GENETIC AND MORPHOLOGICAL DIFFERENTIATION AND PHYLOGENY IN THE AUSTRALO-PAPUAN SCRUBWRENS (, )

LESLIE CHRISTIDIS,1'2 RICHARD $CHODDE,l AND PETER R. BAVERSTOCK 3 •Divisionof Wildlifeand Ecology, CSIRO, P.O. Box84, Lyneham,Australian Capital Territory 2605, , 2Departmentof EvolutionaryBiology, Research School of BiologicalSciences, AustralianNational University, Canberra, Australian Capital Territory 2601, Australia, and 3EvolutionaryBiology Unit, SouthAustralian Museum, North Terrace, Adelaide, South Australia 5000, Australia

ASS•CRACr.--Theinterrelationships of 13 of the 14 speciescurrently recognized in the Australo-Papuan oscinine scrubwrens, Sericornis,were assessedby protein electrophoresis, screening44 presumptivelo.ci. Consensus among analysesindicated that Sericorniscomprises two primary lineagesof hithertounassociated : S. beccarii with S.magnirostris, S.nouhuysi and the S. perspicillatusgroup; and S. papuensisand S. keriwith S. spiloderaand the S. frontalis group. Both lineages are shared by Australia and . Patternsof latitudinal and altitudinal allopatry and sequencesof introgressiveintergradation are concordantwith these groupings,but many featuresof external morphologyare not. Apparent homologiesin face, wing and tail markings, used formerly as the principal criteria for grouping species,are particularly at variance and are interpreted either as coinherited ancestraltraits or homo- plasies. Distribution patternssuggest that both primary lineageswere first split vicariantly between Australia and New Guinea, and then radiated independently on each land massunder the influence of paleoclimaticchange. Dispersal between Australia and New Guinea is indicated only in the magnirostrissublineage and is either very recent or just broken. Received13 April 1987, accepted18 April 1988.

WItH thornbills () and Although this reorganization was controver- (),scrubwrens of the Sericornis sial, it has not been subjectedto testing other Gould form one of three major assemblagesof than by brief and conflicting comparisonsof species within the Australo-Papuan family morphological and behavioral traits (Keast Acanthizidae. In its strictest sense (RAOU 1978a, b; Schodde 1981). Checklist Committee 1926; Mayr 1986), the We studied the relationships among the group comprisessmall, brown, thin-billed in- speciesof Sericornissensu Mayr (1986) as deter- sectivoresthat live sedentarilyin pairs or small mined by allozymicvariation at 44 presumptive communal groups in the middle- and under- loci. The monophyly of this group, corrobo- strata of subtropical and temperate forests in rated by DNA-DNA hybridization (Sibley and montane New Guinea and coastal Australia Ahlquist 1985), is not in question.Material of (Figs. 1 and 2). all but one specieswas examined, but we pre- General agreement on the compositionand sentdata for only 11of the 14.Of the 3 excluded relationships of Sericorniswas reached in the or missing,the TasmanJanScrubtit (S. magnus) revisionsof Mayr (1937),Mayr and Rand (1937), was found to be unrelated to the core lineages Mayr and Serventy (1938), and Keast (1978a). of Sericornis(Christidis and Schodde in prep.); Their conceptsof the genus were maintained the (S. maculatus)is not ac- in essenceby Mayr (1986), who accepted 14 cepted here as a speciesbecause it intergrades species.Six are endemic to New Guinea, seven with the White-browed Scrubwren(S. frontalis; endemic to Australia and one (S. beccarii) is cf. Condon 1951) and shows no significant al- shared(Figs. 1 and 2). Schodde(1975) expanded lozymic differentiationfrom it (Christidisand Sericornisto include 5 allied mono- and di-typic Schoddein press);and tissuesof the Vogelkop generic groups from drier sclerophylloushab- Scrubwren (Sericornisrufescens) were unavail- itats in Australia (cf. Schodde and McKean 1976). able. 616 The Auk 105: 616-629. October 1988 October1988] BiochemicalSystematics ofScrubwrens 617

Fig. 1. Geographicaldistribution of taxa of the Sericornismagnirostris primary lineage.

Fig. 2. Geographicaldistribution of taxa of the METHODS Sericornisfrontalis primary lineage. Liver, breast muscle and heart tissues, stored in liquid nitrogen,were examinedfrom 127individuals sideredclosest to coresericornine lineages (Meise 1931, of 11 species,together with thoseof 6 geographically Mayr 1937),but falls outsidethe sericornineassem- spreadindividuals of the BrownThornbill (Acanthiza blage.In-group comparisonswere also made within pusilia).Geographically isolated subspeciesof the the sericornineassemblage of Schodde(1975), but Large-billed Scrubwren (S. m. magnirostris,S. m. viri- becausethey did not affect the monophyly of Seri- dior) and Yellow-throated Scrubwren (S.c. citreogu- cornisas defined here, they are reported elsewhere laris,S.c. cairnsi)were included to provide a frame for (Christidisand Schoddein prep.). differentiation across all taxonomic levels. The sam- Morphometricand qualitativemorphological data plesand their collectionlocalities are itemized in Ta- for comparisonwere collated from specimensin the ble 1. Australian National Wildlife Collection, Canberra Individuals were screened electrophoretically for (ANWC), augmentedwith material from the Queens- 37 enzymesystems representing 44 presumptiveloci land Museum, Brisbane (QM), Australian Museum, (Christidisand Schoddein press).Initially, represen- Sydney(AM), and,for S.rufescens, American Museum tatives of 5 diverseacanthizid genera, including Ser- of Natural History, New York (AMNH). Information icornis,were analyzed to determine optimal condi- on altitudinal distribution in New Guinea was drawn tions for each enzyme system. It allowed us to from specimenrecords in ANWC and AM, and from maximize resolution of the allelic bands and so detect banding records in the Australian Banding most of the variation in our subsequentscreening of Scheme, Australian National Parks and Wildlife Ser- Sericornis. Genetic distances between taxa were esti- vice, Canberra. matedby the measuresof Nei (1978)and Rogers(1972), from which UPGMA phenograms(Sneath and Sokal RESULTS 1973) and distance-Wagnertrees (Farris 1972) were constructed.A further analysis of fixed (nonpoly- Allelic frequencies,heterozygosities, and genetic morphic) differencesamong electromorphswas per- distances.--Of the 44 loci screened, 12 were formed by the cladisticmethod of Hennig (1966). monomorphic in all species(E.C. numbers in Highly polymorphicloci and low frequencyallelo- parentheses):Got-2 (2.6.1.1), Idh-2 (1.1.1.42),Tpi morphsshared across lineages were culled, to reduce (5.3.1.1), Sod(1.15.1.1), Fum (4.2.1.2), Ak (2.7.4.3), confusion from unclarified plesiomorphic statesand Ck-3(2.7.3.2), Gdh (1.4.1.3), Gpt-2 (2.6.12), Acon-1 convergence(cf. Pattonand Avise 1983).Accordingly (4.2.1.3), Mdh- 1 and Mdh-2 (1.1.1.37). A1lelic fre- Gpdh,Est-1, Est-3, and the low frequencyallelomorphs quenciesat the other 32 varying loci are given at PepL-Pro and Acon-2were excluded.The rationale and modes of these procedureswere outlined in in Table 2, excluding thoseof 5% or less. For Christidis (1987). the species of Sericornis,mean heterozygosity Acanthizapusilia was usedto root the Wagner tree calculatedby the directcount estimate (Nei 1975) and asthe out-groupfor cladisticanalysis (Patton and was 0.035, and the averageproportion of poly- Avise 1983) becauseit is a member of the genuscon- morphicloci 18.4%(Table 2). The mean genetic 618 CHRISTIDIS,SCHODDE, AND BAVERSTOCK [Auk,Vol. 105

TABLEi. Localitiesand samplesizes (n) of speciesexamined. Voucher specimensin AustralianNational Wildlife Collection,CSIRO, Canberra;and in AustralianMuseum, Sydney.

Symbola Locality n Sericornisperspicillatus SPE Efogi, 13 S. arfakianus SAR Kaftmui and Taft, Papua New Guinea 4 S. magnirostrismagnirostris SMA KangarooValley, New South Wales 2 Cambridge Plateau, New South Wales 6 Conondale Range,Queensland 3 Dawes Range,Queensland 1 Clarke Range,Queensland 2 S. m. viridior SMV Atherton, Queensland 4 Helenvale district, Queensland 1 S. beccarii SBE Silver Plains, Cape York Peninsula, Queensland 3 McIlwraith Range, Queensland 6 Tari district, Papua New Guinea 3 S. nouhuysi SNO Efogi, Papua New Guinea i0 S. papuensis SPA Efogi, Papua New Guinea 3 S. spilodera SSP Tari district, Papua New Guinea 2 S. citreogulariscitreogularis SCI KangarooValley, New South Wales 2 CambridgePlateau, New South Wales 6 Conondale Range, Queensland 3 S.c. cairnsi SCC Atherton, Queensland 6 S. keri SKE Atherton, Queensland 2 Mt. Lewis, Queensland 4 S. humills SHU Tasmania b 6 S. frontalis SFR South-eastern Australia b 28 Acanthizapusilla APU Eastern Australia, Tasmania 6 Symbolsused in Figs. 4 and 5. A completelisting of localitiesis given in Christidisand Schodde(in press). distance (Nei 1978) among all 11 specieswas Geneticdifferentiation.--The UPGMA pheno- 0.197 (SD 0.068, SE 0.001); and that between grams derived from both Rogers' (1972) and them and Acanthizapusilla was 0.345 (SD 0.054, Nei's (1978) distanceswere identical and sep- SE0.005), indicating substantial differentiation. arate Sericornisinto 5 groups of species,all of In comparison,the averageinterspecific and in- which are closer to one another than to Acan- tergenericdistances in Parulidae, Emberizidae, thiza pusilla(Fig. 3). The first group comprises Muscicapidaeand Icteridae were 0.10 and 0.26 the Pale-billedScrubwren (S. spilodera) and Pap- respectively(Avise and Aquadro 1982). uan Scrubwren (S. papuensis),between which As expected,genetic distanceswere greater the distancewas greater than any among the among speciesthan subspecies(Table 3), but other 4 groups.Thus these2 specieswere clus- several values were contradictory. Nei's (1978) tered because of their distinctiveness from re- distanceswere only 0.046between the allospe- maining taxa.The other 4 groupswere: S. citreo- cific Brown and White-browed scrubwrens (S. gularis;New GuineanS. perspicillatus-S. arfakianus; humills,S. frontalis),and were 0.02-0.03 among S. magnirostris-S.beccarii-S. nouhuysi; and Austra- the Australian Large-billed Scrubwren (S. m. lian S.frontalis and its allies.The lastassemblage magnirostris),morphologically dissimilar Trop- comprisedS. frontalis,the TasmanianS. humills, ical Scrubwren (S. beccarii),and New Guinean and the (S. keri), and was (S. nouhuysi).In contrast,the closestto the magnirostriscluster. In the magni- pairsof subspeciesof the Large-billedand Yel- rostrisgroup, S. m. viridiorof northeastQueens- low-throated(S. citreogularis) scrubwrens, which land was the most distant member allozymical- are almostidentical in appearance(Mayr 1937), ly (seeabove). differed by distancesof 0.032 and 0.036. Com- Phylogeneticclustering.--Derived from values pared to measureswithin other avian species for Rogers' (1972) genetic distance, the dis- (Corbin 1977, Avise and Aquadro 1982), these tance-Wagnertree (Fig. 4) is rooted by Acan- valueswere rather high and equivalent to pro- thizapusilla. Unlike the UPGMA procedure,the tein differentiation at specieslevel in parulid Wagner algorithm does not assumea constant warblers (Avise et al. 1980). rate of protein evolution, and this is reflected October1988] BiochemicalSystematics ofScrubwrens 619 in thediffering branch lengths among taxa (Figs. amongthe 5 clusters.Here the alignmentof S. 3 and 4). It implies, for example,that S. citreo- magnirostriswith the S. frontalisgroup in the gulariscairnsi in northeast Queenslandhas UPGMA phenogramwas distorted because ple- undergonea 2.5-fold fasterchange in proteins siomorphieslinking them were not eliminated since divergencethan its allosubspecies,S.c. by out-group comparison.The alternative ar- citreogularis,in central east Australia. Despite rangementcorroborated by both distance-Wag- this basicdifference, the Wagnernetwork iden- ner and cladisticanalysis is the morelikely phy- tified the same 5 clustersof speciesas in the logeny.It indicatestwo primary lineagesin the pheneticanalysis. One major discrepancywas genus.In one, the close-knitmagnirostris lin- the positionof the S.frontalis assemblage which eage,the S.perspicillatus and S. magnirostris groups was aligned most closelywith S. citreogularis,form two sublineages.The other, the frontalis leaving the Buff-facedScrubwren (S. perspicil- lineage,is more diffuse.It includesS. keri as a latus)and Olive Scrubwren(S. arfakianus)as the closeally of S. frontalis,and S. citreogularisand sister-groupof the magnirostrisassemblage. S. papuensisgroup as more divergent sublin- Cladistic analysisdistinguished further be- eages. tween ancestral and derived states among the Plumagepatterns, hitherto the principal cri- allozymes(Fig. 5). Plesiomorphieswere iden- teria for grouping the speciestaxonomically, tified from comparisonwith Acanthizapusilla, are at variancewith this phylogeny.By consen- and the synapomorphiesdefining each clade sus,they align the speciesin two different clus- are listed in Table 4. They corroboratedonce ters(Mayr 1937,1986; Galbraith and Parker1969; more the 5 terminal clustersidentified by both Parker 1970; Keast 1978a; and Diamond 1985). UPGMA and distance-Wagneranalyses and split In one,sharing plain plumagewith red iris and them into two primary lineages. On one side unbarred or faintly barred tail, 2. papuensis,2. of the split were the S. perspicillatusand S. mag- spilodera,and 2. keriare groupedwith the per- nirostrisgroups; the remainderwere on the oth- spicillatusand magnirostrissublineages. In the er. A closeconnection between the perspicillatus other,with seeminglyhomologous white marks and magnirostrisgroups was supportedby the on face and wing coverts and sexual dimor- Wagner tree (Fig. 4), and, to a lesserextent, by phism in the duskinessof the lores, 2. beccarii the UPGMA phenogram(Fig. 3). Moreover,the is aligned with the 2. frontalisgroup and 2. cit- synapomorphyat locusGda(a), which defined reogularis.There are, nevertheless, anomalies this clade, did not occur in any of the other consistentwith the protein phylogeny. 2. bec- specieswhich were all fixed for the ancestral cariihas the red iris and plain tail of the mag- state,indicating that convergenceat this locus nirostrissublineage; and 2. spiloderashares dusky is unlikely. breastspotting with some membersof the 2. Membersof the other primary lineage(com- frontalissuperspecies. The plain plumageof oth- prisingS. citreogularis,the frontalisgroup and S. er species,moreover, may mask relationships papuensis-S.spilodera) were not so close,as in- (Galbraith and Parker 1969). It is tempting to dicated in the UPGMA phenogram and dis- dismissthe morphologicalclusters as unsound, tance-Wagnertree (Figs. 3 and4). Althoughtheir basedon undiscriminatedplesiomorphies and cladeis alsodefined by a singlesynapomorphy, homoplasiesof unknown genetic background 6 Pgd(c),the allele was present in S. nouhuysi, and adaptive value. Becauseprotein phyloge- together with state (d). This may reflect con- niesmay be distortedby convergences(Stewart vergence.Convergence may explain much of et al. 1987), we examined additional evidence the observed variation at this locus in Sericornis from geographicaland ecologicaldistribution (Table 2), but it does not strengthen the hy- and genetic intergradation. pothesisof monophylyof the group. Distributionand foraging niche.--Patterns of al- lopatry and sympatry,associating taxa inversely DISCUSSION to their affinities, separate2. papuensisand 2. spiloderafrom the perspicillatusand magnirostris Concordanceof lineages.--The 3 differentanal- groups,align 2. beccariiwith 2. nouhuysi,and are ysesof electrophoreticdata define interspecific consistent with links between 2. keri and the alliancesthat are largely concordant.Sericornis frontalissublineages, and between2. beccariiand comprises5 clear-cut terminal clusters of species. the magnirostrissublineage. There is some discord, however, in the links In New Guinea, speciesthat coexist--up to 620 CHRISTIDIS,SCHODDE, •a•D BAVra•STOCK [Auk,Vol. 105

TABLE2. Frequenciesof electromorphsat the 32variable loci, proportion of polymorphiclocP (P), and mean heterozygosity(H) in all speciesexamined. Alphabetic designations of allelesrefer to their anodalmobilities, where A > B > C. Unless indicated otherwise,alleles are present at a frequency of 100%.

Locus (E.C. no.)

Pep L-A GPDH Pep L-PrO GOT-I Pep L-G-G 6PGD Species (3.4.11) (1.1.1.8) (3.4.11) (2.6.1.1) (3.4.11) (1.1.1.44)

Sericornisfrontalis 15(0.96) C(0.06)D(0.83) G(0.10) C A C(0.89)D(0.06) C(0.98)

S. humills B D C A C(0.92)D(0.08) A(0.17) B(0.08)C(0.75) S. keri A(0.14) B(0.86) D(0.50) G(0.50) C A B B(0.07)C(0.93) S. m. magnirostrisB D(0.71)H(0.14) C A(0.93)B(0.07) A(0.18)B(0.11) C(0.71) B S. m. viridior B D(0.20) G(0.60) H(0.10) A(0.40) C(0.60) A B(0.90)C(0.10) B I(0.10) S. beccarii A(0.13) 15(0.87) G C A 15(0.33)C(0.67) B S.c. citreogularisA(0.36) 15(0.64) G A(0.18)C(0.82) A A(0.32)15(0.14) C(0.50) C S.c. cairnsi B G A(0.17) C(0.83) A A(0.17) 15(0.25)C(0.58) C S. nouhuysi B D(0.85) G(0.15) C(0.96) A B(0.SI C(0.19) 15(0.46)C(0.54) S. perspicillatus B A(0.50)G(0.50) C(0.54)D(0.46) A 15(0.85)C(0.15) D S. papuensis B A(0.17)H(0.83) C A A(0.17)15(0.83) C S. spilodera B A(0.25)H(0.75) C C B C S. arfakianus B(0.88)C(0.12) A C(0.88)E(0.12) A C D Acanthizapusilia B E(0.08)H(0.17) I(0.75) C A A(0.08)B(0.84) C(0.08) B

three at any one site--are membersof different ric. Between S. papuensisand S. perspicillatus,as protein lineages and forage in separateforest between S. spiloderaand S. arfakianus,there was strata. Those that belong to the same or sister too much altitudinal overlap to be explainedby sublineageglean in the samestrata and replace competitive exclusionbetween closely related one anotheraltitudinally (Fig. 6, Table 5, cf. Fig. congeners.The first pair appeared to be parti- 5). Diamond (1969, 1972, 1985) aligned S. pa- tioned at least partly by feeding zone (Beehler puensisand S. spiloderawith the perspicillatus et al. 1986).In the second,S. spilodera,larger in group becausehe recordedthem in a successive size and with a longer and heavier bill than S. altitudinal sequence,but our data from six dis- arfakianus(Table 5), may alsoforage differently cretemountain systemsshow a different pattern (cf. Schoener 1965, Willson 1971). (Fig. 6). Only memberswithin the perspicillatus Membersof the third group in New Guinea, and papuensisgroups were consistentlyallopat- S.beccarii and S.nouhuysi, also replace one another

TABLE 2. Continued.

Locus (E.C. no.)

ENOL GA3PD PK GPT-2 CK-4 GPI MP1 EST-1 Species (4.2.1.11) (1.2.1.12) (2.7.1.40) (2.6.1.12) (2.7.3.2) (5.3.1.9) (5.3.1.8) (3.1.1.1)

S. frontalis B B C(0.96) C B(0.13)C(0.87) C(0.96) A A(0.08)B(0.92) S. humills B B C C C C A 15(0.67)C(0.25) D(0.08) S. keri B B C A B(0.29)C(0.71) C A(0.71) B(0.29) B(0.50)C(0.43) D(0.07) S. m. magnirostris B B C C A(0.11)C(0.89) C A(0.89)B(0.11) B(0.96)C(0.04) S. m. viridior B B C C A(0.30) C(0.70) C A(0.80) B(0.20) B(0.90)C(0.10) S. beccarii B B C C A(0.21)C(0.79) C A(0.79) B(0.13) B C(0.08) S.c. citreogularis C B C C A(0.09)C(0.91) C(0.86)D(0.14) A(0.96) A(0.77)B(0.23) S.c. cairnsi C B C C A(0.33) C(0.67) C A A(0.58) 15(0.42) S. nouhuysi B B C C A(0.15)C(0.85) B(0.19)C(0.81) A(0.96) B(0.77)C(0.23) S. perspicillatus B B C B(0.15)C(0.85) A(0.37) C(0.73) B A(0.96) B(0.09)D(0.50) E(0.41) S. papuensis A B A C C B A(0.83)15(0.17) C(0.50)D(0.50) S. spilodera B A A C C A A 15(0.50)D(0.50) S. arfakianus B B C C C B A C A. pusilla B B C C A(0.25) C(0.75) B(0.92)C(0.08) A(0.92) B(0.08) B(0.25)C(0.75)

A locusis consideredpolymorphic if the frequencyof the mostcommon allele doesnot exceed0.99. "Nothing" dehydrogenase,i.e. NADP specificbands, was observedwithout the addition of any substrateto the staining mixture. General protein stainedwith CoomasieBlue. October 1988] BiochemicalSystematics of Scrubwrens 621

TABLE 2. Extended.

Locus (E.C. no.)

G6PD GDA PGM-1 PGM-2 PGM-3 ACON-2 NP (1.1.1.49) NADP nDH 2 (3.5.4.3) (2.7.5.1) (2.7.5.1) (2.7.5.1) (4.2.1.3) (2.4.2.1)

B D(0.98) B A(0.96) B(0.06)C(0.86) C(0.92) C D(0.98) 0(0.06) A(0.17) B(0.83) D B A(0.92) C(0.08) C A A(0.25) B(0.75) D B A(0.43) B(0.43) C(0.14) B A C C C D B(0.79)C(0.07) D(0.14) D A A(0.79) C(0.21) A(0.64) C(0.36) C C D(0.96) B(0.40) C(0.60) D A A A(0.20) C(0.80) C C D

B D A A A(0.96) C C D(0.96) A(0.14) B(0.82) D B A(0.96) C C C(0.77) D(0.18) D(0.86) F(0.09) A(0.08) B(0.92) D B A(0.67) B(0.33) C C C(0.17) D(0.83) C B D A A A C C D B D(0.96) A A C C C D(0.92) E(0.08) B D B A C C C(0.83) D(0.17) D B D B A C C C D B D A A C C B(0.13) C(0.87) D A(0.83) B(0.17) D B A C C B(0.08)C(0.92) D(0.25) E(0.33)

altitudinally (see below) and overlap with patric with every other speciesin mainland membersof the other two protein lineagesat rain forests, all of which are members of the all levels. In Australia, S. beccariiis allopatric frontalisprotein lineage (Figs. ! and 2, cf. Figs. with both S. frontalisand S. magnirostris(Parker 4 and 5). Within the frontalisgroup allopatry is 1970, Keast 1978a, Boles 1979) where its mode more ecologicalthan geographical.S. frontalis of foraging (it creepsover stemsand lianas in is replacedby S. humillsin Tasmania,and over- the lower-mid strata of forest) is like that of lapsS. keriand S. citreogulariscompletely (Fig. magnirostris,not frontalis(Table 5). S. nouhuysi 2). Ecologicalseparation is greatestbetween keri also feeds in the same manner but lower, usu- and frontalis,and vergeson local allopatry.S. ally within 2 m of the ground (Beehler et al. keri is endemic to the substageof closed rain 1986). forest above ca. 650 m altitude in northeast Among Australiantaxa, S. magnirostrisis sym- Queenslandand frontalisoccupies the uncano-

TAnrE 2. Extended.

Locus (E.C. no.)

EST-2 EST-3 I DH- 1 ALD LDH- 1 LDH-2 (3.1.1.1) (3.1.1.1) (1.1.1.42) (4.1.2.13) (1.1.1.27) (1.1.1.27) GP-I• GP-23 GP-33 GP-4• P •I

B(0.92) B A(0.06) C(0.92) D A B B A B B 34.1 0.035 B B C D A(0.83) B(0.17) B B A B B 20.5 0.038 A(0.43) B(0.57) B C D A B B A B B 18.2 0.055 B B A(0.07) C(0.93) C A B B A B B 25.0 0.054 B B C C A B B A B B 18.2 0.068 B B C C A B B A B B 13.6 0.034

B B C E A B B A B A 25.0 0.033 B B C E A B B A B A 15.9 0.045 B B C C A B B A B B 18.2 0.035 B(0.96) A C(0.96) D(0.04) B A(0.92) B(0.08) B B A B B 27.3 0.049 B A C A A B A A A B 11.4 0.038 B B C -- A B B A B B 4.5 0.011 B B C -- A B B A B B 6.8 0.017 B C C F C A C B B B 22.7 0.042 622 CHRISTIDIS,SCHODDE, AND B^VEI•STOClC [Auk, Vol. 105

T^BLE3. Genetic distancemeasures. Upper matrix, Rogers(1972) genetic distance;lower matrix, Nei (1978) unbiasedgenetic distance.

Species I 2 3 4 5 6 7 8 9 I. Sericornisfrontalis -- 0.082 0.11! 0.112 0.147 0.134 0.150 0.178 0.116 2. S. humilis 0.046 -- 0.165 0.162 0.195 0.192 0.190 0.206 0.165 3. S. keri 0.075 0.129 -- 0.196 0.173 0.187 0.200 0.232 0.166 4. S. m. magnirostris 0.085 0.131 0.165 -- 0.076 0.053 0.204 0.230 0.064 5. S. m. viridior 0.114 0.164 0.141 0.032 -- 0.070 0.202 0.224 0.082 6. S. beccarii 0.116 0.173 0.170 0.022 0.033 -- 0.189 0.216 0.062 7. S.c. citreogularis 0.116 0.166 0.167 0.186 0.172 0.176 -- 0.071 0.202 8. S.c. cairnsi 0.152 0.189 0.208 0.222 0.209 0.212 0.036 -- 0.230 9. S. nouhuysi 0.091 0.143 0.137 0.023 0.041 0.031 0.183 0.220 -- 10. S. perspicillatus 0.182 0.234 0.202 0.164 0.135 0.163 0.228 0.271 0.144 11. S. papuensis 0.237 0.291 0.252 0.312 0.278 0.300 0.241 0.285 0.267 12. S. spilodera 0.166 0.224 0.183 0.232 0.201 0.225 0.212 0.257 0.197 13. S. arfakianus 0.168 0.203 0.179 0.148 0.127 0.162 0.221 0.268 0.122 14. Acanthizapusilia 0.323 0.359 0.355 0.322 0.297 0.352 0.391 0.421 0.333

pied shrubberiesof the forestedge and streams ferent localities at different altitudes) that there at the samelevels there. Becausefrontalis enters is a patchy sequenceof intergradationbetween rain forestswherever elsethey occurin its range beccariiand nouhuysiin . and foragesin the samezone as keri (Table 5), The more uniformly intermediate populations thispartitioning may be enforcedby competi- on isolated ranges away from the central cor- tion. S. citreogulariscoexists with frontalisand dillera may be stabilizedhybrids, proportionate keri in the same forests where the larger and in their traits to the genetic input from respec- heavier-billed citreogularisfeeds almost exclu- tive parental members (cf. Hartert 1930). sively on the forestfloor and may take prey of In Australia, S. beccariiintergrades with S. a different averagerange and size(Table 5; Keast magnirostris,not S. frontaliswhere it abuts on 1978a;cf. Hespenheide1973, 1975;Cody 1978). them on the eastern foot of Cape York Penin- Here the degreesof sympatryare consistent with sula (Schoddeand Ford in prep.; contraParker the greater genetic distancebetween citreogu- 1970, Boles 1979). Boles (1979) recorded beccarii larisand the other 2 species(Figs. 3 and 4). and magnirostrisas sympatric, but our seriesre- Geneticintergradation.--Intergradation through vealed a complexzone of introgressionbetween introgressivegene flow hasbeen recordedonly them in all differential traits, even color of bill among membersof the S. frontalissuperspecies (cf. Table 5). and between S. rnagnirostris,S. beccariiand S. Morphologicalcorrelations.--Patterns of geo- nouhuysi.Its sequencesin the frontalisgroup, graphicaldistribution, foraging niche, and ge- which implicatesS. hurnilis,were describedby neticintergradation are consistentwith the pro- Mayr (1937),Condon (1951,1954), Green (1969), tein phylogenywherever it wasdiscrepant with and Ford (1985), and analyzed electrophoreti- traditional morphologicalgroupings. Accord- cally by Christidisand Schodde(in press).It is ingly, we interpret thosesimilarities in plum- consistentwith both the proteinphylogeny and age marks that conflict with the protein phy- traditional morphological groupings. logeny as unresolved plesiomorphies and Intergradationamong membersof the mag- homoplasies.Discriminating them and their nirostrisprotein lineages conflicts with mor- polarity is not possible,except in terminal sub- phology.In New Guineawhere they abut (Fig. lineages,until the phylogeny of the broader 1), S. beccariiis linked to S. nouhuysiby a suite sericornineassemblage (Schodde 1975) is clar- of populationsidentified as "S. virgatus"(Rand ified because other members of the sericornine and Gilliard 1967: 359; Diamond 1969, 1985; Gil- assemblageshare both plain and analogously liard and Le Croy 1970;cf. Mayr 1986). Despite marked plumageswith Sericornis(sensu stricto). sightobservations of sympatry(Diamond 1985), For membersof terminal sublineages,character it is our impression(from spot-sampledpopu- statesin sisterlineages indicate that plain plum- lationsat differentstages of intermediacyat dif- age was acquiredsecondarily in S. kerias were October1988] BiochemicalSystematics of Scrubwrens 623

TABLE 3. Extended.

10 11 12 13 14

0.198 0.235 0.175 0.175 0.326 0.243 0.277 0.222 0.209 0.353 0.220 0.253 0.194 0.195 0.350 0.190 0.293 0.235 0.169 0.287 0.160 0.270 0.215 0.153 0.260 0.181 0.277 0.226 0.169 0.292 0.240 0.241 0.223 0.220 0.369 Genetic Distance (Nei 1975l 0.267 0.278 0.255 0.258 0.386 Fig. 3. UPGMA phenogrambased on Nei's (1978) 0.162 0.259 0.201 0.139 0.282 geneticdistance values (Table 3). -- 0.214 0.233 0.111 0.269 0.199 -- 0.206 0.229 0.326 0.231 0.210 -- 0.196 0.341 ginning in magnirostrisin central eastAustralia 0.079 0.232 0.196 -- 0.260 0.316 0.334 0.355 0.300 -- and ending in S. nouhuysiin New Guinea. By these criteria, S. rufescens,of which no tissues were available but which is as small and slen- der-billed as S. perspicillatusand occupiesthe patterned face and wing covertsand pale bill sameniche, is the vicar of that speciesand not in S. beccarii(Table 5, cf. Fig. 5). stubby-billed S. papuensisin the mountains of Variationsin bodily proportionscorrelate best Cendrawasi,west New Guinea (alsoMayr 1937, with niche occupiedand environment(Table 5, 1986; cf. Diamond 1985). cf. Schoener1965, Hespenheide 1973). In tarsus, Zoogeographicalsynthesis. wAncestral Sericornis arborealspecies were consistentlyshorter legged presumablyfirst split into two primary lineages than those such as the Australian members of in cool rain forestson the formerly unbroken the frontalisgroup which forage near and on Australo-southern New Guinean continent, the forest floor. Those scrubwrensthat probe probably at sometime within the last 5-10 mil- the cranniesof branchesand stems(rnagnirostris lion years(cf. Sibley and Ahlquist 1985). Then sublineage)or feed on the ground (Australian as the continent dried in the late Plio-Pleisto- frontalisgroup) have disproportionatelylonger and more slender beaks than higher level fo- SHU SKE liage- and twig-gleaners.Taxa at lower tropical latitudes have disproportionatelyshorter tails than thoseat higher temperatelatitudes, a char- /• SMA acteristicalso recored in the related genusAcan- thiza(Keast 1978c). In the frontalisgroup, S. keri SCC S• SMV hasa shortertail than southernS. frontalis, while in the magnirostrissublineage there is a south to north sequencein diminishing length, be-

TABLE4. Electromorphsdefining clades in Fig. 5.

Clade Apomorphic characters SPA o, 0,02o?, • APU I Pep L-G-G (D) 2 CK-4 (S), ALD (D) a Roger's (1972) D 3 GP-4 (A), ENOL (C) Fig. 4. Distance-Wagner tree based on Rogers' 4 PK (A) 5 6PGD (C) (1972) genetic distancevalues (Table 3). The tree is 6 G6PD (C) rooted by the out-group method of Farris (1972), to 7 PGM-2 (A), ALD (C) • Acanthizapusilla (APU). Other symbolsrefer to Seri- 8 6PGD (D) cornispapuensis (SPA), S. spilodera(SSP), S. citreogularis 9 GDA (B) citreogularis(SCI), S.c. cairnsi(SCC), S. keri (SKE), S. 10 GP-1 (B), GP-2 (B), LDH-1 (A), LDH-2 (B) humilis(SHU), S.frontalis (SFR), S. arfakianus(SAR), S. ' Charactersin which the plesiomorphicstate could not be deter- perspicillatus(SPE), S. nouhuysi(SNO), S. magnirostris mined. magnirostris(SMA), S.m. viridior(SMV), S.beccarii (SBE). 624 C,PaSTiVis,Sc•ovvE, AND BAVERS?OCK [Auk,Vol. 105

TAI3Lœ5. Morphologicaltraits, foraging strata, and distributionof the speciesof Sericornisin comparisonwith thoseof Acanthizapusilla. Data on foragingstrata from Diamond1972, Keast 1978a, Beehler et al. 1986(p. 31), and personalrecords. Symbols in parenthesesin Table indicateincomplete expression. The sample sizesof malesand females,respectively, for statisticalanalysis of relative size and proportionare in paren- thesesfollowing speciesnames. Parenthetic numbers after size are wing measurementsof male samples only; and thoseafter the following ratiosare for the sexescombined. Measurements are means + 1 SD. Wing measuredas the flattenedchord; tail, from tip to baseof centralrectrices; bill, as the exposedculmen; and tarsus,from heel to fore-edgeof last undivided scuteat baseof toes.

S. per- S. arfakianus S. nouhuysi S. beccarii S. m. magni- S. rufescens spicillatus olivaceus oorti minimus rostris Trait (7, 6) (10, 10) (10, 10) (10, 10) (10, 10) (10, 10) Russet face wash - - - + - (+) Russetfrons scalloping - - - (+) (Acanthiza-like) Black lores/frons .... (+) - Pale superciliary stripe .... + -- (broken) Pale subocular bar + - Pale malar line (+) - Blackspotting/streak- ing on throat/breast Diffuse olive-grey + (+) --,(+) (+) throat/breast streak- ing White tipping on alu- -,(+) + la/primary coverts Black subterminal tail (+) (+) (-) bar White tail tipping

Bill color dark dark dark dark pale dark Iris color chestnut chestnut chestnut chestnut chestnut chestnut Sexual dimorphism in face/throat marks Sexualdimorphism in (+) + + + + - size Size small small small large medium medium- 53.0 + 1.4 55.0 + 1.3 54.5 + 1.3 64.1 + 2.0 58.2 ñ 1.3 small 56.3 ñ 1.6 Tail: wing ratio short short short short short long 0.69 ñ 0.03 0.71 ñ 0.03 0.71 ñ 0.02 0.68 ñ 0.02 0.70 ñ 0.02 0.79 ñ 0.03

Bill: wing ratio medium medium medium long long long 0.18 _+ 0.01 0.18 ñ 0.01 0.18 ñ 0.01 0.19 ñ 0.01 0.20 ñ 0.01 0.21 ñ 0.01 Tarsus:wing ratio medium medium medium medium medium medium 0.35 ñ 0.02 0.36 + 0.01 0.36 ñ 0.02 0.37 ñ 0.02 0.36 + 0.02 0.37 ñ 0.01 Foraging strata arboreal arboreal arboreal substage subarbo- arboreal foliage stems real stems (2-20 m) (0.5-3 m) stems (1-20 m) (0.5-15 m) Distribution Cendra- New New New Cape York Central east wasi, Guinea Guinea Guinea Penin- Australia New sula Guinea

cene,both lineageswithdrew with the rain for- Strait to separatethe Australian and New Guin- eststo its eastand north fringes, taking refuge ean ranges,initiated vicariantsplits within both in the Australian Great Dividing Rangeand ris- lineages. In montane New Guinea the perspicil- ing New Guinean cordillera. These events, de- latus and papuensisgroups diverged from the veloping a dry Arafura plain and, later, Torres Australian magnirostrisand frontalis-citreogularis October 1988] BiochemicalSystematics of Scrubwrens 625

TABLE 5. Extended.

S. p. S. spilodera S. c. papuensis guttatus citreogularis S. keri S. h. humilis S.f. frontalis (10, 6) (10, 6) (10, 10) (6, 10) (10, 10) (10, 10) A. pusilla

_ _ + -- (+) + --

_ _ + -- (+) + -- - + - - + + scallop- streaked _ _ _ (+) + - _

.... (+) + -

+,(+) - _ (+) (+) (+) + (+ other races)

(+ other races) dark pale dark dark dark medium dark chestnut chestnut chestnut chestnut cream cream chestnut _ _ + (+) + + -

- + + + + + + medium medium large medium large medium 58.6 + 1.0 62.2 + 1.0 68.6 + 1.8 60.3 + 1.4 64.0 + 1.6 58.4 _+ 2.6 short medium long medium long long 0.68 + 0.02 0.74 + 0.02 0.79 + 0.2 0.73 + 0.03 0.82 + 0.03 0.80 + 0.04 short medium medium long long long 0.15 _+ 0.01 0.17 + 0.01 0.17 + 0.01 0.19 _+ 0.01 0.20 + 0.01 0.19 + 0.01 medium short long long long long 0.35 + 0.01 0.32 + 0.00 0.40 + 0.02 0.39 + 0.02 0.40 -+ 0.01 0.39 + 0.02 lower arbo- arboreal floor-sub- substage- substage- substage-floor aboreal rea! fo- stage floor floor (0-2 m) (0.5-10 m) liage (0-1 m) (0-i m) (0-2 m) (0.5-5 m) New New East Aus- North-east Tasmania South-east Australia Guinea Guinea tralia Australia Australia

groups,respectively. Since then, all 4 secondary 1969, Gilliard and Le Croy 1970). This recon- lineageshave undergoneminor radiationssep- structionis generally concordantwith the pa- aratelyin Australiaand New Guinea,with only leogeographicand environmentalhistory of the one, the magnirostrisgroup, dispersedbetween Australo-Papuancontinent of the times (Doutch the 2 land masses (see below, also Diamond 1972; Walker 1972; Bowler 1976, 1982; Dow 1977; 626 CnPaST•D•s,SCHODDE, ,•ND B,•VERSTOCZC [Auk,Vol. 105

US.arfakianus 2,00 1 Samplinglimit US.parspicillatus "C-•S.papuensis :LP•S.spilodar•S.nouhuysia-baccarit

24001 Samplinglirat__2 Fig. 5. Cladogram derived from electromorphs (Table 4), with Acanthizapusilla as the out-group.Bars representthe number of synapomorphiesthat define each clade, and numbers at branch points identify synapomorphicallozymes (Table 4). Letteredsymbols are identified in Fig. 4.

Sampling timit Kemp 1981; Schodde1982; Hope 1983; Pieters et al. 1983; Dow and Sukatmo 1984). An alter- native interpretation (Keast 1978a,b), postulat- 1500[• • ing that Sericornisarose in New Guinea and dis- 1200• m•linglimit persedto Australiain a complexseries of waves, 2100 3 4 is less parsimoniousand at variance with the Sampling limit protein phylogeny and known paleogeogra- Sampling limit phy. 1800 Within New Guinea, the sequence of radia- tion in the secondarylineages reflectsaltitu- dinal replacement(Schodde and Hitchcock1972; Diamond 1973, 1985). Ancestral stocks of the •t500 respectivepairs S. perspicillatus-S.arfakianus and S. papuensis-S.spilodera were probably first sep- 1200 Samplinglimit arated from each other on different eastern and Sampling timit westernor outlying sectorsof the central New Guinean cordillera, perhaps as altitudinal life o Local zonesrose above valley floorsduring the warm- 2too 5 mountainsummits 6 er interglacial epochsof the Pleistocene.Then, as the zones fell during the next or subsequent Samptlng limit glacial periods, the stocksspread out to come 1800 in contactand overlap one another at different altitudes where they had speciated. Sericornis perspicillatusmoved over S. arfakianusand, be- =15oo fore this, S. papuensisover S. spilodera,judged by

•12oo

Not sampled Fig. 6. Altitudinal rangesof speciesof Sericornisat 9OO Not sampled six localities in eastern New Guinea: 1. Kuper-Kaindi Ranges;2. Kukukuku Range; 3. ; 4. ; 5. south scarp, ; 6. Mt. Simpson-Mt. Dayman ranges. Narrow- 800 width bandsrepresent single records and broad-width bands 2 or more in the Australian National Wildlife Collection, CSIRO, and the Australian Bird Banding Scheme, Australian National Parks and Wildlife Ser- vice, Canberra. October1988] BiochemicalSystematics of Scrubwrens 627 their genetic,morphological and ecologicaldis- membersof the perspicillatusgroup more in form tances(Figs. 3 and 6, Table 5). Becausethe mem- and feeding behavior than either of its sister bersof both pairsare so widespreadtoday, we species(Table 5). The intergradation between cannot pinpoint their geographicorigins. Seri- membersof the magnirostrisgroup wherever they cornisrufescens has speciatedsince in isolation abut,together with slight differentiationin their from S. perspicillatusafter stocksof the latter proteins (Table 3), indicatesthat they have not reached the mountains of Cendrawasi when life speciatedcompletely. Their differencesin ex- zones were lower than now. ternal morphology, however, are as great as In Australia, radiation probably followed se- those between any speciesin Sericornis(Table quencesof latitudinal isolation produced by 5). There are 2 likely but unresolved explana- fluctuating connectionsin habitat around the tions: 1. In responseto very recent dispersal to periphery of the continent by climatic oscilla- New Guinea, perhaps during the last glacial tions during the Plio-Pleistocene(Gentilli 1949; epochswhen ancestralstocks of beccariimay have Serventy 1953, 1972; Keast 1959, 1961). First to been confined to tiny pocketsof rain forest on divergein thefrontalis primary lineagewere the the Arafura plain (Nix and Kalma 1972), they frontalis and citreogularisgroups, split in en- may have evolved more rapidly. This would claves of rain forest between northeast and accountfor the divergencein plumage pattern southeastAustralia by an early dry "glacial" but not the conformity in proteins. 2. Magni- epoch.We cannotidentify either region as the rostris,beccarii and nouhuysiare all older but their sourceof either group becauseboth groupsare intergradation in current pluvial times has en- widely sympatric today. Within the frontalis abledalleles for proteinsto spreadmore rapidly group, S. keri then budded off in the northeast than the genesfor morphologicaltraits. during a subsequentdry glacial period, fol- lowed by S. humilisin Tasmania.S. citreogularis ACKNOWLEDGMENTS was probablyalso split into northern and south- ern populationsat about the sametime or a little For permissionto collectmaterial for this study,we later, unless its membersdiverged at a slower are deeply indebted to the Papua New Guinea De- partment of Environment and Conservation,to the rate. Within both citreogularisand frontalis Australian State National Parks and Wildlife Service groups,the ratesof protein change(Fig. 4) was in Queenslandß New South Wales, South Australiaß fastestin the populationsisolated in northeast Tasmania and Western Australia, and the Conserva- Queensland (S. citreogulariscairnsi, S. keri) and tion Commissionof the Northern Territory. Individ- Tasmania (S. humills),probably reflecting past ual help was given by K. Kisokau, Lester Seri, K. bottlenecks in available habitat there. Because Aplin, Dr. and Mrs. G. Maynes, and S. Pruett-Jones its populationsnorth of Brisbaneare as yet un- (ringing data) in PapuaNew Guinea;and by C. Cot- differentiated (Ford 1985; Christidis and ben, F. Crome,R. H. GreenßM. KingßI. J. Mason, L. Schoddein press;pace Mayr 1937,1986), S. fron- MooreßG. O'Neill, L. Pedler, G. B. Ragless,D. Saun- talishas apparently re-expanded north to fringe ders, J. C. Wornbey,H. B. Gill and J. Pedler in Aus- tralia. Curators in Australian state museums and the on S. keri very recently. This could have hap- American Museum of Natural Historyß New York, pened 6,000-8,000years ago after the end of the generouslymade specimensin their care available. last glacial epoch when for the first and only The geographicalfigures were preparedby F. Knightß time in the last 25,000 years the northeast CSIRO Division of Wildlife and EcologyßCanberra. Queenslandtablelands were linked by suitable Christidiswas funded by a CSIRO postdoctoralfel- rain foreststo thosein the south (Nix and Kalma lowship for the duration of the study. 1972). We dedicatethis paper to ProfessorCharles G. Sib- The magnirostrisgroup apparently arose in ley on the occasionof his 70th birthday for his pi- Australia and dispersed to New Guinea, bud- oneeringexample in the molecularstudy of phylog- ding off S. beccariiand S. nouhuysiin the course eny in . of dispersal (also Diamond 1969, Gilliard and Le Croy 1970).It was accompaniedby a shift in LITERATURE CITED foraging zones,from the forest midstageto the AVISE,J. C., & C. F. AQUADRO.1982. A comparative substage,a reversal of the niche switch in the summary of genetic distancesin the vertebrates. frontalisprimary lineage between New Guinea Patterns and correlations. Evol. Biol. 15: 151-185. and Australia. Consistent with this, S. magni- ß J. C. PATTON,•r C. F. AQUADRO. 1980. Evo- rostris,the putative ancestral form, resembles lutionarygenetics of birds.J. Hered. 71: 303-310. 628 CHRISTIDIS,$CHODDE, AND BAVERSTOCK [Auk,Vol. 105

BEEH•ER,B. M., T. K. PRATT,& D. A. ZIMMERMAN. 1986. of the White-browed Scrub- in eastern Aus- Birds of New Guinea. Princetonß Princeton Univ. tralia. Emu 85: 92-96. Press. GALBRAITH,I. C. J., & $. A. PARKER. 1969. The Ath- BOLES,W.E. 1979. Comments on Sericornisbeccarii in erton Scrub-wren Sericornis keri Mathews: a ne- southern Cape York Peninsula.Sunbird 10: 70- glectedAustralian species.Emu 69: 212-232. 72. GENtII•I,I, J. 1949. Foundations of Australian bird BOWLEg,J.M. 1976. Aridity in Australia:age, origins geography.Emu 49: 85-129. and expression in aeolian landforms and sedi- GILLIAP,D, E. T., & M. LE CRO¾. 1970. Notes on birds ments. Earth-Sci. Rev. 12: 279-310. from the , New Guinea. Am. 1982. Aridity in the late Tertiary and Qua- Mus. Novit. 2420: 1-28. ternaryof Australia.Pp. 35-46 in Evolutionof the GREEN,R.H. 1969. The birds of Flinders Island. Rec. flora and fauna of arid Australia (W. R. Barker Queen Victoria Mus. Launceston 34: 1-32. and P. J. M. Greenslade, Eds.). Adelaide, Peacock HARTERT,E. 1930. List of the birdscollected by Ernst Publ. Mayr. Novit. Zool. 36: 27-128. CHRISTIDIS,L. 1987. Biochemicalsystematics within HENNIG,W. 1966. Phylogeneticsystematics. Chica- paleotropicfinches (Aves: Estrildidae).Auk 104: go, Univ. Illinois Press. 380-392. HESPENHFaDœ,H.A. 1973. Ecologicalinferencesfrom --, & R. SCHODDE.Genetic and morphological morphologicaldata. Ann. Rev. Ecol.Syst. 4: 213- differentiation within the White-browed Scrub- 229. wren complex(Sericornis frontalis: Aves). Aust. J. --. 1975. Preycharacteristics and predatorniche Zool. In press. width. Pp. 158-180 in Ecologyand Evolution of COD¾,M.L. 1978. Habitat selectionand interspecific Communities(M. L. Cody and J. M. Diamond, interactionsamong sylviid warblersin England Eds.).Cambridge, Massachusetts, Belknap Press. and Sweden. Ecol. Monogr 48: 351-396. HOPE,G. $. 1983. Australianenvironmental changes: CONDON, H. T. 1951. Notes on the birds of South timing, directions,magnitudes and rates.In Pleis- Australia: occurrence, distribution and taxono- tocene extinctions: the search for causes (P.S. my. $. Aust. Ornithol. 20: 26-68. Martin and R. G. Klein, Eds.). Tucson, Univ. Ar- --. 1954. Bird maps, 1. S. Aust. Ornithol. 21: 41- izona Press. 44. KEAST,A. 1959. Australian birds: their zoogeogra- CORBIN,K.W. 1977. Genetic diversity in avian pop- phy and adaptationsto an arid continent.Pp. 89- ulations. Pp. 291-302 in Endangeredbirds (S. A. 114in Biogeographyand ecologyin Australia(A. Temple, Ed.). Madison, Wisconsin, Univ. Wis- Keast, R. L. Crocker, and C. $. Christian, Eds.). consin. The Hague, Junk. DIAMOND,J.M. 1969. Preliminary resultsof an or- --. 1961. Bird speciationon the Australian con- nithologicalexploration of the north coastalrange, tinent. Bull. Mus. Comp. Zool., Harvard 123:303- New Guinea. Am. Mus. Novit. 2362: 1-57. 495. --. 1972. Avifauna of the EasternHighlands of --. 1978a. Zoogeographyand evolutionary his- New Guinea. Publ. Nuttall Ornithol. Club 12: 1- tory of the scrubwren genus Sericornis.Emu 78: 438. 119-125. 1973. Distribution ecology of New Guinea 1978b. The relationships of the Fieldwren birds. Science 179: 759-769. Calamanthusfuliginosus (Acanthizinae). Emu 78: 1985. New distributional records and taxa 20-24. from the outlying mountain ranges of New --. 1978c. The status of the Guinea. Emu 85: 65-91. Acanthiza katherina. Emu 78: 7-10. DOUTCH,H.F. 1972. The paleogeographyof north- KEMP,E.M. 1981. Tertiary palaeogeographyand the ern Australia and New Guinea and its relevance evolution of Australian climate. Pp. 33-49 in Eco- to the Torres Strait area. Pp. 1-10 in Bridge and logicalbiogeography of Australia(A. Keast,Ed.). barrier:the naturaland culturalhistory of Torres The Hague, Junk. Strait (D. Walker, Ed.). Canberra, Res.Sch, Pacific MAYR,E. 1937. Notes on the genusSericornis Gould. Stud. Publ. BG/3, Australian National Univ. Am. Mus. Novit. 904: 1-25. Dow, D. B. 1977. A geologicalsynthesis of Papua 1986. Acanthizidae.Pp. 409-464 in Check- New Guinea. Canberra, Australian Bur. Miner. of the world: a continuation of the Resour.Geol. Geophys.Publ. 201. work of JamesL. Peters,vol. XI (E. Mayr and G. ß& R. SUKATMO.1984. Later Tertiary to Qua- W. Cottrell,Eds.). Cambridge, Massachusetts, Mus. ternary tectonicsof Irian Jaya.Episodes 7: 3-9. Comp. Zool. FARMS,J.S. 1972. Estimatingphylogenetic trees from -, & A. L. RAND. 1937. Results of the Archbold distance matrices. Am. Nat. 106: 645-668. expeditions. 14. Birds of the 1933-1934 Papuan FORD,J. 1985. Secondarycontact between subspecies expedition. Bull. Am. Mus. Nat. Hist. 73: 1-248. October1988] BiochemicalSystematics ofScrubwrens 629

, & D. L. SERVENTY. 1938ß A review of the 1982ß Origin, adaptationand evolutionof genusAcanthiza Vigors & Horsfield.Emu 38:245- birds in arid AustraliaßPp. 191-224 in Evolution 292. of the flora and fauna of arid Australia (W. R. MEISE,W. 1931ß Zur systematikder gattung Geryg- Barker and P. J. M. Greenslade, Eds.). Adelaide, one. Novit. zool. 31: 317-379ß Peacock Press. NEI, M. 1975. Molecular population genetics and ß& W. B. HITCHCOCK.1972ß BirdsßPp. 67-86 evolution. Amsterdam, North-Holland. in Encyclopaediaof Papuaand New Guinea,vol. 1978ß Estimationof averageheterozygosity 1 (P. Ryan, Ed.). MelbourneßMelbourne Univ. and geneticdistance from a small number of in- Pressß dividuals. Genetics 89: 583-590. ß& J. L. McKF,•N. 1976ß The relations of some Nix, H. A., & J. D. KALM^. 1972ß Climate as a dom- monotypicgenera of Australianoscines. Pp. 530- inant control in the biogeographyof northern 541 in Proc. 16th Int. Ornithol. Congr. (H. J. Frith Australia and New Guinea. Pp. 60-91 in Bridge andJ. H. Calaby,Eds.). Canberra, Australian Acad. and barrier: the natural and cultural history of Sci. Torres Strait (D. Walkerß Ed.). Canberra, Res. Sch. SCHOr•ER, T. W. 1965ß The evolution of bill size Pacific Stud. Publ. BG/3, Australian National differencesamong sympatric congeneric species Univß of birds. Evolution 19: 180-213. PARKER,S.A. 1970ßTaxonomy of the populationsof SERVENTY,D.L. 1953ßSome speciation problems in Sericornisbeccariiinhabiting Cape York Peninsulaß Australian birds:with particular referenceto re- Emu 70: 69-72. lations between Bassianand Eyrean "species- P^TTON,J. C., & J. C. AVISE. 1983. An empirical eval- pairs." Emu 53: 131-145. uation of qualitativeHennigian analysesof pro- --. 1972ßCausal ornithogeography of Australia. tein electrophoretic dataßJ. Molecular Evol. 19: Pp. 574-584 in Proc. 15th Int. Ornithol. Congr. 244-254. Leiden, E. J. Brill. PIETERS,P. E., C. J. PIGRAM, D. S. TRAIL, D. B. DOW, N. SIBLEY,C. G., & J. E. AHLQI3IST.1985. The phylogeny RATM^N,& R. SUKATMO.1983. The stratigraphy and classificationof the Australo-Papuanpasser- of western Irian Jaya.Bull. Geol. Resourc.De- ine birds. Emu 85: 1-14ß velopment Ctr. 8: 14-48ß SNEATH, P. H. A., & R. R. SOKAL. 1973. Numerical RAND, A. L., & E. T. GILLIARD. 1967ß Handbook of taxonomyßSan Francisco,W. H. Freemanß New Guinea birdsß London, Weidenfeld & Nic- STEWART,C.-B., J. W. SCHILLING,& A. C. WILSON. 1987. olson. Adaptiveevolution in the stomachlysozymes of RAOU CHECKLIST CoMMrr'r•,•. 1926. The official foregutfermenters. Nature 330:401-404. checklist of the birds of Australia, 2nd ed. Mel- W^LKr•,D. 1972. Bridgeand barrier. Pp. 399-405in bourne,Royal Australasian Ornithologists Union. Bridgeand barrier:the naturaland culturalhis- ROCERS,J. S. 1972ß Measuresof genetic similarity tory of Torres Strait (D. Walker, Ed.). Canberra, and geneticdistanceß Studies in genetics.7. Univ. Res. Sch. Pacific Stud. Publ. BG/3, Australian Na- Texas Publ. 7213: 145-153. tional Univ. SCHODDE,R. 1975. Interim list of Australian song- WILLSON, M.F. 1971. Seed selection in some North birds: passetines.Melbourne, RAOU. American finches. Condor 73: 415-429. ß 1981ß The taxonomyof the RAOU "Interim List of AustralianSongbirds." Aust. Birds 15:63- 71.