THE AND BIOGEOGRAPHY OF THE THICK-BILLED COMPLEX IN BORNEO

FREDERICK H. SHELDON • WesternFoundation of VertebrateZoology, 1100 Glendon Avenue, LosAngeles, California 90024 USA

ASSTRACT.--TheWestern Foundationof VertebrateZoology expedition to Sabah,East Ma- laysia(North Borneo)collected specimens and field data on Brown-backedFlowerpeckers (Dicaeumeveretti) and Thick-billed (D. agile)in 1982 and 1983. Dicaeumagile had not been known from Borneopreviously, and its presencethere changesthe biogeo- graphic assumptionsSalomonsen (1960) used to classifythe "thick-billed" flowerpeckers. Salomonsen'sD. everettiwas found to be a valid ,but his splitting of the Philippine StripedFlowerpecker (D. aeruginosum)from D. agilewas not justified. agile of Borneo is here placed in the race remotumwith the Malayan form of the species.The Borneanand Malayan races of D. everettiare grouped in the subspecieseveretti. Received 2 October1984, accepted12 February1985.

THEthick-billed flowerpecker complex of In- D. everetti specimenswere thought to be the dia, SoutheastAsia, and the Philippines is list- first from Borneoin 27 years (Pfeifer 1961). In ed by Salomonsen(1967) as comprising three fact, no specimensof D. everettihave been col- species, the Thick-billed Flowerpecker (Di- lected from Borneosince 1901, even though caeumagile), the Brown-backedFlowerpecker (D. several museums have specimens labeled as everetti),and the (D. aeru- such. The two specieshave been confusedfor ginosum).Of thesespecies, only D. agileof India 85 years, largely becausethe scarcityand dis- and Sri Lanka is common and well documented persionof specimenshave led to oversightsin (Ali and Ripley 1974). In other parts of South- the taxonomic literature. east Asia, the three speciesrarely have been The Western Foundation field data and the recorded, and their statuses are uncertain. This new and old specimenmaterial permit a criti- is especiallytrue in Sundaland,the areasof land cal evaluation of Salomonsen's (1960) classifi- such as Malaya, Borneo, and much of the In- cation of the thick-billed flowerpecker com- donesianarchipelago associated with the Sun- plex. His controversial separation of D. everetti da shelf. Dicaeumeveretti in Malaya is known from D. agile(sensu Mayr and Areadon 1947) is from only four specimens.Dicaeum agile in Java strongly supported, but questions are raised is known from only two specimens.Dicaeum about some of his racial taxonomy and about everettiin the Natuna Islands and D. agile in the schemeproposed to explain the radiation Sumatra are known from only one specimen and speciationof the thick-billed flowerpeck- each. ers in the Sunda and Philippine islands. In 1982 and 1983, the Western Foundation of VertebrateZoology expeditionto Sabah,East TAXONOMIC HISTORY Malaysiacollected both D. agileand D. everetti. Salomonsen (1967) divided the thick-billed The D. agilespecimens were thought to be the first from Borneo (Smythies 1981). However, flowerpeckercomplex into three speciesbased museumstudy subsequent to the expeditionhas on his 1960 review: D. agile,encompassing the shown that specimensof D. agilehave existed racesagile, zeylonicum, modestum, remotum, atje- from Borneo, in small numbers, since 1898. The hense,finschi, tinctum, and obsoletum;D. everetti, comprising everetti, bungurense,and sordidum; and D. aeruginosum,composed of affine,aerugi- *Present address:Department of Biology, Yale nosum, and striatissimum. He considered D. ev- University, P.O. Box6666, New Haven, Connecticut eretti and D. aeruginosumto constitute a super- 06511 USA. species. The only sympatric forms in the

606 The Auk 102: 606-612. July 1985 July 1985] BorneanThick-billed Flowerpeckers 6O7

T^BLE1. Rangesand characteristicsof the subspeciesof the thick-billed flowerpeckercomplex.

Wingb Bill Subspecies Locality Dorsum' Underparts Tail (mm) shape agile India, Pakistan Olive-gray Faint streaks Bright spots 55-65' Thick zeylonicum Sri Lanka Dark olive-gray Moderatestreaks Bright spots 55-57' Thick modestum Assam-Indochina Olive-green Moderatestreaks Bright spots 57-61 (13) Thin remotum Malay Peninsula, Olive Moderate streaks Faint spots 57-63 (24) Thin Borneo atjehense Sumatra Olive Moderate streaks Faint spots -- Thin a finschi Java Olive Moderate streaks No spots 62 (1) Thin tinctum Sumba, Lesser Olive Faint streaks Faint spots 54-59 (9) Thick Sundas obsoletum Timor Grayish olive No streaks Faint spots 53-62 (11) Thick everetti Borneo Olive-brown No streaks No spots 54-60 (14) Thin bungurense Great Natuna Olive-brown No streaks No spots 63• Thin Island sordidum Malaya, Riau Olive-brown No streaks No spots 59-61 (3) Thin Archipelago affine Palawan Grayisholive Heavy streaks Bright spots 61-66 (7) Thick aeruginosum Mindoro, Negros, Olive-green Heavy streaks Faint spots 62-68 (8) Thick Cebu, Mindanao striatissimum Luzon, Sibuyan Brownisholive Heavy streaks Faint spots 64-65 (2) Thick aColor terminologyfollows Smithe (1975). bSample sizes are given in parentheses,except in instanceswhere measurementsare taken from indepen- dent sources. ½Measurements from Ali and Ripley 1974. a Measurements from Chasen 1939b. e Measurements from Chasen 1934. complex known to Salomonsenwere D. a. re- cal to the Borneanforms of the speciesD. ev- moturnand D. e. sordidurnof Malaya. The distin- erettiand D. agile,respectively. guishing features of the three speciesare as Chasen(1935, 1939a)grouped the Sundaland follows:D. agilehas a comparativelypale olive forms, sordidurn,everetti, and bungurense,with or gray dotsurn,streaked underparts, and (ex- the Indian race, agile,in the species"Piprisoma" ceptfor finschi)white spottingat the tail tip; D. agile(Piprisorna being a subjectivesynonym of everettihas a browner dorsum, no underpart Dicaeurn,with type speciesagile), and he united streaking,and no tail spotting;and D. aerugi- remotum and modesturn in P. modesturn. Al- nosurnhas the same characteristicsas D. agile though it attempted to recognize the distinc- but generally is larger and darker (Salomonsen riveness of the sordidum-everetti-bungurense 1960). The ranges and characteristicsof these group, this arrangement was counterintuitive. races are summarized in Table 1. It ignored the dissimilaritybetween the Indian Until Salomonsen's review of Dicaeum, the and Sundalandbirds, and it createda disjunct member racesof the thick-billed flowerpecker distribution in which the continental South- complexhad beenclassified in a varietyof ways. east Asian P. modestumseparated the Indian For example,Sharpe (1885) included the 5 then- from the Sunda forms.Had Chasenunited ag- known formsas full speciesof the genusPrio- ile, remotum,and modesturnin one speciesand nochilus.Oberholser (1923) split the group into sordidurn,bungurense, and everettiin another, his 3 genera and 5 species.Mayr and Areadon classification of these taxa would have made (1947) transferredthe group from more morphologicaland geographicalsense. As to Dicaeumand lumped all the racesinto a sin- they were, Chasen'sgroupings did more harm gle species,D. agile.To understand the rela- than good becausethey were believed to be the tionships of the Borneanmembers of the com- product of over-splitting and, consequently, plex, the treatmentsgiven the Malayan forms, were later ignored or attacked in reviews. As sordidurnand rernoturn,must be emphasizedbe- further complications,the few existing speci- causethese Malayan are virtually identi- mens of sordidumand bungurense,deposited in 608 FREDERICKH. SHELDON [Auk, Vol. 102

TABLE2. Specimenrecords of Dicaeumagile collected in Borneo.a

Num- Altitude bet Date Locality Habitat (m) Museum b 3 Dec 1898 Upper Mahakam R. (Kaliman- -- -- RHNL, USNM tan) 4 Dec 1947- Kelabit Plateau, Sarawak 2ø forest, fruit 1,000 AMNH, SM Jan 1948 trees 1 28 Sept 1952 Long Tinkalat, Tinjar R., Sara- 2ø forest 300 SM wak 7 Sept 1956 Long Peso,Kajan R., eastern 2ø forest, village 150-200 MHNP Borneo fruit trees 4 July 1982 SabahSoftwoods, Brumas, Sa- Albiziagroves 250 WFVZ bah l Three poorly prepared specimens(2 in the PMNH and 1 in the SM) collected 9 October 1952 on Mt. Matang, Sarawakhave been omitted from this list becausetheir identities are uncertain.They could be either D. agileor D. everetti. bRHNL = Rijksmuseumvan Natuurlikijke Historie, Leiden; USNM = U.S. National Museum;AMNH = American Museum of Natural History; SM = Sarawak Museum; MHNP = Museum d'Histoire Naturelie de Paris;WFVZ = WesternFoundation of VertebrateZoology; PMNH = PeabodyMuseum of Natural History, Yale University.

the Raffles Museum in Singapore, were inac- morphological counterparts in the Indonesian cessibleduring World War II when the major archipelagoand on the Asian mainland. No review work was undertaken, and Chasen, the simple pattern of D. agile radiation could ac- only activeornithologist familiar with the Ma- count for this anomaly. Salomonsendecided layan birds, was killed. that the Philippine forms were derived from Deignan (1945), apparently in responseto D. everettion Borneo,but he placed them in a Chasen'staxonomic arrangement, condemned separatespecies (D. aeruginosum)because of their the splitting of "Piprisoma"on the Asiatic main- distinctivehess from D. everetti. To reflect the land and lumped all mainland forms in P. agile logic of his radiation scheme,however, Salo- modesturn,including by implication sordidurn. monsen chose to unite D. everetti and D. aeru- Then followed a series of important works cit- ginosumas a superspecies. ing Deignan (1945) and discounting the dis- Deignan (1960) investigated the validity of tinct sordidum-everetti-bungurensegroup. Mayr separatingD. everettifrom D. agileby examin- and Areadon (1947) merged all the Asian (in- ing the sole Borneanspecimen of the thick- cluding Philippine) racesinto D. agile,ignored billed flowerpecker complex in the U.S. Na- sordidurn,and called all Malayan forms modes- tionalMuseum. Deignan pointed out •hat this turn, as did Delacour (1946, 1947). specimenhad white tail spots,a fact in direct Salomonsen(1960: 5) departed from the trend contradiction with Salomonsen's assertion that to lump the thick-billed flowerpeckersfor two everettiof Borneohad no tail spots.Salomonsen reasons. First, he realized that the two forms examined only four Bornean specimens--too occurring together on the Malay Peninsula, few to discoverthat D. agilealso occurson Bor- sordidurn and remotum, were distinct. Second, neo; Deignan, by chance,had come acrossone Salomonsen found that the form he examined of the D. agile specimens that Salomonsen from Borneo was "very similar" to sordidurnin missed. Malaya.He believedthat sordidurnderived from Apart from Deignan's initial criticisms,Sal- the Bornean everetti and, therefore, that they omonsen'sproposals generally were accepted should be joined as D. everetti,originally de- (duPont 1971, King et al. 1975, Medway and scribedby Sharpe(1877). Salomonsenretained Wells 1976, Smythies 1981).

remotumin D. agile.This arrangementleft the MATERIALS AND METHODS Philippine races,which have white tail spot- ting, streakedunderparts, and olive backs,geo- Primary data for this paper come from the speci- graphically removed from their most similar men material and field notes collectedby members July 1985] BorneanThick-billed Flowerpeckers 609

T^BLE3. Specimenrecords of Dicaeumeveretti collected in Borneo.

Num- ber Date Locality Museuma Remarks 1 18757 Bintulu, Sarawak BMNH Type (Sharpe) 1877 1 18777 Labuan, North Borneo (Sabah) BMNH Plate in Sharpe 1879 7 1891-1898 Kuching, Sarawak SM, USC -- 2 24 Dec 1893 Kenepai, central Borneo (Kalimantan?) RHNL -- 2 Nov-Dec 1897 Tutong R., Brunei AMNH -- 1 19017 Lawas, northwestern Borneo (Brunei) BMNH -- 1 3 Mar 1983 Binsulok, Sabah WFVZ Adult male in kerangas aBMNH = BritishMuseum of Natural History;SM = SarawakMuseum; USC = Universityof Singapore Collection;RHNL = Rijksmuseumvan NatuurlikijkeHistorie, Leiden; AMNH = AmericanMuseum of Nat- ural History;WFVZ = WesternFoundation of VertebrateZoology. and associatesof the Sabahexpedition and from pre- and feet dark gray. Measurements:wing 57 mm, viouslycollected specimens (see Tables 2, 3). One D. culmen from feathers 7.6 mm, gape 11.4 mm, everettiand 4 D. agileskins were collectedduring tarsus 12.6 mm, tail 29 mm. This description March1981 to September1983. These were measured alsoapplies to the other Borneanspecimens of and comparedwith 15 other D. agileand 13 other D. this species.The sexesare alike. Measurements everettispecimens from Borneo.The Western Foun- dationspecimens also were comparedwith skinsof (n = 14): wing 57.5 + 2.0 mm (54-60), culmen all the other forms in the thick-billed flowerpecker 7.7 + 0.4 mm (6.8-8.4), gape 10.5 + 0.6 mm complexexcept bungurense and atjehense,which are (9.5-11.1), tarsus12.7 + 0.6 mm (12.4-13.5), tail knownonly from 1 specimeneach, stored in the Brit- 28 + 1 mm (25-30). ish and Bogormuseums, respectively (see Table 1). FIELD OBSERVATIONS DESCRIPTIONS AND MEASUREMENTS Few, if any, reliable published accountscan Dicaeumagile.--The BorneanD. agilehas its be found of D. agileand D. everettiin Borneo head and back uniformly olive, with a slightly becauseof their apparentrarity and becauseof yellower rump. The wing covertsand flight the danger of confusingthe two species. feathersare strongly edged with yellowish ol- Dicaeumagile.--Dicaeum agile were observed ive-green. The breast is pale buffy gray mod- in June and July 1982 during a survey in groves erately streakedwith grayishbrown. The tail of the exotic pulp-wood tree, Albizia falcataria, is marked with faint but distinct white spots at Sabah Softwoods near Tawau, Sabah (4ø32'N, on the tips of the inner vanesof the outer three 117ø34'E).The birds always foraged in the can- or four rectrices. Its maxilla is black or dark opy of the older trees(25-30 m) in mixedflocks gray, mandibledark gray with a paler base,iris with sunbirds and other small insectivores. light brown with an orangetinge, and legsand These flocks usually contained from 3 to 10 feet black. The sexesof this speciesare alike. flowerpeckers,but occasionallylarge flocks Measurements[size + SD (range), n = 16]: wing numbered from 15 to 30 flowerpeckers. Di- 60 + 1 mm (57-63), culmen 7.7 ___0.5 mm (7.2- caeumagile fed mainly on the tiny caterpillars 8.5),gape 10.0 + 0.5 mm (9.6-10.8),tarsus 12.4 + (e.g. Euremablanda) that infested the plantation, 0.5 mm (11.9-13.6), tail 30 + 1 mm (28-32). and this abundant food source may have been Dicaeumeveretti.--The single specimenof D. responsiblefor attracting such large numbers. everetti, an adult male, is olive-brown on the Tom Harrisson of the Sarawak Museum (spec- head and back and darker brown on the tail. imen label data) collected a out of a flock The edgesof the primariesare slightly tinged of 30 individuals in a tall fruit tree and another with olive. Underneath, it is light grayish- from the top of a village coconut palm in the brown on its chest and flanks and somewhat Kelabit Plateau of Sarawak. Charles Francis of paler on the throat, center of chest,and belly. the SabahForest Department (pers. comm.) ob- There are no obvious streaks on the breast and served three birds at 40 m in primary forest at no white spots in the tail. Its maxilla is dark Lumerau near Lahad Datu in October 1982 gray, mandible gray, iris dull yellow, and legs (5ø12'N, 118ø52'E).The birds were foraging in 610 FREDERICKH. SHELDON [Auk,Vol. 102

a viny bush in moderately disturbed coastal kerangas,a habitat of podzolized soil dominat- ed by shrubsand stunted trees.It did not sing, but chipped constantlyas it foraged. Its testes were enlarged (5 x 3 ram). D. M. Batchelor(Smythies 1981, pers. comm.) reported seeing D. everettion several occasions in Mawao, Klias, Bongawan, and Lumahat in the southwesterncoastal region of Sabahin the early 1960's.He watched a pair at Lumahat in February 1962 adding to a nest some5-6 m up a tree. "The nest appearedto be a felty pouch, whitish in appearance,which suggestedthat it was built of lalang seed heads or perhaps ka- pok down; it was suspendedfrom the end of one of the outer branchesand was quite inac- Fig. 1. The distribution of the thick-billed flow- cessible"(pers. comm.). Unfortunately, because erpeckercomplex in Sundalandand outlying areas. of the possibility of his confusingspecies, it is Symbolsindicate specificsites of all the known rec- not certain whether Batchelor saw D. everetti or ords for the correspondingraces; square = Dicaeum D. agile.However, the plate in Smythies(1960), agileremotum, diamond = D. a. atjehense,circle = D. a. which was the guide used by Batchelor,is an finschi,triangle = D. everetti(including the Malayan excellent likeness of D. everetti, and the area sordidurnand the Natunan bungurense).Numbers in- where he saw the birds is adjacentto the West- dicate general racial distribution for thoseforms that ern Foundation's collecting site and presents have been more commonlycollected or recorded;1 = similar habitat. D. a. modesturn,2 = D. a. tincture, 3 = D. a. obsoletum, 4 = D. aeruginosumaffine, 5 = D. a. aeruginosum,6 = D. a. striatissimum. Question marks indicate taxonomic RESULTS AND DISCUSSION uncertainty. The discoveryof D. agileon Borneosupports Salomonsen's(1960) separation of D. everetti the tree tops along a river edge, and Francis from D. agile.These are morphologicallydis- noted the diagnostictail wagging (seebelow), tinct species,sympatric in Malaya and Borneo. breaststreaking, and white tail spotting.Pfef- An assessmentof the Bornean D. agile'saffini- fer (1961) reported D. agileas commonin fruit ties and a teevaluationof the subspecificstatus gardensof the village of Long Pesoon the Ka- of the BorneanD. everetti,however, suggest that jan River, easternKalimantan in early Septem- modificationof Salomonsen'sracial taxonomy ber 1956, but he was not able to find them in is required.In addition,the presenceof D. agile similar situationselsewhere. Pearson(1975) re- on Borneo changesthe assumptionsused by corded this speciesas uncommon and occur- Salomonsenin his speculationson the specia- ring in the upper strataand canopyof primary tion of D. aeruginosum.A simpler explanation forest in Kutai Reserve, eastern Kalimantan. of the relationship of this speciesto the Sunda However, he collectedno specimensand pro- birds is now possible. vided no description. Dicaeumagile of Borneois indistinguishable Although it is an obscure,dull brown bird, from D. a. remotum(Robinson and Kloss), 1915, D. agile in Borneo has the unmistakable habit of Malaya and here is assignedto that subspe- of wagging and fanning its tail from side to cies.The Borneanbird also is similar to atje- side. This tail-waggingbehavior (perhapsre- henseand modestum,which are geographically lated to the white tail spotting) makes this juxtaposedto the Malayan remotum(see Fig. 1). speciesthe easiestof the nondescriptflower- In fact, Chasen's(1939b) description of atje- peckersto identify at long distances. henseand Deignan's (1945, 1960) discussionsof Dicaeumeveretti.--We saw D. everettionly once modestum indicate that these races and remotum in 3 years, at Binsulok, Sabah(5ø31'N, 115ø40'E). probably could be lumped into modestum.I have A bird was found foraging on 3 March 1983 in not examined atjehenseand cannot venture a July1985] BorneanThick-billed Flowerpeckers 611 soundopinion on its taxonomy.However, I find when Sundaland was emergent in the Mio- remotum to have a somewhat browner back and Pliocene and early Pleistocene, and again in whiter underparts than modesturnand choose, the late Pleistocene.In particular, the north- therefore, to follow Salomonsen (1960) in western Bornean distribution of D. everetti, a keeping these two races separate. More speci- speciesthat seemsto be restrictedto poor soil mens of Malayan and Sumarranbirds are need- habitats such as kerangas,fits neatly with what ed before these problems can be resolved with is known of the phytogeographyof the region. confidence. The other Sunda races of D. agile Ashton (1972) noted a strong affinity between are easily distinguished from modesturn,remo- the eastern Malayan, Riau, and northwestern tum, and atjehense(see Table 1). Borneanpodzolic soil floras.These floras were The affinity between modesturnand the Bor- once part of a continuum that developedin the nean remotumalso is evidencedby the tail wag- middle and late Pliocene following the Mio- ging and fanning behaviorshared by theseraces Plioceneuplift of the present central Bornean (King et al. 1975). This behavior, which is un- mountains, when the bulk of Sundaland was usual for flowerpeckers,probably plays a role above sea level. All records of D. everetti fit the in speciesrecognition and may relate to the distribution-habitat pattern of the podzolic soil white tail spotting.For this reason,it would be floras(see Fig. 1). In addition, someother Sun- a useful taxonomictool in determining wheth- da endemic birds, such as the Hook-billed Bul- er a race suchas finschi,which lackswhite tail bul (Setorniscriniger) and Gray-breastedBabbler spots,merits specificstatus. Unfortunately, the (Malacopteronalbogulare), approximate this pat- only field records available are those of remo- tern. turn and modesturn. Dicaeumaeruginosum, which has all the defin- Salomonsen(1960) apparentlyrecognized the ing charactersof D. agile,should be joined with establishedD. everettiraces--sordidum, bungu- D. agileas originally proposedby Delacourand rense,and everetti--mainly because of their Mayr (1946) and Mayr and Amadon (1947). The geographicseparation. He stated(1960: 14) that existenceof D. agileon Borneofills a gap in the the specimen of bungurense"differs from the distribution of the thick-billed flowerpeckers other two subspeciesof D. everettionly in being and providesthe biogeographiclink between slightly larger" and that sordidurn"is extraor- D. aeruginosumin the Philippines and D. agile dinarily like the Borneaneveretti." Indeed, there in the Indonesianarchipelago and on the Asian are no substantial differences between sordidurn mainland. and everettiin bill shape,plumage, or soft-part How the present distribution of the thick- colors, and the apparent average size differ- billed flowerpeckercomplex developed is still encesare not verifiably significant becauseof not clear. Salomonsen(1960) proposedthat the the small number of specimens(only four are group radiated from D. agilestock that origi- known). I consider sordidurn Robinson and nated on the Asiatic mainland and dispersed Kloss, 1918, a synonym of nominate everetti through the Malay Peninsula and Sunda Is- Sharpe, 1877. However, becausethere is only lands,eventually reachingthe Philippines.But one specimenof bungurense,which I have not to view the present-day distribution of these examined, it is difficult to assessits racial valid- speciesas the result of an island-hopping in- ity. Its larger size (wing 63 mm, Chasen 1934) vasionis probablytoo simplean approach.Most is not far out of the range for the other sup- speciesof Dicaeumand Prionochilusoccur in the posed races,and it is intermediate between the Philippinesand Sundas,thereby implying that two in geographicposition. On the other hand, the roots of the Dicaeidae lie in these islands living on a small island, it may have diverged (Mayr and Amadon 1947). It is most likely that more quickly than "sordidurn"and everetti. the D. agile-everettiprogenitor originated in The current distribution of D. agile and D. Sundaland or the Philippines and that the ra- everettion Malaya and Borneois consistentwith diation of the group was driven by geographic what is known about Sunda paleogeography. changesin the Tertiary followed by eustaticsea- The Bornean avifauna is most similar to that of level changesin the Quaternary, creating con- Malaya in terms of sharedspecies (Banks 1937), ditions for reiterative isolation and coloniza- and this similarity is attributable largely to the tion. The mainlandforms are probablyoutliers, continentaland drainageconditions prevailing and the rarity of many of the island formsmay 612 FREDERICKH. SHELDON [Auk, Vol. 102

be the result of the disappearanceor spottiness DEIGNAN, H. G. 1945. The birds of northern Thai- of Quaternary habitats. land. U.S. Natl. Mus. Bull. 186: 1-616. 1960. Remarks on the flower-pecker, Di- caeurnagile (Tickell). Bull. Brit. Ornithol. Club 80: ACKNOWLEDGMENTS 142-144. DELACOUR,J. 1946. Notes on the taxonomy of the I thank Patrick Andau, Anthony H. Bledsoe,Jody birds of Malaysia.Zoologica 31: 1-8. Kennard, Kenneth C. Parkes,Eleanor Hß Stickney, 1947. Birds of Malaysia. New York, Mac- and David R. Wells for their advice and help while millan. I prepared this manuscript.I gratefully acknowledge ß & E. MAYR. 1946. Birds of the Philippines. the assistanceof the following museumsand agen- New Yorkß Macmillan. cies for the loan of specimensor permissionto un- DUPONTßJ.E. 1971. Philippine birds. Delaware Mus. dertake research: American Museum of Natural His- Monogr. Set. 2: 1-480. tory, British Museum of Natural History, Museum KING, B. F., M. WOODCOCK, & E. C. DICKINSON. 1975. d'Histoire Naturelie de Paris, Peabody Museum of A field guide to the birds of South-east Asia. Natural History, Rijksmuseumvan Natuurlikijke Londonß Collins. Historie of Leiden, Sarawak Museum, University of MAYR, E., & D. AMADON. 1947. A review of the Di- SingaporeMuseum, U.S. National MuseumßWestern caeidae. Amer. Mus. Novitates No. 1360. Foundation of Vertebrate Zoology, Socio-Economic MIDWAY, LORDß & D. R. WELLSß 1976. The birds of Unit of the MalaysJanPrime Minister's Department, the Malay Peninsula,vol. 5. London,Witherby. SabahChief Minister's DepartmentßSabah Forest De- OBERHOLSER,H. C. 1923. A review of the partment, and SabahSoftwoods Sdn. Bhd. The fund- Prionochilus Strickland and its closest allies. Ohio ing and encouragement for this research came from J. Sci. 23: 287-294. the Western Foundation of Vertebrate Zoology and, PEARSON,D. S. 1975. A preliminary survey of the in particularß from Ed N. Harrisonß Presidentßand birds of the Kutai Reserve, Kalimantan Timur, Lloyd F. Kiff, Director. This is paper no. 10 of the Indonesia. Treubia 28: 157-162. Western Foundation of Vertebrate Zoology Sabah PkEFFER,P. 1961. Etuded'une collectiond'oiseaux Project. de Borneo. Oiseau 31: 9-29. SALOMONSEN,F. 1960. Notes on flowerpeckers (Avesß LITERATURE CITED Dicaeidae).2ß The primitive speciesof the genus Dicaeum. Amer. Mus. Novitates No. 1991. ALI, S., & S. D. R•PLI¾. 1974. Handbook of the birds 1967. Dicaeidae. Pp. 166-208 in Check-list of India and Pakistan, vol. 10. Bombay, Oxford of birds of the world (R. A. Paynter, Ed.). Cam- Univß Pressß bridge, Massachusetts,Museum of Comparative ASHTON,P.S. 1972ß The Quaternary geomorpho- Zoology. logical history of western Malesia and lowland SHARPE, R. B. 1877. Contributions to the ornithol- forestphytogeography. Pp. 35-49 in The second ogy of Borneoßpart 2. Ibis 1: 1-25. Aberdeen-Hullsymposium on MalesJanecology 1879. A list of the birds of Labuan Island (P.S. Ashton and M. Ashtonß Eds.). Aberdeen, and its dependencies. Proc. Zool. Soc. London Univß Hull. 1879: 317-353. BANKSßE. 1937. The distribution of Bornean birds. 1885. Catalogueof the birds in the British Sarawak Mus. J. 4: 453-496ß Museum, vol. 10. London, British Museum (Nat- CHASEN,F.N. 1934. Nine new races of Natuna birds. ural History). Bullß Raffles Mus. 9: 92-97. SMITHI, F. B. 1975. Naturalist's color guide. New 1935. A handlist of MalaysJanbirds. Bull. YorkßAmerican Museum Natural History. Raffles Mus. 11: 1-389ß SMYTHIES,B.E. 1960. The birds of Borneo. London, 1939a. The birds of the Malay Peninsulaß Oliver and Boyd. vol. 4. LondonßWitherby. 1981. The birds of Borneo, 3rd ed. Kuala ß 1939b. Preliminary diagnosesof new birds Lumpur, SabahSociety and Malayan Nature So- from north Sumatra II. Treubia 17: 183-184ß ciety.