phylogeographic patterns, molecular and vocal

www.elsevier.com/locate/ympev

Molecular Phylogenetics and Evolution 44 (2007) 154–164

Phylogeographic patterns, molecular and vocal differentiation,and species limits in Schiffornis turdina (Aves)

Arpad S. Nyari *

Natural History Museum and Biodiversity Research Center, University of Kansas, 1345 Jayhawk Blvd., Dyche Hall, Lawrence, KS 66045-2454, USA

Received 6 July 2006; revised 25 January 2007; accepted 17 February 2007Available online 27 February 2007

Abstract

Establishing species limits can be challenging for organisms in which few variable morphological characters are available, such asSchiffornis turdina, a Neotropical suboscine bird of long-debated taxonomic affinities. Apart from its dull plumage and secretive behav-ior, this taxon is well-known for its subtle but discrete within-species geographic variation in vocalizations. Phylogeographic reconstruc-tion based on three mitochondrial markers sampled across much of the species’ range reveals substantial structuring, concordant withrecognized areas of endemism in Neotropical lowland forests. Monophyly of S. turdina was weakly supported by the combined dataset,as was the basal position of the Guyanan Shield population with regard to other S. turdina clades. Based on the results from both geneticand a preliminary, qualitative analysis of vocalizations, I recommend revised species limits to reflect more accurately the evolutionaryhistory of this complex.� 2007 Elsevier Inc. All rights reserved.

Keywords: Schiffornis; Aves; Phylogeography; Neotropics; Andes; Vocalizations; Species limits

1. Introduction

Phylogeographic patterns among lowland South andCentral American birds have recently been explored withthe aid of molecular markers, revealing previously unap-preciated degrees of geographic differentiation (Aleixo,2002, 2004; Burns and Naoki, 2004; Cheviron et al.,2005; Eberhard and Bermingham, 2005; Lovette, 2004;Marks et al., 2002). Such detailed studies of Neotropicaltaxa have also greatly improved the picture of overallavian diversity in the region, especially for taxa for whichfew diagnosable morphological characters have made spe-cies-level taxonomy problematic. Such is the case ofSchiffornis, an enigmatic and difficult genus including(at present) 3 species that have long challenged taxono-mists (Ames, 1971; McKitrick, 1985; Prum and Lanyon,1989; Sibley and Ahlquist, 1985; Sibley and Monroe,1990; Prum et al., 2000; Irestedt et al., 2001; Johansson

1055-7903/$ - see front matter � 2007 Elsevier Inc. All rights reserved.

doi:10.1016/j.ympev.2007.02.020

* Fax: +1 785 864 5335.E-mail address: [email protected]

et al., 2002; Chesser, 2004), resulting in recent recognitionof a novel higher taxon, the Tityrinae (including Schiffor-

nis and several other problematic suboscines). Althoughhigher-level relationships of the Tityrinae have receivedconsiderable attention (Chesser, 2004; Ericson et al.,2006), the details of variation, distribution, and specieslimits of the component taxa remain poorly known(Prum and Lanyon, 1989). Here, I focus on Schiffornis

turdina, a medium-sized, sexually monochromatic (Eaton,2005) dull-colored, secretive bird distributed throughoutNeotropical humid lowland forests from southeasternMexico south to northern Bolivia and the Atlantic Forestof southeastern Brazil (Fig. 1). S. turdina, throughout itsbroad geographic distribution exhibits subtle but discretevariation in plumage coloration and vocalizations (Stilesand Skutch, 1989; Ridgely and Tudor, 1994; Snow,2004).

Presently, 13 subspecies are recognized within S. turdina

(Peters, 1979; Snow, 2004); although the existence of multi-ple species has been suggested (Ridgely and Tudor, 1994;Howell and Webb, 1995; Hilty, 2003; Snow, 2004), no

mailto:[email protected]

Fig. 1. Distribution map for Schiffornis turdina (adapted from Ridgely and Tudor, 1994; Snow, 2004) showing samples included in the genetic analysis asdotted circles. Outlines represent individual phylogroups, numbered corresponding to clades recovered through phylogenetic analyses (see Fig. 2). Insettree provides a cross-reference to the overall phylogeographic relationships within S. turdina.

�A.S. Nyari / Molecular Phylogenetics and Evolution 44 (2007) 154–164 155

rangewide treatment has as yet addressed any feature of itsphenotype or genotype. Descriptions of geographic varia-tion in this species have focused on definitions of subspe-cies in terms of subtle differences in plumage hue andintensity, and body size. Plumage types range from darker,brownish-olive forms (S. t. veraepacis, S. t. acrolophites,

S. t. rosenbergi, S. t. aenea) to brighter, rufescent forms(S. t. panamensis, S. t. stenorhyncha). Birds of S. t. olivacea,S. t. amazona, and nominate S. t. turdina are mostlyuniform olive-brown overall (Ridgely and Tudor, 1994;Snow, 2004).

Males are polygynous, advertising their presence onwidely spaced territories through syncopated, melodious,whistled songs, usually given at long intervals. Geographicvariability of these songs is striking, ranging from two-noted whistles (S. t. veraepacis and S. t. olivacea), three-noted, slower and more drawn-out songs (S. t. amazona)to four-noted, faster songs (S. t. panamenis and S. t. aenea;Ridgely and Tudor, 1994) (Fig. 3). Upon attracting afemale by means of vocalizations, no physical courtshipdisplays are performed (Skutch, 1969; Prum and Lanyon,1989; Snow, 2004), which distinguishes Schiffornis frommanakins and cotingas, where mate choice centers aroundritualized visual displays of bright, sexually dimorphicplumage (Prum and Johnson, 1987; Prum, 1990; Robbins,1983).

Since vocalizations play a pronounced role in the lifecycle of this species, plumage characters may provide fewuseful characters in understanding its variation. The utilityof molecular markers for inferring phylogeographic

patterns has been established firmly (Avise, 2000), and con-stitutes a powerful approach in cases such as the presentone, in which few diagnosable morphological charactersare available for inferring historical relationships or estab-lishing species limits. A modern reconstruction of historicalpatterns of geographic population structure in S. turdina

will add an important contribution to the pool of taxa thatcan serve as a basis for future integrative and comparativephylogeographic analyses (Avise, 2000; Zink et al., 2001).As such, the purpose of this study is to use molecular char-acters to explore (1) monophyly of S. turdina, (2) aspects ofphylogeographic variation across its range, and (3) geo-graphic structure and patterns of vocal differentiation forcomparison with patterns of genetic differentiation.

2. Materials and methods

2.1. Taxon sampling and laboratory protocols

Schiffornis was represented in this study by 41 individu-als; of these, 38 covered most of the range of S. turdina,with samples from nearly every recognized subspecies(Table 1 and Fig. 1). The nominate subspecies of the Atlan-tic Forest of southeastern Brazil, for which no fresh/frozentissue was available, was sampled via a toepad from amuseum study skin (FMNH 191688); laboratory work onthis sample was conducted by the Genetics Lab, Depart-ment of Systematic Biology, National Museum of NaturalHistory and National Zoological Park, following estab-lished in-house protocols (Fleischer et al., 2000, 2001).

Table 1Taxa included in this study, with sample sources (all museum voucher specimens), and GenBank sequence accession numbers

Taxon Subspecies Sourcea Sample Locality ND2 COI cyt b

IngroupSchiffornis turdina veraepacis KUNHM 2115 Mexico, Silvituc EF458501 EF458586 EF458543Schiffornis turdina veraepacis MZFC 14587 Mexico, Chiapas EF458499 EF458584 EF458541Schiffornis turdina veraepacis MZFC 14589 Mexico, Chiapas EF458498 EF458583 EF458540Schiffornis turdina veraepacis MZFC 10543 Mexico, Quintana Roo EF458500 EF458585 EF458542Schiffornis turdina veraepacis LSUMNS 16118 Costa Rica, Puntarenas EF458504 EF458589 EF458546Schiffornis turdina dumicola LSUMNS 9885 Panama, Chiriquı EF458502 EF458587 EF458544Schiffornis turdina dumicola LSUMNS 9887 Panama, Cocle EF458503 EF458588 EF458545Schiffornis turdina panamensis LSUMNS 9882 Panama, Canal Zone EF458524 EF458608 EF458567Schiffornis turdina panamensis LSUMNS 9883 Panama, Canal Zone EF458525 EF458609 EF458568Schiffornis turdina panamensis LSUMNS 1352 Panama, Darien EF458522 EF458607 EF458565Schiffornis turdina panamensis LSUMNH 2261 Panama, Darien EF458523 – EF458566Schiffornis turdina rosenbergi ANSP 2230 Ecuador, Esmeraldas EF458505 EF458590 EF458547Schiffornis turdina rosenbergi LSUMNS 11820 Ecuador, Esmeraldas EF458506 EF458591 EF458548Schiffornis turdina rosenbergi ANSP 3531 Ecuador, Azuay EF458507 EF458592 EF458549Schiffornis turdina olivacea AMNH ROP 164 Venezuela, Bolivar EF458495 EF458580 EF458537Schiffornis turdina olivacea KUNHM 1265 Guyana, Iwokrama Reserve EF458489 EF458574 EF458531Schiffornis turdina olivacea KUNHM 3937 Guyana, N slope of Mt. Roraima EF458491 EF458576 EF458533Schiffornis turdina olivacea KUNHM 5793 Guyana, Barima River EF458490 EF458575 EF458532Schiffornis turdina amazona LSUMNS 7550 Venezuela, Amazonas EF458511 EF458596 EF458553Schiffornis turdina amazona LSUMNS 20362 Brazil, Amazonas EF458494 EF458579 EF458536Schiffornis turdina amazona LSUMNS 36666 Brazil, Rondonia EF458521 EF458606 EF458564Schiffornis turdina amazona ANSP 5792 Ecuador, Sucumbios EF458513 EF458598 EF458555Schiffornis turdina amazona LSUMNS 2552 Peru, Loreto EF458512 EF458597 EF458554Schiffornis turdina amazona KUNHM 889 Peru, Loreto EF458514 EF458599 EF458556Schiffornis turdina aenea ANSP 4450 Ecuador, Zamora-Chinchipe EF458508 EF458593 EF458550Schiffornis turdina aenea ANSP 5100 Ecuador, Sucumbios EF458510 EF458595 EF458552Schiffornis turdina aenea LSUMNS 5543 Peru, San Martin EF458509 EF458594 EF458551Schiffornis turdina wallacii USNM 10460 Guyana, Sipu River EF458493 EF458578 EF458535Schiffornis turdina wallacii USNM 10470 Guyana, Sipu River EF458492 EF458577 EF458534Schiffornis turdina wallacii FMNH 391536 Brazil, Amapa EF458497 EF458582 EF458539Schiffornis turdina wallacii FMNH 391537 Brazil, Amapa EF458496 EF458581 EF458538Schiffornis turdina wallacii FMNH 391539 Brazil, Para (S bank) EF458517 EF458602 EF458559Schiffornis turdina wallacii FMNH 391540 Brazil, Para (S bank) EF458519 EF458604 EF458561Schiffornis turdina steinbachii LSUMNS 1984 Peru, Pasco EF458515 EF458600 EF458557Schiffornis turdina steinbachii LSUMNS 9545 Bolivia, Pando EF458516 EF458601 EF458558Schiffornis turdina steinbachii FMNH 322499 Peru, Cuzco EF458518 EF458603 EF458560Schiffornis turdina steinbachii LSUMNS 13831 Bolivia, Santa Cruz EF458520 EF458605 EF458563Schiffornis turdina turdina FMNH 191688 Brazil, Minas Gerais – – EF458562Schiffornis virescens KUNHM 307 Paraguay EF458526 EF458610 EF458569Schiffornis virescens KUNHM 3830 Paraguay, Caazapa EF458527 EF458611 EF458570Schiffornis major KUNHM 1426 Peru, Madre de Dios EF458528 EF458612 EF458571

OutgroupLaniocera hypopyrra KUNHM 1408 Peru, Madre de Dios EF458529 EF458613 EF458572Laniisoma elegans ANSP 1558 Ecuador, Morona-Santiago EF458530 EF458614 EF458573

a Museum abbreviations: LSUMNS - Louisiana State University Museum of Natural Science, Baton Rouge; ANSP - Academy of Natural Sciences,Philadelphia; FMNH - Field Museum of Natural History; AMNH - American Museum of Natural History; MZFC - Museo de Zoologıa, Facultad deCiencias, Universidad Nacional Autonoma de Mexico; USNM – US National Museum of Natural History; KUNHM - The University of Kansas NaturalHistory Museum.

156 �A.S. Nyari / Molecular Phylogenetics and Evolution 44 (2007) 154–164

S. virescens (2 samples) and S. major were included to rep-resent the remaining congener sister species, and Laniocera

hypopyrra and Laniisoma elegans were included as out-group taxa, based on extensive recent higher-level studies(Barker et al., 2004; Chesser, 2004).

Total genomic DNA was extracted from frozen or alco-hol-preserved tissue samples using standard Qiagen tissueextraction protocols (Qiagen, Valencia, CA). Sequencesof the mitochondrial genes NADH dehydrogenase subunit2 (ND2), cytochrome b (cyt b), and cytochrome c oxidasesubunit I (COI) were used as molecular markers. ND2

was amplified using the primers L5216 and H6313 (Soren-son et al., 1999) and a standard PCR thermal cycling pro-tocol with initial denaturation at 94 �C, annealing at 55 �C,and extension at 72 �C. Protocols and primers for cyt b andCOI amplification followed Johansson et al. (2002) andHebert et al. (2004), respectively. The targeted fragmentof the COI gene corresponds to the one proposed to serveas the unique DNA marker for the Consortium for theBarcode of Life (CBOL) and the All Birds Barcoding Ini-tiative (ABBI) (Stoeckle, 2003; Hebert et al., 2004). AllPCR amplifications were carried out in 25 ll reactions


using Amersham PureTaq RTG PCR beads (AmershamBiosciences). Amplified double-stranded PCR productswere cleaned with Agencourt AmPure PCR purificationsystem (Agencourt Bioscience Corp.), and visualized on ahigh-melt agarose gel stained with ethidium bromide.

Purified PCR products were cycle-sequenced with ABIPrism BigDye v3.1 terminator chemistry using the sameprimers as for each PCR reaction. Cycle sequenced prod-ucts were further purified using CleanSEQ (Agencourt Bio-science Corp.) magnetic beads and finally sequenced on anABI 3130 automated sequencer. Sequences of both strandsof each gene were examined and aligned in Sequencher 4.1(GeneCodes Corp., 2000), and a final data matrix of con-tiguous sequences assembled using ClustalX 1.8 (Thomp-son et al., 1997). Datasets for each gene were alignedwith homologous sequences from the chicken (Gallus gallus

domesticus) genome from GenBank (Desjardins and Mor-ais, 1990) to verify base composition, amino acid transla-tion, and assure mitochondrial origin of amplifiedproducts.

2.2. Phylogenetic analyses

Evaluation of the molecular dataset and phylogeneticreconstruction were performed using the programs PAUP*

4.0 (Swofford, 2002) and MrBayes 3.1 (Huelsenbeck andRonquist, 2001; Ronquist and Huelsenbeck, 2003). Out-group taxa included for all the analyses were Laniocera

hypopyrra and Laniisoma elegans (Barker et al., 2004;Chesser, 2004). Prior to analyzing the combined dataset,a partition homogeneity test (Farris et al., 1994) was car-ried out in PAUP* (with 1000 heuristic replicates), to detectany incongruences among the phylogenetic signals of thethree genes; this step would determine if subsequent analy-ses could be conducted on the overall combined dataset.

Phylogenetic analyses included both parsimony (MP)and likelihood (ML) approaches. MP reconstructions werecarried out in PAUP* through heuristic searches, with 1000random stepwise addition replicates and TBR branchswapping. The dataset was further explored through threecharacter weighting schemes: (1) equal weighting of all sub-stitutions, (2) downweighting of third position transitionsby a factor of 5 relative to transversions, and (3) down-weighting third codon positions by a factor of 5 relativeto other positions.

Bayesian ML (BML) analyses were conducted usingMrBayes 3.1, with a default flat prior probability densityand flat distribution of nucleotide substitutions and basefrequencies. The Markov chain Monte Carlo searchparameters included a general time reversible model(nst = 6) with the molecular dataset partitioned by geneand codon position, and was run for 5 · 106 generationswith default chain heating conditions, sampling every 100generations. The topologies sampled from the first 25%of generations were discarded as an initial ‘‘burn-in,’’ afterhaving examined the analysis for stationarity by plotting�lnL against generation time.

ML analyses of the complete dataset were performed inPAUP*, using a heuristic search with 100 addition stepwiseaddition sequence replicates. Parameter estimation wasestablished through a best fit model of evolution recoveredvia a hierarchical likelihood ratio test (hLRT) and Akaikeinformation criterion (AIC), in ModelTest 3.7 (Posada andCrandall, 1998). Because the model that the AIC identifiedas best fitting the GTR + I + G model, whereas hLRTidentified the TrN + G model, both models of evolutionwere explored. However, only the results fromGTR + I + G are presented because the two models pro-duced identical topologies. For both MP and ML analyses,nodal support was assessed via bootstrapping with 1000and 100 random addition replicates, respectively. Differ-ences in rate heterogeneity across lineages were assessedby comparing likelihood scores for the ML topology withand without a molecular clock enforced, where twice thedifference in log likelihood value was compared to a v2 dis-tribution with n � 2 degrees of freedom (n = number oftaxa).

2.3. Vocalizations

Suboscines have long been regarded as the only passe-rines that have innate songs, lacking the learning capacityand associated forebrain nuclei for song acquisition char-acteristic of oscine passerines (Kroodsma and Konishi,1991), although at least one exception to this rule is known(Baptista and Kroodsma, 2001). S. turdina advertisementsong recordings were obtained from the Macaulay Libraryof Natural Sounds (Cornell University, Ithaca, NY; Fig. 3)and from a commercially available recording (Krabbe andNilsson, 2004). Representative songs covering the species’range were included to provide independent estimates ofintraspecific variation, and to contrast vocal divergencewith genetic differences. Although several vocal sampleswere available for most described subspecies, lack of multi-ple song samples per individual and population samplesprecluded detailed, quantitative analyses of song charactervariation. Therefore, after reviewing all available record-ings, a characteristic loudsong (Zimmer and Isler, 2003)was selected for each geographic region as representativefor the population or subspecies, for a preliminary qualita-tive analysis.

Recordings were visualized as sound spectrograms pro-duced with the aid of the bioacoustics software RAVEN1.2 (Cornell Lab of Ornithology, Charif et al., 2004). Rep-resentative spectrograms from each geographic region wereexamined for characters such as number of notes, note fre-quency range, and duration. Based on these characteristics,geographic song variation was assessed, and results weresubsequently contrasted with those obtained from themolecular characters. To illustrate geographic song varia-tion in comparison with molecular findings, each spectro-gram was exported from RAVEN as an image file, andthen converted to a vectorial image in Adobe Illustrator(Adobe Systems Inc.) by retracing the dominant spectrum


of each song, while maintaining accurate frequency andtemporal ranges for descriptive interpretation (Fig. 3).

3. Results

3.1. Sequence dataset characteristics

The aligned three-gene dataset resulted in a total of 2475base pairs, of which 1020 were of ND2, 615 of COI and 840of cyt b. For the single ancient DNA sample, 462 base pairsof cyt b were included in the analysis. Of the 2475 basepositions, 839 (33.9%) were variable and 629 (25.4%) werephylogenetically informative. For the ingroup, among thethree genes analyzed, COI had the highest percentage ofinformative sites (21.9%), but mostly at third codon posi-tions (60.0%; Table 2); no second codon position baseswere phylogenetically informative in the ingroup in COI(Table 2).

Sequence alignment was straightforward and compari-sons to the respective published chicken sequences revealedsimilar base compositions, with no insertions, deletions, oranomalous stop-codon placements that would have ren-dered protein-coding regions non-functional. Thus, I con-cluded that the genes were of true mitochondrial originand did not represent nuclear pseudogenes (Aleixo, 2002;Cheviron et al., 2005; Marks et al., 2002). Nucleotide fre-quencies showed a bias towards low levels of guanine,but were well overall within the frequency ranges of previ-ously published mitochondrial datasets.

The average pairwise distance (uncorrected P) betweeningroup and outgroup taxa was 17.1%. S. major differedon average by 14.6% from other Schiffornis members.Within the remainder of the ingroup, sequence divergences

Table 2Summary of sequence attributes for ND2, COI, and cyt b sequences of 38 Sc

Gene Total sites Variable sites Informative sites Percent of sit

1st

ND2 1020 289 (28.3%) 213 (20.8%) 17.3 (10.6)COI 615 156 (25.4%) 135 (21.9%) 7.8 (5.8)Cyt b 840 216 (25.7%) 171 (20.3%) 11.8 (8.2)

Table 3Percent pairwise uncorrected P distances among phylogroups and outgroups,

Phylogroups 1 2 3 4

1 —2 8.4 —3 0.8 8.2 —4 3.1 8.6 3.2 —5 8.5 5.4 8.4 8.6 8.1 5.1 7.9 8.7 9.6 9.4 9.5 9.S. virescens 9.2 9.2 9.1 9.S. major 14.4 14.9 14.9 14.

To outgroup 17.2 17.7 17.3 17.Distances within phylogroups 0.2 0.3 0.3 0.

among major lineages ranged from 0.8% (between theMexican and western Ecuadorian clades) to 9.6% (betweenthe Mexican and Guyanan Shield clade; Table 3). Withinmajor lineages (see Section 4), mean sequence divergenceranged from 0.1% (Guyanan Shield clade) to 1.7% (withinthe SE Amazon and Atlantic forest clades).

3.2. Phylogenetic analysis

A partition homogeneity test revealed no significant dif-ferences in phylogenetic signal (P = 0.60) between the threegenes, so all analyses were performed on the combineddataset. Examination of saturation plots for each gene(not shown) with numbers of transitions plotted againstoverall uncorrected P distances showed evidence of satura-tion at third codon positions beyond 10% sequencedivergence.

Maximum parsimony analysis identified 12 most parsi-monious trees (length = 1396, CI = 0.562, RI = 0.873;excluding uninformative sites). Bayesian and ML analysesrecovered the same topology as the MP consensus tree(likelihood scores of �lnL = 10511.45 and �lnL =10513.40, respectively). Schiffornis as a whole was recov-ered as monophyletic with S. major as basal to S. virescens

and several geographically distinctive S. turdina phylo-groups (Figs. 1 and 2). All three phylogenetic methods indi-cated high levels of phylogeographic structure within S.

turdina, with high bootstrap/posterior probabilities formost clade support (all P90%), except for the CentralAmerican clade (clade 1, Fig. 2), which had only 65%ML bootstrap support. In contrast, deeper nodes werecharacterized by weak support (e.g. nodes A and B inFig. 2). Different weighting schemes explored under the

hiffornis turdina samples

es variable by codon (informative) Nucleotide frequencies

2nd 3rd %A %C %G %T

9.7 (5.0) 57.9 (47.0) 30.5 32.4 9.1 280.0 (0.0) 68.3 (60.0) 24.8 30.6 14.9 29.73.2 (2.8) 62.1 (50.0) 25.6 31.3 13.7 29.4

and percent within-phylogroup variation

5 6 7 S. virescens S.major

5 —1 4.3 —2 9.5 9.1 —2 9.3 8.9 9.2 —4 14.3 14.4 15.0 14.7 —

3 17.0 16.8 17.0 17.13 1.2 1.7 0.1 — —

Fig. 2. Phylogeographic relationships within Schiffornis turdina based on Bayesian likelihood analysis. Numbers above branches represent posteriorprobabilities, while ML bootstrap scores are indicated below; *indicates <50% bootstrap support. S. turdina samples are indicated by their collectinglocalities, while nodes A–D and the seven phylogroups are referenced throughout the text. The tree is rooted with Laniisoma elegans and Laniocera

hypopyrra (not shown).


MP criterion all recovered the same topology. Nodal sup-port for basal relationships improved slightly upon down-weighting third position transitions: node A supporting amonophyletic S. turdina received MP bootstrap supportof 80%, and node B rendering the Guyanan Shield popula-tions as basal relative to other clades improved to 90% ascompared to the equally-weighted analysis.

MP and BML analyses based solely on COI failed torecover monophyly of S. turdina, placing S. virescens

within S. turdina, as sister to the Guyanan Shield popula-tions; differentiation and reciprocal monophyly of clades1 and 3 were also not recovered. Analyses of the combineddataset, however, managed to recover monophyly of S. tur-

dina, although with relatively poor support (<50% MLbootstrap support; Fig. 2). Similar low support was recov-ered for the placement of the Guyanan Shield populationas basal to other S. turdina clades (66% ML bootstrap sup-port; Fig. 2). The Guyanan Shield populations are sister to

Fig. 3. Preliminary summary of vocal variation throughout the distribution of Schiffornis turdina, with dots indicating recording localities; letters areassociated with each spectrogram and locality. Numbered outlines summarize the geographic extent of song similarity. Inset tree provides a reference ofsong groups with regard to molecular phylogroups. Recording sources: MLNS (Macaulay Library of Natural Sounds, Cornell University, Ithaca, NY):65777 (a), 11604 (b), 42847 (c), 89455 (d), 57396 (e), 63187 (f), 79747 (g), 13221 (h), 81019 (j), 66100 (k), 108452 (l), 43356 (m), 51903 (n), 115380 (o), 66103(p), 60229 (q); Krabbe and Nilsson (2004): LXXXIB 1–15 (i).


a clade comprised of two groups of trans- and cis-Andeanorigin (nodes C and D; Fig. 2), respectively. One clade(node C) unites populations from Mexico south to westernPanama, as well as western Ecuadorian lowland popula-tions. Within this group, the basal populations belong tothe cis-Andean foothill populations of eastern Ecuadorand central Peru. The second major group (node D)includes populations from the Amazon headwaters in thewest as sister to samples from the southeastern Amazonand the Atlantic forest. Also part of this assemblage areindividuals from Panama east of the Panama Canal (Figs.1 and 2).

Finally, the likelihood ratio test was unable to detect sig-nificant rate heterogeneity across lineages (v2 = 50.09,P > 0.15), suggesting a relatively uniform manner of molec-ular evolution in this lineage. However, for lack of appro-priate calibrations for splitting events within Schiffornis, nodating was attempted.

3.3. Vocalizations

Examination of spectrograms from recordings through-out a significant part of the range of S. turdina indicatesnotable between-population song variation, with consistentgeographic patterns. Geographic patterns of song typescontain well-defined entities, distinguishable in note struc-ture, frequency range, and temporal characters. Accordingto overall song characteristics, 5 main groups are diagnos-able (Fig. 3). Referring to the phylogeographic structurerecovered from the molecular dataset, these song types

are distributed as follows: (1) a 2-note song with a some-what longer (1–1.5 s), upslurred first note and a second,brief (0.2 s) ‘‘tu’’-like note in molecular clades 1, 3 and 7;(2) a 4-noted, rapid succession of song elements (the firstending slightly down-slurred and the second rising in fre-quency), ending with two short ‘‘tu’’ notes in clade 4; (3)a 3-note whistled song, of one longer note followed bytwo shorter elements ending upslurred, in clades 5 and 6;(4) the distinctive song of the nominate subspecies (songo, Fig. 3), which compared to other Amazonian-type songshas a distinctive, narrow frequency range in all three notes;and (5) a rapid, upslurred 3-note whistled song of broadfrequency range (2–4.6 kHz) in clade 2.

4. Discussion

4.1. Phylogeographic patterns

Interpretations of historical biogeographic events acrossthe Neotropics were reviewed by Haffer (1997), but withoutthe benefit of detailed phylogenetic information. However,recent molecular phylogenetic studies have provided newperspectives on speciation in the region (Aleixo, 2002,2004; Bates et al., 1999; Burns and Naoki, 2004; Chevironet al., 2005; Eberhard and Bermingham, 2005; Lovette,2004; Marks et al., 2002; Perez-Eman, 2005). Phylogeneticanalysis of my combined dataset resolved a monophyleticS. turdina complex, although support for the basal nodethat excludes S. virescens was weak (node A, Fig. 1). Basedon the three mitochondrial genes analyzed, S. turdina con-


sists of several well-differentiated sets of populations (Figs.1 and 2). Here, I delineate individual phylogroups based ontheir degree of molecular differentiation and geographicallydisjunct distributional areas.

Phylogroups identified on molecular grounds includepopulations: (1) from Southern Mexico south to westernPanama (clade 1; S. t. veraepacis and S. t. dumicola), (2)eastern Panama (clade 2; S. t. panamensis, and S. t. steno-

rhyncha), and (3) in the western lowlands of Ecuador (clade3; S. t. rosenbergi and S. t. acrolophites). (4) Sister to theseare cis-Andean populations from the eastern slopes of theAndes in Ecuador and Peru, (clade 4; S. t. aenea). TheAmazon Basin includes members of populations distrib-uted in (5) the headwaters of Amazon tributaries of thelowlands of Southern Venezuela to eastern Ecuador, cen-tral Peru, and northwestern Bolivia (clade 5; S. t. amazo-

na); (6) areas from Cuzco, Peru, through Bolivia andPara, Brazil (clade 6; S. t. wallacii, S. t. steinbachii, S. t. tur-

dina), and including the single sample available from theAtlantic Forest of Brazil (S. t. intermedia from easternmostpart of the Atlantic Forest of Brazil remained unsampled);and (7) the Guyanan Shield and lowlands north of thelower Amazon River (S. t. olivacea and S. t. wallacii)(Peters, 1979; Ridgely and Tudor, 1994; Snow, 2004).

The differentiated lineages identified here within S. turdi-

na coincide well with known areas of endemism of Neo-tropical lowland forests (Cracraft, 1985; Cracraft andPrum, 1988). For instance, the lineages identified withinthe Amazon Basin (clades 5 and 6) and the Guyanan Shieldclade (clade 7) are separated by substantial genetic dis-tances (4.3–9.5%), supporting the idea that the lower Ama-zon River and its tributaries represent an important barrierfor terra firme forest birds (Aleixo, 2004). Moreover, eventhrough the rather sparse sampling in the Amazon Basin,we can observe even finer-scale phylogeographic structur-ing, as seen within groups 5 and 6 (Fig. 2), where individ-uals from Peru and Bolivia, and Brazil and Bolivia,respectively, are on average 2.5% different from membersof their respective groups. Clades in Mesoamerica andEcuador (clades 2 and 3) are interrupted geographicallyby taxa of Amazonian affinity (clade 7).

The lowland forests of northern Colombia are thoughtto have facilitated faunal exchange around the northerntip of the Andes during humid Pleistocene and post-Pleis-tocene climates (Haffer, 1967a,b). Since no samples wereavailable from the lowlands of either Colombia or theCaribbean coastal lowlands of Venezuela, these hypothesescannot yet be tested using genetic evidence. However,vocalizations of populations from Falcon, Venezuela, andDarien, Panama, show conserved note structure and asso-ciated temporal and frequency characteristics (songs p andq, Fig. 3), which suggests a single clade from eastern Pan-ama through the northeastern lowlands of Colombia tonorthwestern Venezuela.

The rise of the Andes played a major role in early sepa-ration of the Choco-Central American and Amazonianfaunas, and several studies have concluded that the Andean

uplift occurred after vicariant events that isolated theAtlantic Forest from the main forest mass of the Amazon(Cracraft and Prum, 1988; Eberhard and Bermingham,2005; Cheviron et al., 2005; Marks et al., 2002). However,in this study, phylogeographic affinities of cis-Andean foot-hill populations (clade 4) remain with trans-Andean popu-lations and not with adjacent lowland Amazonianpopulations (Brumfield and Capparella, 1996; Haffer,1967a). What is more, the Amazon–Atlantic Forest split(to the extent that it is a split) was clearly late, considerablylater than splits associated with the Andean chain. Consid-ering that S. virescens is also distributed in the AtlanticForest of southeastern Brazil, where it replaces S. t. turdina

altitudinally (Ridgely and Tudor, 1994; Snow, 2004), thesituation would make the Atlantic Forest more of a biogeo-graphic hybrid area (Cracraft and Prum, 1988; Markset al., 2002; Costa, 2003).

Comparable to other phylogeographic studies involvinglowland Neotropical populations, this study identified well-differentiated lineages throughout the species’ range, even iflack of resolution at basal nodes precludes conclusionsabout sequences of vicariant events. Similar patterns oflow phylogenetic support for relationships among areasof endemism have been found in other studies with densertaxon sampling (Aleixo, 2004; Marks et al., 2002; Perez-Eman, 2005) and larger molecular datasets (Lovette, 2004).

4.2. Species limits and vocalizations

Vocalizations play an important role in ornithology,providing key characters for inferring species limits, partic-ularly in suboscine birds (Isler et al., 1998, 1999, 2005;McCracken and Sheldon, 1997). Although this study aimedto provide independent genetic and vocal data sets for arevised view of phylogeographic divisions within the cur-rent S. turdina, insufficient sampling precluded more thor-ough analysis of song character variation. Majorgeographic divisions in song similarity were consistent withthose in the genetic analysis, although not as finelyresolved. In contrast to the molecular dataset, my analysesof vocalizations supported only five subgroups, comparedto the seven groups recovered from genetic characters.Only slight vocal differentiation can be observed betweenwhat are genetically highly divergent clades: e.g. the Guya-nan Shield populations vs. populations of Central Americaand the Choco (clades 7 vs. 1; 7 vs. 3; Table 3) and amongthe two largely Amazonian clades (clade 5 vs. 6; Table 3).

Three species concepts are seeing some degree of use inornithology: the Biological Species Concept (BSC; Mayr,1963), the Phylogenetic Species Concept (PSC; Cracraft,1983), and the Evolutionary Species Concept (ESC; Wiley,1980), each with its own suite of conceptual and opera-tional criteria. In sorting through the phylogeographic pat-terns and distinct forms outlined above, one challenge isthat of establishing a ‘best’ treatment of each form or setof forms under each concept. The PSC and ESC bothemphasize monophyly of species taxa, which in the present


example is fulfilled for each of the phylogroups discussedabove. The PSC, which emphasizes diagnosability of pop-ulations, would likely recognize 6 or 7 current S. turdina

populations as full species: all of the groups listed above,except that clades 1 (Mesoamerica) and 3 (western Ecua-dor) are not distinct vocally and not markedly distinct inmolecular characters, although they are reciprocally mono-phyletic. An ESC view would probably best also recognize6 or 7 species, on the grounds that the phylogroups are allreciprocally monophyletic, and appear to be evolving alongindependent evolutionary trajectories.

BSC species limits, however, are more complex to estab-lish. In the case of S. turdina, forms that are distributedparapatrically and that are vocally (Fig. 3) and genetically(Fig. 2) distinct indisputably merit species status: S. vera-

epacis of Mexico, Central America, western Ecuador, andthe Guyanan Shield (including clades 1, 3 and 7, above);S. stenorhyncha of eastern Panama and the lowlands ofnorthern Colombia and northwestern Venezuela (clade2); S. aenea of the lowlands of western Ecuador and Peru(clade 4); S. amazona of the western Amazon Basin (clade5), and S. turdina in the southeastern Amazon Basin andthe Atlantic Forest (Fig. 1). Each of these forms is geneti-cally distinct in spite of occurring in close proximity toanother form, and hence likely reproductively isolatedand recognizable under the BSC. Arguments could bemade for further subdivision based on odd distributionalpatterns and genetic distinctiveness, such as between clades1 and 3, but these arguments (e.g. range disjunctions) donot fall clearly into BSC thinking, which focuses on repro-ductive isolation. Since proper establishment of specieslimits should always rely on multiple, independent charac-ter-based diagnostics (DeSalle et al., 2005), I hereby proposethe above as a taxonomic treatment for the assemblage.

This study also constitutes one of the few rangewide sur-veys of a Neotropical bird species to include the COI genefragment (Eberhard and Bermingham, 2005; Lovette,2004). In this study, COI on its own would not have per-mitted recovering of monophyly of S. turdina and detectionof the same level of phylogroup detail. Using a criterionbased on a 10 times average intraspecific variation for flag-ging genetically divergent taxa as distinct species (Hebertet al., 2004) would not be easily applicable, as the overallaverage intra-S. turdina differentiation is 6.7%.

5. Conclusions

The present study offers first insights into the phylogeo-graphic structure of an enigmatic Neotropical bird, S. tur-

dina. Consistent patterns of geographic structure weredemonstrated, including fairly dramatic genetic differentia-tion of regional populations. Although vocal characters didnot support the same levels of geographic subdivision asthe molecular dataset, they nevertheless were useful in con-firming five subgroups as meaningful in phenotypic terms.Further genetic sampling is needed in zones of potentialintergradation in eastern Panama and the Colombian

Choco, and along the major tributaries of the AmazonRiver to establish the status of those populations. Thiswork would ideally be conducted via targeted playbackexperiments and collecting of the same individual, creatinghighest-quality voucher specimens for species-limitsclarification.

Acknowledgments

This research was supported by the University of Kan-sas General Research Fund, and by generous assistancefrom Richard Prum, whose excitement for suboscinesguided me towards Schiffornis. Tissue samples were pro-vided by the Louisiana State University Museum of Natu-ral Science; Academy of Natural Sciences, Philadelphia;Field Museum of Natural History; American Museum ofNatural History; Museo de Zoologıa, Facultad de Cien-cias, Universidad Nacional Autonoma de Mexico; USNational Museum of Natural History; and University ofKansas Natural History Museum. I am thankful to theField Museum of Natural History for allowing subsam-pling of a S. t. turdina study skin. I am also indebted toall the field collectors for their efforts towards accumulat-ing information-rich, vouchered specimens, without whichthis research would not have been possible. For their gen-erous assistance with recordings, I thank the MacaulayLibrary of Natural Sounds, Cornell University, as well asthe numerous contributors to sound libraries. Robert Flei-scher and Dana Hawley were extremely generous with labwork on the ancient DNA sample. Special thanks to Town-send Peterson and Mark Robbins for their guidancethroughout. For laboratory assistance and troubleshootingI thank Michael Grose, Jennifer Pramuk, Shannon DeVa-ney, Brett Benz, and Elisa Bonaccorso, while Keith Barkerprovided valuable help with phylogenetic programs. The fi-nal version of this manuscript benefited from helpful sug-gestions by two anonymous reviewers, and by AndresCuervo. Finally, I thank Monica Papes for continuousencouragement and support.

References

Aleixo, A., 2002. Molecular systematics and the role of the ‘‘varzea’’ –‘‘terra-firme’’ ecotone in the diversification of Xiphorhynchus wood-creepers (Aves: Dendrocolaptidae). Auk 119, 621–640.

Aleixo, A., 2004. Historical diversification of a terra-firme forest birdsuperspecies: a phylogeographic perspective on the role ofdifferent hypotheses of Amazonian diversification. Evolution 58,1303–1317.

Ames, P.L., 1971. The morphology of the syrinx in passerine birds. Bull.Peabody Mus. Nat. Hist. 37, 1–194.

Avise, J.C., 2000. Phylogeography: The History and Formation of Species.Harvard University Press, Cambridge.

Baptista, L.F., Kroodsma, D.E., 2001. Avian bioacoustics. In: del Hoyo,J., Elliot, A., Sargatal, J. (Eds.), Handbook of the Birds of the World,Mousebirds and Hornbills, vol. 6. Lynx Edicions, Barcelona, Spain,pp. 11–52.

Barker, F.K., Cibois, A., Schikler, P., Feinstein, J., Cracraft, J., 2004.Phylogeny and diversification of the largest avian radiation. Proc. Nat.Acad. Sci. 101, 11040–11045.


Bates, J.M., Hackett, S.J., Goerck, J.M., 1999. High levels of mitochon-drial DNA differentiation in two lineages of antbirds (Drymophyla andHypocnemis). Auk 116, 1039–1106.

Brumfield, R.T., Capparella, A.P., 1996. Historical diversification of birdsin northwestern South America: a molecular perspective on the role ofvicariant events. Evolution 50, 1607–1624.

Burns, K.J., Naoki, K., 2004. Molecular phylogenetics and biogeographyof the Neotropical tanagers in the genus Tangara. Mol. Phylogenet.Evol. 32, 838–854.

Charif, R.A., Clark, C.W., Fristrup, K.M., 2004. Raven 1.2. CornellLaboratory of Ornithology, Ithaca, NY.

Chesser, R.T., 2004. Molecular systematics of the New World suboscinebirds. Mol. Phylogenet. Evol. 32, 11–24.

Cheviron, Z.A., Hackett, S.J., Capparella, A., 2005. Complex evolution-ary history of a Neotropical lowland forest bird (Lepidothrix coronata)and its implications for historical hypotheses of the origin ofNeotropical avian diversity. Mol. Phylogenet. Evol. 36, 338–357.

Cracraft, J., 1983. Species concepts and speciation analysis. In: Johnston,R.F. (Ed.), Current Ornithology, vol. 1. Plenum Press, New York, pp.159–187.

Cracraft, J., 1985. Historical biogeography and patterns of diversificationwithin the South American areas of endemism. In: Buckley, P.A.,Foster, M.S., Morton, E.S., Ridgely, R.S., Buckley, F.G. (Eds.),Neotropical Ornithology. Ornithological Monographs No 36. Am.Ornithol. Union, Washington, DC, pp. 49–84.

Cracraft, J., Prum, R.O., 1988. Patterns and processes of diversification:speciation and historical congruence in some Neotropical birds.Evolution 42, 603–620.

Costa, L.P., 2003. The historical bridge between the Amazon and theAtlantic Forest of Brazil: a study of molecular phylogeography withsmall mammals. J. Biogeogr. 30, 71–86.

DeSalle, R., Egan, M.G., Siddall, M., 2005. The unholy trinity: taxonomy,species delimitation and DNA barcoding. Phil. Trans. R. Soc. B. 360,1905–1916.

Desjardins, P., Morais, R., 1990. Sequence and gene organization of thechicken mitochondrial genome. A novel gene order in highervertebrates. J. Mol. Biol. 212, 599–634.

Eaton, M.D., 2005. Human vision fails to distinguish widespread sexualdichromatism among sexually ‘‘monochromatic’’ birds. Proc. Nat.Acad. Sci. 102, 10942–10946.

Eberhard, J.R., Bermingham, E., 2005. Phylogeny and comparativebiogeography of Pionopsitta parrots and Pteroglossus toucans. Mol.Phylogenet. Evol. 36, 288–304.

Ericson, P.G.P., Zuccon, D., Ohlson, J.I., Johansson, U.S., Alvarenga, H.,Prum, R.O., 2006. Higher-level phylgeny and morphological evolutionof tyrant flycatchers, cotingas, manakins, and their allies (Aves:Tyrannida). Mol. Phylogenet. Evol. 40, 471–483.

Farris, J.S., Kallersjo, M., Kluge, A.G., Bult, C., 1994. Testing signifi-cance of incongruence. Cladisics 10, 315–319.

Fleischer, R.C., Olson, S., James, H.F., Cooper, A.C., 2000. The identityof the extinct Hawaiian eagle (Haliaeetus) as determined by mito-chondrial DNA sequence. Auk 117, 1051–1056.

Fleischer, R.C., Tarr, C.L., James, H.F., Slikas, B., McIntosh, C.E., 2001.Phylogenetic placement of the po‘o-uli (Melamprosops phaeosoma)based on mitochondrial DNA sequence and osteological characters.Stud. Avian Biol. 22, 98–103.

GeneCodes Corp., 2000. Sequencer, Version 4.1, GeneCodes Corp., Inc.,Ann Arbor, MI.

Haffer, J., 1997. Alternative models of vertebrate speciation in Amazonia.Biodivers. Conserv. 6, 451–476.

Haffer, J., 1967a. Speciation in Colombian forest birds west of the Andes.Am. Mus. Novit., 2294.

Haffer, J., 1967b. Some allopatric species pairs of birds in northwesternColombia. Auk 84, 343–365.

Hebert, P.D.N., Stoeckle, M.Y., Zemlak, T.S., Francis, C.M., 2004.Identification of birds through DNA barcodes. PLoS Biol. 2,1657–1663.

Hilty, S.L., 2003. Birds of Venezuela. Princeton University Press, NJ.

Howell, S.N.G., Webb, S., 1995. A Guide to the Birds of Mexico andNorthern Central America. Oxford University Press, NY.

Huelsenbeck, J.P., Ronquist, F., 2001. MRBAYES: bayesian inference ofphylogenetic trees. Bioinformatics 17, 754–755.

Irestedt, M., Johansson, U.S., Parson, T.J., Ericson, P.G.P., 2001.Phylogeny of major lineages of suboscines (Passeriformes) analysedby nuclear DNA sequence data. J. Avian Biol. 32, 15–25.

Isler, M.L., Isler, P.R., Whitney, B.M., 1998. Use of vocalizations toestablish species limits in antbirds (Passeriformes: Thamnophilidae).Auk 115, 577–590.

Isler, M.L., Isler, P.R., Whitney, B.M., 1999. Species limits in antbirds(Passeriformes: Thamnophilidae): the Myrmotherula surinamensis com-plex. Auk 116, 83–96.

Isler, M.L., Isler, P.R., Brumfield, R.T., 2005. Clinal variation invocalizations of an antbird (Thamnophilidae) and implications fordefining species limits. Auk 122, 433–444.

Johansson, U.S., Irestedt, M., Parson, T.J., Ericson, P.G.P., 2002.Basal phylogeny of the Tyrannoidea based on comparisons ofcytochrome b and exons of nuclear c-myc and RAG-1 genes. Auk119, 984–995.

Krabbe, N., Nilsson, J., 2004. Birds of Ecuador. Bird Songs InternationalBV, Netherlands.

Kroodsma, D.E., Konishi, M., 1991. A suboscine bird (Eastern Phoebe,Sayornis phoebe) develops normal song without auditory feedback.Anim. Behav. 42, 477–487.

Lovette, I.J., 2004. Molecular phylogeny and plumage signal evolution ina trans Andean and circum Amazonian avian species complex. Mol.Phylogenet. Evol. 32, 512–523.

Marks, B.D., Hackett, S.J., Capparella, A.P., 2002. Historical relation-ships among Neotropical lowland forest areas of endemism asdetermined by mitochondrial DNA sequence variation within theWedge-billed Woodcreeper (Aves: Dendrocolaptidae: Glyphorhynchus

spirurus). Mol. Phylogenet. Evol. 24, 153–167.Mayr, E., 1963. Animal Species and Evolution. Belknap Press of Harvard

University Press, Cambridge, Massachusetts.McCracken, K.G., Sheldon, F.H., 1997. Avian vocalizations and phylo-

genetic signal. Proc. Nat. Acad. Sci. USA 94, 3833–3836.McKitrick, M.C., 1985. Monophyly of the tyrannidae (Aves): comparison

of morphology and DNA. Syst. Zool. 34, 35–45.Perez-Eman, J.L., 2005. Molecular Phylogenetics and biogeography of the

Neotropical redstarts (Myioborus; Aves, Parulinae). Mol. Phylogenet.Evol. 37, 511–528.

Peters, J.L., 1979. In: Check-list of Birds of the WorldMuseum ofComparative Zoology, Vol. 8. Cambridge, Massachusetts.

Posada, D., Crandall, K.A., 1998. ModelTest: testing the model of DNAsubstitution. Bioinformatics 14, 817–818.

Prum, R.O., Johnson, A.E., 1987. Display behavior, foraging ecology, andsystematics of the Golden-winged Manakin (Masius chrysopterus).Willson Bull. 99, 521–539.

Prum, R.O., Lanyon, W.E., 1989. Monophyly and phylogeny of theSchiffornis group (Tyrannoidea). Condor 91, 444–461.

Prum, R.O., 1990. Phylogenetic analysis of the evolution of displaybehavior in the Neotropical Manakins (Aves: Pipridae). Ethology 84,202–231.

Prum, R.O., Rice, N.H., Mobley, J.A., Dimmick, W.W., 2000. Apreliminary phylogenetic hypothesis for the cotingas (Cotingidae)based on mitochondrial DNA. Auk 117, 236–241.

Ridgely, R.S., Tudor, G., 1994. In: The Birds of South AmericaTheSuboscine Passerines, vol. 2. University of Texas Press, Austin, TX.

Robbins, M.B., 1983. The display repertoire of the Band-tailed Manakin(Pipra fasciicauda). Wilson Bull. 95, 321–342.

Ronquist, F., Huelsenbeck, J.P., 2003. MRBAYES 3: bayesian phyloge-netic inference under mixed models. Bioinformatics 19, 1572–1574.

Sibley, C.G., Ahlquist, J.E., 1985. Phylogeny and classification of NewWorld suboscine passerines (Passeriformes: Oligomyodi: Tyrannides).In: Buckley, P.A., Foster, M.S., Morton, E.S., Ridgely, R.S., Buckley,F.G. (Eds.), Neotropical Ornithology. American Ornithologists’Union, Washington, DC, pp. 396–430.


Sibley, C.G., Monroe, B.L., 1990. Distribution and Taxonomy of theBirds of the World. Yale University Press, New Haven, CT.

Skutch, A.F., 1969. Life histories of Central American birds III. PacificCoast Avifauna 35, 1–580.

Snow, D.W., 2004. Family pipridae (manakins). In: del Hoyo, J., Elliot,A., Christie, D.A. (Eds.), Handbook of the Birds of the World,Cotingas to Pipits and Wagtails, vol. 9. Lynx Edicions, Barcelona, pp.110–169.

Sorenson, M.D., Ast, J.C., Dimcheff, D.E., Yuri, T., Mindell, D.P., 1999.Primers for a PCR-based approach to mitochondrial genome sequenc-ing in birds and other vertebrates. Mol. Phylogenet. Evol. 12, 105–114.

Stiles, G.F., Skutch, A.F., 1989. A Guide to the Birds of Costa Rica.Cornell University Press, Ithaca, New York.

Stoeckle, M., 2003. Taxonomy, DNA and the barcode of life. BioScience53, 2–3.

Swofford, D.S., 2002. Phylogenetic Analysis Using Parsimony (*and othermethods), Version 4. Sinauer, Sunderland, MA.

Thompson, J.D., Gibson, T.J., Plewniak, F., Jeanmougin, F., Higgins,D.G., 1997. The ClustalX Windows interface: flexible strategies formultiple sequence alignment aided by quality analysis tools. NucleicAcids Res. 24, 4876–4882.

Wiley, E.O., 1980. The evolutionary species concept reconsidered. Syst.Zool. 27, 17–26.

Zimmer, K.J., Isler, M.L., 2003. Family thamnophilidae (typicalantbirds). In: del Hoyo, J., Elliott, A., Christie, D. (Eds.),Handbook of Birds of the World, vol. 8. Lynx Editions, Barcelona,pp. 448–681.

Zink, R.M., Kessen, A.E., Line, T.V., Blackwell-Rago, R.C., 2001.Comparative phylogeography of some aridland species. Condor 103,1–10.

phylogeographic patterns, molecular and vocal

Documents