The Condor 93:738-745 0 The Cooper Ornithological Society I99 1

PHYLOGENY AND BIOGEOGRAPHY OF THE RIFLEBIRDS (AVES: PARADISAEIDAE)’

BRUCE M. BEEHLER NHB Room 336, Smithsonian Institution, Washington,D.C., 20560

R. J. SWASY 79 Cambridge Terrace,Malvern, South Australia 5061

Abstract. We performed a phylogenetic and biogeographicstudy of the riflebird Ptiloris (Paradisaeidae). Our analysis, which includes sonogramsdocumenting intertaxo- nomic variation in male advertisement vocalizations and a reassessmentof key morpho- logical characters, indicates that (1) the genus includes two distinct clades (P. victoriael paradiseusand the P. magnijicuscomplex); (2) within the P. magnificuscomplex, the eastern Papuan form may constitute a heretofore unrecognizedsibling ;and (3) the bioge- ography of Ptiloris provides additional confirming evidence that the Torres Strait has been of minor importance as a biogeographicbarrier for Australo-Papuan rainforest . Key words: Systematics;biogeography; Ptiloris; Paradisaeidae;riflebirds.

INTRODUCTION Our attention was drawn to the riflebirds as The riflebird genus Ptiloris (Paradisaeidae) in- part of the junior author’s studies of variation in cludes the only group of birds of paradise that avian vocal dialects of specieswith populations has speciatedacross the Torres Strait, the shallow on either side of Torres Strait. Unpublished field water passageseparating from Aus- recordings made by the senior author indicated tralia (Walker 1972, Schodde and Calaby 1972, that analysis of regional variation in Ptiloris vo- Kikkawa et al. 198 1). Riflebirds are stocky, sex- calizations could help clarify understanding of ually dimorphic forest-dwellers that are best riflebird relationships (see Diamond 1972, known for the persistentloud calls given by males Schodde 1974, Cooper and Forshaw 1977, Beeh- from regular display perches (e.g., P. victoriae) ler et al. 1986, Coates 1990). or randomly throughout their forest territories In our preliminary studies we also found that (e.g., P. magnzjcus; Gilliard 1969, Beehler and geographicvariation in the pattern of feathering Pruett-Jones 1983). The speciesare polygynous on the base of the culmen (see Mayr and Rand and promiscuous, the males establishing display 1937, Schoddeand Hitchcock 1968) appeared to courts and the females alone raising offspring corroborate the geographic trends in vocaliza- without assistance(Cooper and Forshaw 1977, tion, and thus constituted an additional feature Diamond 1986). of unrecognized systematic significance. In light Traditional systematictreatments (Mayr 1962, of the recent systematic and biogeographicanal- Gilliard 1969, Cooper and Forshaw 1977) rec- ysis of Australasian passerinesmade by Cracraft ognize three allopatric species:the Magnificent (1986), we believe it is useful to present a revised Riflebird [Ptiloris magnificus (Vieillot)], com- assessmentof the riflebirds, one of Cracraft’s fo- prising three subspecies(P. m. magnzjcus, P. m. cal taxa. alberti Elliot, and P. m. intercedens Sharpe) and We have reassessed:(1) character evolution in inhabiting New Guinea and the Cape York Pen- the genus Ptiloris, (2) the cladistic relationships insula of Australia; Victoria’s Riflebird (Ptiloris in the five regional taxa, and (3) the impact of a victoriaeGould); and the (Pti- revised generic phylogeny on species status of loris paradiseus Swainson). The last two are the Papuan forms and on the biogeographic sig- monotypic and inhabit small rangesin rainforest nificance of the Torres Strait. refuges in eastern Australia (Gilliard 1969). MATERIALS AND METHODS For our vocal analysis, we assembled tape re- I Received7 January 1991. Final acceptance11 April cordings of male advertisement calls in five pop- 1991. ulations of Ptiloris from New Guinea (interce-

t7381 PHYLOGENY AND BIOGEOGRAPHY OF PTZLORIS RIFLEBIRDS 739

FIGURE 1. Sonogramsof Ptiloris populations. A. P. m. alberti from Iron Range, Queensland. B. P. m. magnificus,hwani, West Sepik Province, Papua New Guinea. C. P. victoriae, Paluma, Queensland. D. P. paradiseus,Bunya Mountains, Queensland. E. P. m. intercedens.Morobe Province, Papua New Guinea. dens, magnifcus) and Australia (alberti, morphology, and a single vocal character(Tables paradiseus, and victoriae). The Australian re- 1 and 2). This included a reassessmentof six of cordings were made by RJS, the Papuan record- Cracraft’s characters and 11 additional traits. ings were made by BMB. These were converted Plumage and bill morphology were coded after to sonograms (log scale, wide band) for visual examining a series of five each of male and fe- comparison (Fig. 1) using a Kay Sonagraph in male specimens from collections of the Ameri- the Department of Vertebrate Zoology, National can Museum of Natural History (New York) and Museum of Natural History (Washington, D.C.) National Museum of Natural History. We ex- For cladistic analysis of Ptiloris we developed cluded characters that showed significant intra- a charactermatrix for the five regional taxa (mag- populational variation. ni~%us,intercedens, alberti, paradiseus, and vic- The data in our systematic character matrix toriae) and two designated outgroups ( were subjected to the PAUP branching algo- lawesii and superba). Our choice of rithm, versions 2.4 and 3.0 (Swofford 1985, outgroups follows Cracraft (1986) as well as tra- 1990). We performed series of PAUP runs both ditional determinations of sister-groups to the with characters unordered and ordered. We or- riflebirds (e.g., Schodde 1976, Diamond 1972). dered characters based on putative primitive- We followed the methodology detailed in Cra- derived sequencesdelineated with referenceboth craft (1986). For analysis, we used a series of to the outgroup taxa and the distribution of char- thirteen plumage characters,three aspectsof bill acter states in other related taxa (Kluge 1976). 140 BRUCE M. BEEHLER AND R. J. SWABY

TABLE 1. Character states for the genus Ptiloris. Characterswere weighted equally, preventing us from unconsciously skewing the analysis by em- CHARACTER ONE: male advertisement vocal- phasizing characters that support personal bi- ization-(O) one or more harsh notes;(1) two or several ases. Because our analysis involved only seven loud growl notes, (2) several upslurred, musical wolf taxa, using PAUP version 2.4 we were able to whistles (see sonograms,Fig. 1). use the “alltrees” mode which produces an ex- CHARACTER TWO: feathering on base of cul- men of male-(O) extended anteriorly, (1) small portion haustive search of all possible branching com- of base unfeathered, (2) unfeathered ridge deep and binations. M. C. M&trick ran the characterma- expanded posteriorly. trix for us on PAUP version 3.0 using the “branch CHARACTER THREE: iridescentbreast-shield of and bound” mode. male- (0) winged, (1) small, unwinged, (2) large, unw- inged. We plotted distributional recordsfor the three CHARACTER FOUR. black velvety featherspos- allotaxa in the magnijicus complex using data terior to breast shield of male-(O) absent, (1) forming from the literature (Schoddeand Hitchcock 1968, a narrow band, (2) forming a broad band. Gilliard 1969, Diamond 1972, Cooper and For- green velvet feathering pos- CHARACTER FIVE: shaw 1977, Coates 1990); from specimen records terior to breast shield of male-(O) absent, (1) narrow, (2) extensive. from the American Museum of Natural History, CHARACTER SIX: flank plumes- (0) absent, (1) National Museum of Natural History, and Aus- very brief, (2) subequal or equal to length of tail, (3) tralian National Wildlife Collection; and from longer than tail. unpublished distributional recordsbased on voice CHARACTER SEVEN: occipital plumes-(O) brief, (1) elongated. (mostly of BMB). Our map showing distribution CHARACTER EIGHT: central rectrices-(0) vel- of the entire genus is based on the above sources vety purple-black, (1) blue-purple, (2) iridescentgreen. plus data from MacDonald (1973) and Blakers CHARACTER NINE: female ventral pattern-(O) et al. (1984). fine dark barring, (1) dark scalloping,(2) dark spotting. CHARACTER TEN: female dorsal plumage col- or-(O) gray-brown, (1) medium brown, (2) reddish brown. RESULTS CHARACTER ELEVEN: pale streakingon crown CHARACTER ANALYSIS of female-(O) prominent, (1) reduced, (2) much re- duced or absent. Our analysis of vocalizations (Fig. 1) shows that CHARACTER TWELVE: male flight feathers-(O) (1) Ptiloris victoriae and P. paradiseus share the unspecialized, (1) blunt-tipped, sexually distinct, (2) same coarse yaaaas call; (2) the regional taxa with emarginate outer primaries. CHARACTER THIRTEEN: upper throat pattern alberti and magnijkus sharethe upslurred “wolf- of female-(O) with fine dark speckling,(1) unspeckled. whistle” call unlike that of any other riflebird. CHARACTER FOURTEEN: Transcribed as woiiieet- woit!, the call is usually in bill length-(O) female’s shorter than or nearly equal two notes but on occasion as many as four in a to the male’s, (1) bill longer than male’s, series; and (3) the growled voice of intercedens CHARACTER FIFTEEN: rectrix shape of male- (0) narrow and typical, (1) broad and blunt-tipped. [uRAUow-urauow] sharessome characteristics CHARACTER SIXTEEN: male wing length(arc)- with that of P. victoriaelparadiseus but is dis- (0) less than 160 mm, (1) more than 180 mm. tinct, and was classifiedas an intermediate char- CHARACTER SEVENTEEN: bill shape of fe- acter state. There is evidence of slight vocal vari- male-(O) short and straight, (1) elongatedand slightly decurved, (2) elongatedand much decurved. ability within populations (Schodde in litt., for

TABLE 2. Character matrix for Ptiloris.

characters TaXa I 2 3 4 5 6 7 8 9 10 II 12 13 14 15 16 17

Parotia Zawesiit 0 0 1 0 0 2 1 0 0 1 2 2 0 0 1 0 0 Lophorinat 00000010000000000 P. victoriae : 112 2 10 2 2 0 0 11110 1 P. paradiseus 1 1 2 2 1 2 0 0 1 1 1 0 2 P. alberti ; 2 2 1 1 3 : 1: 11 11; 111 P. magnificus 2 2 1 1 3 : 1 0 22: 100111 P. intercedens 1 1 2 1 1 2 10 110 111 PHYLOGENY AND BIOGEOGRAPHY OF PTZLORZS RIFLEBIRDS 741

magi

paradiseus

victoriae

B P T I L 0 R I S

victoriae paradiseus alberti magnificus intercedens

Parotia

tree length = 33 steps

FIGURE 2. (A) Geographic distribution of the five basal taxa of Ptiloris in New Guinea and Australia, with postulated vicariance barriers correspondingto the numbered branching events in the cladogram. (B) Phylogenetic hypothesis for the genus Ptiloris, rooted to the outgroupsParotia lawesii and Lophorina superba.Characters uniquely defining each node, numbered l-4, are: (1) chars. 7, 15; (2) chars. 3, 4, 5, 6, 14; (3) chars. 3, 4, 5, 8, 16; (4) chars. 1, 2, 6, 11. Steps and consistencyindex (c.i.) based on unordered characters. C.i. generated with uninformative charactersdiscounted.

P. m. alberti), but not to the extent that our clas- One plumage character, female dorsal color sification of character states would be altered. (character 11 of Cracraft 1986; our character 10) The pattern of culmen feathering showed a may be the product of character convergencere- trend from heavily feathered anteriorly (primi- lated to the regional presence or absence of in- tive) to unfeathered along the ridge of the culmen terspecific flocking. This character distinguishes (derived). P. m. magnijicus and P. m. alberti the two Papuan populations (russet-backed fe- shared the derived condition, and the other five males) from the three Australian populations Ptiloris tended toward the primitive condition (gray-brown-backedfemales). We believe that the exhibited by the outgroups Lophorina and Par- evolution of female dorsal plumage has experi- otia. We are uncertain of the functional impor- enced regional plumage convergence, influenced tance of this feathering pattern in Ptiloris, al- by the considerable importance mixed foraging though it seemsto be of importance to courtship flocks in New Guinea and their absencein Aus- display in the genus Parotia (Schodde and Mc- tralia. In New Guinea, there is evidence of strik- Kean 1974). ing plumage convergenceamong speciesthat 142 BRUCE M. BEEHLER AND R. J. SWABY

Ptiloris “magnificus”

I intercedens

magnificus/alberti clade

ltercedens

I

FIGURE 3. Geographic distribution of the Ptilorismugnijicus complex in New Guinea and Australia. Black circles = mugnz@cus;open circles = intercedens;black triangles = alberti.The easternmostblack circle marked “a” representsa single vocal record of a male giving the mugnificus-typeadvertisement call (Coates 1990). The circled dot indicated as “b” is location of the putative intermediate specimen (see text).

forage together in mixed flocks. These are Dia- and P. m. intercedens.Within the latter clade, P. mond’s “brown and black flocks,” which include m. intercedens,the population from eastern New birds of paradise, pitohuis, drongos, honeyeat- Guinea, comprisesthe sister-form to the albertil ers, and assortedother species(Diamond 1987). magnificus pair (Fig. 2). With the characters Although we propose that the taxonomic dis- unordered, the tree comprises 33 steps, with a tribution of female dorsal color may be the prod- consistency index (excluding uninformative uct of convergence,we still include the character characters)of 0.893. With all charactersordered, in our matrix for two reasons. It allows for the the tree is of 37 stepsand has a consistencyindex possibility that we are wrong in our assumption (excluding uninformative characters) of 0.8 11. about convergence, and it permits us to test the Both constituted the sole most parsimonious of robustness of our novel branching pattern. 935 trees compared in each PAUP run, and there Our other characterswere based on a revision was no difference in branching order. and expansion of Cracraft’s character list, and Our mapping of the distribution of the three depends to a greater degreeon those male plum- allotaxa of the magnz$cuscomplex shows that age traits that are related to courtship display. In intercedens differentiated as an eastern isolate addition, we focus on traits that require straight- (Fig. 2a, 3). Thus the main body of New Guinea, forward objective measurement (length or shape) western New Guinea and northernmost Queens- rather than minor variation in color or irides- land comprise the range of the magnijicuslalberti cence (Cracraft’s characters7, 8, 10, 12). We be- clade, and eastern New Guinea-primarily the lieve we have based our character matrix on a Papuan peninsula-comprises the home of in- detailed knowledge of the biology of Ptiloris, tercedens. based on extensive experience with the genus in Because there are no Ptiloris populations in the field. the high northwest-to-southeast trending cordil- lera of New Guinea, there are two contact zones CLADISTICS AND GEOGRAPHY where intercedensand its sister-form meet, one Our cladistic analysis produced a phylogenetic in the northern watershed (corresponding to the tree that pairs P. paradiseuswith P. victoriaeand Ramu River) and another in the southern wa- forms a triplet of P. m. alberti, P. m. magn$cus, tershed (corresponding to the Purari River). PHYLOGENY AND BIOGEOGRAPHY OF PTZLORZSRIFLEBIRDS 143

The populations of basal magniJicusand al- cedens)shares few spectrographicsimilarities with berti,situated on either side of Torres Strait, show the musical, upsweeping wolf whistles of mag- only minor levels of morphological differentia- nijicus,from central and western New Guinea. tion, and their vocalization and culmen-feath- The contact zone of vocal dialects occursin both ering character statesare equivalent (Fig. 1, Ta- the northern and southern watersheds. What, ble 2). then, is its significance? Given that the species is promiscuous and that males compete for the DISCUSSION right to mate with females, we supposethat vo- The difference between Cracraft’s (1986) and our calization may be of considerable biological im- cladogram is the branching within the “magni- portance. Notable, too, is that the call itself is ” clade. Whereas Cracraft found albertito very simple (Fig. 1) and that there is little or no individual or regional variation (Gilliard 1969, be the outgroup of that triplet, we found inter- Beehler et al. 1986, Coates 1990). We doubt that cedensto be the outgroup. Three charactersnot either the growl or the musical whistle could be used by Cracraft and which support our revised easily derived from the other by frequency mod- branching were male voice, culmen feathering, ulation of the syrinx. and plume length. Because Cracraft’s character matrix and its coding differed from ours it is not Becauseof the overall morphological similar- ity between the two Papuan populations, it is not valid to grade the resulting cladograms by level clear whether any ecological differentiation has of homoplasy. Nonetheless, our better-resolved tree employs new charactersthat are important yet developed between the eastern and western features of the biology of the group, and thus we allotaxa. There are no important differences in believe the branching shown in our cladogram measurements between the two Papuan popu- is a closer approximation of the phylogenetic his- lations (Beehler, unpubl.). tory of Ptiloris. The east-west boundary of the two Papuan forms corresponds to two major distributional SPECIES STATUS OF ALLOTAXA barriers, the Ramu and Purari river basins. (The Cracraft (1986) used the “basal taxon” or “phy- record of intercedensfrom Lake Kutubu, west of logenetic species” concept for morphologically the Purari break, is in error [Schoddeand Hitch- distinct allotaxa. We agree that this is necessary cock 1968, Schodde, pers. comm.].) for the purposesof cladistic analysis, but in some Coates (1990) reported hearing the western di- instances the resulting nomenclature is not con- alect near Yalumet, east of the Ramu, and near cordant with current ornithological usage.Those to where W. S. Peckover heard the easterndialect who employ the “biological species concept” (seeFig. 3). This implies that the two forms may (sensuMayr 1963) may wonder which of the five be able to co-exist. In addition, Schodde(Coates basal taxa treated in our analysis merit full spe- 1990) reports a specimen with intermediate cul- cies status in the Mayrian sense.We suggestthat men feathering characteristics(a possiblehybrid) Ptilorisparadiseus and P. victoriaeare closely at Futei, just east of the Purari barrier. It is ob- related sisterforms that each merit speciesstatus. vious, then, that more fieldwork will be required The difference in female bill shape implies the to determine the nature of the interaction be- evolution of distinct dietary or foraging habits. tween the two Papuan forms. Is contact leading The situation with the Ptilorismagn$cus com- to coexistence, localized introgression, or inter- plex is less clear. Certainly the populations mag- specificterritoriality? Are there populations with nificusand albertiare conspecific, but the para- “hybrid” vocal characteristics?This would be patric disposition and vocal distinctness of relatively simple to determine by a field survey. intercedensand magnificusmight suggesttreat- We urge field workers to examine populations of ment of intercedensas a distinct sibling species. Ptilorisin the vicinity of the Ramu and Purari basins in order to settle this question. PAPUAN DIFFERENTIATION The initial phenomenon that focused our atten- TORRESIAN BIOGEOGRAPHY tion on Ptiloris’was the strikingly different vo- Our revised phylogeny suggestsa minor rework- calizations of male riflebirds in eastern and west- ing of Cracraft’s (1986) general biogeographichy- central New Guinea (Figs. 1, 3). The guttural pothesisfor the Australasian avifauna. As earlier double-growl of the eastern population (inter- postulated by Schodde and Calaby (1972) and 744 BRUCE M. BEEHLER AND R. J. SWABY

Kikkawa et al. (198 l), the importance of the Tor- vation, for permission to observe and record birds in res Strait as a vicariance barrier should be con- Papua New Guinea. Part of the data nresented herein was collected while conductingfield studies supported sidered minor. It appears that phylogenetic dif- by the Smithsonian Institution. For assistanceand ferentiation in taxa that range from the Papuan guidanceof various kinds while in PapuaNew Guinea, mainland southward to the eastern rainforests of we thank K. Kisokau, I. Ila, N. Kwapena, N. Kawa, Australia has been more heavily influenced by A. Balo, M. Yamapao, E. Broom, G. Velai, M. Lucas, habitat barriers between the three eastern rain- R. Goga, B. Finch. and the staff of the Wau Ecoloev Institute: M. C. McKitrick ran our charactermatrix &I forest refuges within Australia than between PAUP 3.0. J. T. Marshall taught BMB how to make Australia and New Guinea (Fig. 2). In Ptiloris, the sonogramsand provided much useful advice on the major bouts of speciation apparently took the subject.We thank the Bird Division, National Mu- place across continental nonforest habitat bar- seum of Natural History, and the Department of Or- nithology, American Museum of Natural Historv. for riers (northern New South Wales/east-central permissionto study bird specimensin their care. Valu- Queensland/northern Queensland for paradi- able criticisms of earlier drafts were provided by P. F. seuslvictoriaelalberti).Differentiation within the Cannell, J. Cracraft, J. A. Endler, G. R. Graves, A. R. magnificusclade is only minor acrossTorres Strait Keast, M. C. McKitrick, S. L. Olson, R. Schodde,and (at the racial level), and is overshadowed by the R. A. Simpson. degreeof east-west differentiation on the Papuan mainland. Acknowledging that New Guinea and LITERATURE CITED northern Australia were linked by a broad cor- BEEHLER,B. M., T. K. PRATT,AND D. A. ZIMMERMAN. ridor of dry land during much of the Pleistocene 1986. Birds ofNew Guinea. PrincetonUniv. Press. (Walker 1972, Kikkawa et al. 198 l), the geo- Princeton, NJ. graphic differentiation of the Ptiloris magnificus BEEHLER,B. M., ANDS. G. PRuET~-JONES.1983. Dis- complex in Australia and New Guinea appears play dispersion and diet of birds of naradise-a comparison of nine species. Behav. Ecol. Socio- relatively straightforward.We envision that there biol. 13:229-238. existed a continuous Australo-Papuan distribu- BLAKERS,M., S. DAVIES, AND P. REILLY. 1984. The tion of the parent (ancestral) form, followed first atlas of Australian Birds. Melbourne University by a vicariance event across the Ramu/Purari Press, Carlton, Victoria, Australia. COATES,B. J. 1990. Birds of Panua New Guinea. the barrier (step 3 on Fig. 2), followed by a subse- passetines. Vol 2. Dove Publications, Aldkrlv,__ quent break of the trans-Torresian populations Queensland,Australia. in the late Pleistocene (step 4 in Fig. 2). COOPER,W., AND J. FORSHAW. 1977. The birds of East-west differentiation on the Papuan main- paradise and bowerbirds. Collins, Sydney, Aus- land is widespread in birds and other vertebrates tralia. CRACW, J. 1986. Origin and evolution of conti- (see, for example Mayr 1942, Schodde and nental biotas: speciationand historical congruence Hitchcock 1972, Diamond 1972, Flannery 1990), within the Australian avifauna. Evolution 40:977- and appears to outweigh the geographic differ- 996. entiation that has occurredacross the Torres Strait DIAIVIOND,J. M. 1972. Avifauna of the eastern high- (Schoddeand Calaby 1972, Kikkawa et al. 198 1, lands ofNew Guinea. Publ. Nuttall Omithol. Club No. 12. Beehler, unpubl. analysis of forest birds). The re- DLWOND, J. M. 1986. Biology of birds of paradise markable genetic homogeneity of the tree Acacia and bowerbirds. Ann. Rev. Ecol. Syst. 17:17-37. mangium on either side of Torres Strait (Moran D~OND, J. M. 1987. Flocks of brown and black et al. 1989) corroborates this pattern. As indi- New Guinean birds: a bicolored mixed speciesfor- aging association.Emu 87:20 l-2 11. cated by the pattern of speciation in the riflebirds FLANNERY,T. 1990. Mammals of New Guinea. Rob- (Fig. 2) for Australasian rainforest birds geo- ert Brown Assoc., Carina, Queensland,Australia. graphic differentiation has occurred primarily GILL-, E. T. 1969. The birds ofparadise and bow- between humid forest refugesin eastern Austra- er birds. Weidenfeld and Nicolson, London. lia and in an east-west pattern in New Guinea KXKAWA, J., G. B. MONTEITH,AND G. INGRAM. 1981. Cape York Peninsula: maior region of faunal in- (Diamond 1972). Torres Strait is a modem fea- terchange, 1697-1742. ZnA. R.-Keast [ed.], Eco- ture that has been of minor importance in this logicalbiogeography of Australia. Vol. 3. W. Junk, process. The Hague, The Netherlands. KLUGE, A. -1976. Phylogenetic relationships in the ACKNOWLEDGMENTS Pygopodidae.Misc. Publ. Museum of Zool. Univ. Michigan, no. 152. Ann Arbor, MI. We thank the Government of Papua New Guinea, es- .__ ~ MACDONALD,^- ___..J. _D. _1973. _ Birds of Australia. H. F. pecially the Department of Environment and Conser- L ti. Witherby, London. PHYLOGENY AND BIOGEOGRAPHY OF PTZLORZS RIFLEBIRDS 145

MAYR, E. 1941. List of New Guinea birds. American and cultural history of Torres Strait. Australian Museum of Natural Historv. New York. Nat. Univ., Canberra, Australia. MAYR, E. 1962. The Paradisaeidae,p. 181-203. In SCHODDE,R., AND W. B. HITCHCOCK. 1968. Contri- E. Mayr and J. C. Greenway [eds.], Checklist of butions to Papuasian ornithology. I. Report on birds of the world. Museum of Comparative Zo- the birds of the Lake Kutubu area, Territory of ology, Cambridge, MA. Papua and New Guinea. Wildl. Res. Tech. pap. MAYR, E. 1963. speciesand evolution. Har- no. 13. C.S.I.R.O., Melbourne, Australia. vard University Press, Cambridge, MA. SCHODDE,R., AND W. B. HITCHCOCK.1972. Birds, p. MAYR, E., AND A. L. RAND. 1937. Results of the 67-86. In P. A. Ryan [ed.], Encyclopaediaof Pa- Archbold Expeditions. 14. Birds ofthe 1933-1934 pua New Guinea. Vol. 1. Melbourne Univ. Press, Papuan Expedition. Bull. Amer. Mus. Nat. Hist. Carlton, Victoria, Australia. 73:1-248. SCHODDE,R., ANDJ. L. MCI&AN. 1974. Distribution, MORAN, G. F., 0. MUONA, AND J. C. BELL. 1989. , and evolution of the gardener bower- Acacia mangium: a tropical foresttree of the coast- birds Amblyornisspp. in easternNew Guinea, with al lowlands with low genetic diversity. Evolution descriptions of two new subspecies.Emu 735 l- 43:213-235. 60. SCHODDE,R. 1974. Review of ‘avifauna of the east- SWOFFORD,D. L. 1985. PAUP: phylogeneticanalysis ern highlands of New Guinea.’ Wilson Bull. 86: using parsimony, version 2.4, program manual. 183-186. Illinois Natural History Survey, Champaign, IL. SCHODDE,R. 1976. Evolution in the birds-of-para- SWOFFORD,D. L. 1990. PAUP: phylogeneticanalysis dise and bowerbirds, a resvnthesis.Proc. XVI Int. using parsimony, version 3.0, program manual. Omithol. Congr. (1974):137-149. Illinois Natural History Survey, Champaign, IL. SCHODDE,R., AND J. H. CALABY. 1972. The bioge- WALKER,D. [ED.] 1972. Bridge and barrier: the nat- ographyof the Australo-Papuanbird and mammal ural and cultural history of Torres Strait. Austra- faunas in relation to Torres Strait, p. 257-300. In lian Nat. Univ., Canberra, Australia. D. Walker [ed.], Bridge and barrier: the natural