interpres, Turn sron e.v-c-One Poliolim n a s cinereus collingwoodi, White-browed Rail.'*' - Mathews in 193q flock of five and one flock of two were ob­ micro~esiaeJ served on Yap Island, August 6 and 16, described the subspecies, a;ci respectively. stated that Micronesian specimens were inter­ Crocethia alba, Sanderling.'*' -Occasional mediate between collinguroodi from the Phil­ individuals were found along the beaches. ippines and specimens from Vulcan Island. Demigretta sacra, Reef Heron.-of 14 Examination of 32 specimens of "microne individuals observed on beaches and mudflats siae" and a series of collinguioodi failed to from July 28 to August 16, 11 were dark substantiate his remarks. Mayr (in litt.) gray, one was light gray, and two were white stated, "To this subspecies (collingiooodi) be in color. long also the specimens from Talaut Islarids, Nycticorax caledonicus, Rufous Night Palau, Guam, and Ruk. There are some Heron.-Two were in the edge of jungle characters peculiar to each locality but the dif­ adjoining the"Nif airfield. ferences betwe<:n populations of these islands Ixobryc hus sinensis, Chinese Least Bit­ tern. '*' ----,-These bitterns were most numerous around the water-filled bomb craters in the Nif airfield; on one occasion, 49 were seen in 4 hours. However, individuals were pres­ ent around permanent and temporary rain­ filled depressions in all parts of the islands. Each taro patch, no matter how small, had at least one bittern . They were abundant in the mangroves, in places a bird every 100 yards. Mayr (1945: 285) stated that this species was mostly nocturnal. I found it usually active in the daytime-feeding, call­ ing, and flying about from one pond to an­ FIG. 3. Grassy swamp, 20 feet elevation, Rumu, .i other. Its call was not heard at night. Yap Island . Habitat of Poliolimnas cinereus, Birds of Yap- FIsHER 59 are not sufficiently constant and conspicuous to permit separation from collingwoodi." These rails were present in all suitable taro patches or marshy areas that were investi­ gated. Th ey were not found in saltwater marshes but were present in marshes of brack­ ish water less than 200 yards from the sea and only 2 to 5 feet above sea level. All areas found to be occupied by this species had three things in commo n: mudflats, shallow water (2 to 4 inches) , and clumps of heavy marsh grasses. ,Apparently the size of a taro patch has FIG. 4. Elevated trail in sea level swamp, Bala­ little to do with the number of rails present. bat, Yap Island. Habitat of lx obrycbus and Polio­ limnas. I never found more than one pair in any isolated marsh or taro patch, and I investi­ group, except for occasional straying by a gated 13 such places, ranging in size from chick; when a chick strayed the parent called, circular areas 50 feet in diameter to swamps k-uk , k-uk, k-uk, in contrast to the usual, containing 3 to 5 acres. A pair was collected hard , ku k, kuk, ku k. So near to each other .from each of two smaller taro patches; during did these birds stay that I was able to collect the following 3 weeks no rails were seen all six with a single shot from a 4 1O-gauge there, although the areas were visited every shotgun at a distance of about 25 yards. other day during that time. On two occasions The adults were a male and a female. The individuals were flushed from dense grass chicks were covered with black down. Their some 50 yards from swampy areas. On both legs were dull blue. The distal third of both occasions the birds flew strongly and swiftly the upper and lower mandibles was dirty­ to cover in the swamp. I never saw this ivory in color; the middle third, extending species more than 5 feet from heavy, grassy proximally to the middle of the external cover. nares, was black. The basal third was washed­ Two trips were made at night to areas in yellow. The call of these chicks, whose body which rails fed in the daytime. In 4yz hours length was 2 inches, was much like that of of observation no birds were seen and no day-old domestic chickens, but finer, shrille r, calls were heard. and weaker. At Balabat, at 10:30 A. M ., on August 11, Ducula oceanica monacha, Micronesian I saw an adult run across a small opening in Pigeon.*-These were not numerous on the the middle of an abandoned taro patch over­ islands, but one could be certain of seeing grown with water grass. A few seconds later five or six in a half-day. It was seldom seen another adult followed; it wa s collected. in the savanna area, except at ' dusk when When I retrieved the bird I found I had also there was considerable movement between collected a chick. At 2:30 P. M. on August 12 wooded regions. As a rule, the birds were in a similar area, I watched two adults and first observed high in the barer branches of , a bro()d of four chicks feeding in water 1 inch trees 30 to 60 feet in height, but one was deep. During the entire 40 minutes this collected while it rested on a nest some 20 group was watched, both adults kept their feet from the ground in a breadfruit tree; the 'tails up in the air and jerked them vertically nest was old and empty. The call is a single at 5-second intervals. They kept in a close augh or ungh. ()O PACIFIC SCIENCE, Vol. IV, January, 1950

differs from that race by the more extensive white tips on the tail feathers and broader white edges on the feathers of the breast and middle of the abdomen. The rufous zone on the lower back is more reduced. There is a more distinct grayish wash in the black por­ tions of the tail feathers. The similarity is so striking that I wouldn 't be surprised if Yap had actually been colonized from Malaita. Although found on all the islands, they are most abundant on Tamil. Th e edges of low undergrowth in lowland jungles and the edges of mangrove swamps seemed to be fav­ orite habitats, but a few were seen far back FIG. 5. Typical taro patch swamp, Omin , Map in the jungle. Without exception, the birds Island. A pair of rails nested here. were found in twos or in family groups; three broods of nearly fledged young (three in each Gallicolumba xanthonura, Ground Dove." brood ) were following pairs on Augustll. -Comparison of specimens from Yap and the They are constantly on the move and are Marianas revealed little significant difference. wary, but they, at this season at least, came The only female from Yap was somewhat in toward squeaking noises. They responded darker above and particularly darker below best to the sound of two pieces of seasoned in the throat and breast; specimens from the wood being knocked together, as when ashes Marianas are more rufous cinnamon in these from a pipe were knocked out by hitting it parts of the body. Size seemed to be identical against a gunstock. As one ' moved along a in the two series. trail these birds frequently flew out ahead, This species is uncommon on the island; displaying .vigorously. They became most I saw only 14 (singles, except for a male active just after sundown and before total and a female collected together) in 4 weeks. darkness. It appeared to be limited to mangrove th ick­ Monarcha godeffroyi, Monarch Fly­ ets in or near water and was not seen any­ catcher.*' - Specimens collected showedmuch where in the interior of Yap Island. individual variation in extent of the black, Rhipidura rufifrons versicolor, Rufous­ brown, and white coloration. Two birds in fronted Fantail.*' -Mayr (in litt. ) wrote as brown coloration, similar to M. takatsukae follows: from Tinian, were taken. Both were adult The Yap race which I had not previously females, on the basis of degree of ossification seen is more distinct from the races of the of the skull and development of the ovaries; Marianas Islands than I had expected. The one was feeding nearly grown young. Imma­ extensive, buffy, ochraceous wash of belly and tures collected were dirty-black or brownish­ flanks is missing. The well-defined, clear, black with dirty-white markings. Males were white upper throat is another striking differ­ white-backed with white ventra, and the fe­ ence as are the gray-brown flanks. The upper­ males had a white ring of varying extent parts are darker and more rufous with the about the neck. . contrast between rump and back much less pronounced. Actually, the Yap bird is, in On all islands of Yap, the species was the coloration of its underparts, much more abundant along the trails in heavy jungle. similar to the Solomon Islands races and in In this habitat the species was usually found particular to brunnea from Malaira Island. It 20 or more feet up in the vegetation; not 10 Birds of Yap:-PIsHER 61

other places, this species is in frequent con­ flict with other species; on Yap, Myzomela cardinalis and Monarcha were the chief vic­ timsof this conflict. Myzomela cardinalis kurodai, Cardinal Honey-eater.*' - It is unfortunate that the series of skins from Yap contained only one of a female, for the races are more clearly pronounced in the females. The one female agrees closely with an adult female k obayashi from Palau; both have distinct olive edges to the tail feathers, and the red is restricted FIG. 6. Sea level swamp, Balabat, Yap Island . tothe top of the crown and does not extend Habita t of Poliolimnas and Ixobrychus; in sur­ to the nape; the comparative blackness of rounding mangroves Gallicolumba xantbonura was the scapulars in the Yap bird is perhaps due collected. to its fresh plumage. The only difference the 'low understory as was Rbipidura. How­ between these two females is that the red of ever, this flycatcher was also observed perched the crown and throat of the specimen from on bare pandanus limbs ( 10 to 20 feet high) Yap is more cherry red, 'less scarlet, and there in the savann a region; here, it flew out after is a more pronounced 'zone without red be­ insects in' typical flycatcher fashion. Mayr tween the upper throat and breast and be­ ( 1945: 92 ) noted that this seldom occurred tween the crown and back. The wing and .in Monarcha; he also stated that the tail tail of the Yap race appear to be somewhat drooped vertically as the bird perched motion­ longer. The bill of the Yap birds appears to less. The tail was never drooped in the be longer than that of the Palau birds, but monarchs of Yap. measurements do not confirm this visual im ~ Aplonis opacus kurodai, Micronesian Star­ pression. The edge of the newly molted pri­ ling.*-Three specimens from Yap agree maries seems more olivaceous in Palauan fairly well with orii Takatsukasa and Yama­ than in Yap males. The under-tail coverts shina from the Palaus, but the gloss on the of birds from Y ap seem blacker than in birds birds from Yap seems to be more blue green, from Palau. On the basis of .the specimens less bottle green; this may be due to slightly before me I would not be inclined rorecog­ greater wear on the specimens from the nize k urodai, but it must be considered that Palaus. The bill seems slightly longer on rhe the Yap series is in full molt. (Mayr, in litt.) birds from Yap (Mayr, in lirr.). The honey-eaters were present in all suit­ The species was abundant on all the islands able habitat on all islands. They were mosz in all types ofcover except the savanna. Birds abundant in low brush, no matter whether were most numerous, however, in the tops of this brush was in the jungle, in ' otherwise high trees and were seldom, if ever, encount­ open country, or in a mangrove swamp. Iii ered in low, dense brush. The tops of coconut their brush habitat the birds appeared to be trees were much used as singing perches, .and weak flyers, their fluttery flights seldom hau trees (Hibiscus tiliaceus) were frequented longer than 15 feet, but on August 21 one for their fruits. No nesting birds were found, was observed over the center of Tomil Har­ but one juvenile was observed (August 7 ) bor, more than Y2 nautical mile from land. begging and receiving food, and an adult was As a rule, they were observed in pairs, but observed carrying food on August 16. As in on August 7 a brood of three was following 62 PACIFIC SCIENCE, Vol. IV, January, 1950 a pair. On August 10, a female (collected) differences between the two species are, how­ was starting a nest in the outer tips of ? ever, in the shape of the bill, which is nearly branch some 10 feet from the ground. twice as long as in sanfordi, and in the eye­ Another female (collected, but unfit for a ring which is obsolete in sanfordi. The ex­ specimen) was observed carrying food on posed culmen measures about 14 millimeters August 10. in oleaginea and over 20 in sanfordi." (Mayr, Zosterops conspicillata hypolais, Bridled in litt.) White-eye.*-The Yap subspecies, hypolais, In the month spent on Yap, only two indi­ is strikingly different from all the other races viduals were observed; one on August 6, 1 of the species. It is by far the most grayish mile northwest of Yaptown, Yap Island, and race of the species. The eye-ring is inconspic­ one on August 7, less than Y.4 mile from the uous and the dark area under the eye is gray­ same area. One was in a bush overhanging ish rather than blackish. The white loral a swamp in dense jungle, and the other was region is much reduced. The upperparts are in low, thick underbrush 200 yards from a gray', with a greenish tinge, most conspicuous swamp. on rump and crown. The underparts are Lonchura punctulata, Weaver-Finch.*­ somewhat like those of rotensis but slightly The .three skins agree best with a series of more yellowish, less buffy, particularly on the the race cabanisi from the Philippines. They flanks. The size seems the same as in most are also rather similar to the race topela from other races of this species in which the wing the Asiatic mainland but are distinguished varies between 55 and 58 millimeters in from this subspecies by the fine barring on length in adult males. (Mayr, in litt.) their flanks and by a less extensive brown This species was usually confined to the area on the throat. (Mayr, in litt.) undergrowth at the edge of the jungle. One This weaver-finch was found in all open could find it in numbers from 2 to 15 in and grassland areas on the islands. Flocks of almost every such area 100 yards long. 40 were observed in the shorter, dry grass . Rukia [Kubaryum] oleaginea, Large Yap of the savanna regions, and they were espe­ White-eye. *-This species differs from the cially numerous in flocks of 5 to 20 birds related Zosterops conspicillata by being along grassy roads and wide trails where they larger, darker, and more brownish olive. "In were feeding on seeds. Some immatures were general coloration it is remarkably similar to begging for and receiving food on August 16. sanfordi from Ponape. It is somewhat darker, This species probably was introduced on particularly on the crown, forehead, and Yap. lares. There is an extensive blackish zone m the malar region and under the eyes, which REFERENCES is barely indicated in sanfordi. The under­ parts also are darker and more grayish, while FISHER, HARVEY 1. 1948. Locality records in sanfordi they are more brightly greenish of Pacific Island reptiles and amphibians . buff. Rump and tail, as well as the edges of Copeia 1948 (1) : 69. the primaries and secondaries,are duller. The MAYR, ERNST. 1945. Birds of the Southwest legs are duller, more grayish horn color rather Pacific. xix + 316 pp. Macmillan Co.. than yellow as in sanfordi. The two principal New York.