Sericornis, Acanthizidae)
Total Page:16
File Type:pdf, Size:1020Kb
GENETIC AND MORPHOLOGICAL DIFFERENTIATION AND PHYLOGENY IN THE AUSTRALO-PAPUAN SCRUBWRENS (SERICORNIS, ACANTHIZIDAE) LESLIE CHRISTIDIS,1'2 RICHARD $CHODDE,l AND PETER R. BAVERSTOCK 3 •Divisionof Wildlifeand Ecology, CSIRO, P.O. Box84, Lyneham,Australian Capital Territory 2605, Australia, 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 species: 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 New Guinea. 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 (Acanthiza) and gerygones Although this reorganization was controver- (Gerygone),scrubwrens of the genus 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 Spotted Scrubwren (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 differentiation from 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 Bird 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 those of 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, Papua New Guinea 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