phylogenetic relationships and historical biogeography of tribes … · (lepidoptera: nymphalidae)...
TRANSCRIPT
Biological Journal of the Linnean Society
, 2005,
86
, 227–251. With 5 figures
© 2005 The Linnean Society of London,
Biological Journal of the Linnean Society,
2005,
86
, 227–251
227
Blackwell Science, LtdOxford, UKBIJBiological Journal of the Linnean Society0024-4066The Linnean Society of London, 2005? 2005862227251Original Article
PHYLOGENY OF NYMPHALINAEN. WAHLBERG
ET AL
.
*Corresponding author. E-mail: [email protected]
Phylogenetic relationships and historical biogeography of tribes and genera in the subfamily Nymphalinae (Lepidoptera: Nymphalidae)
NIKLAS WAHLBERG
1
*, ANDREW V. Z. BROWER
2
and SÖREN NYLIN
1
1
Department of Zoology, Stockholm University, S-106 91 Stockholm, Sweden
2
Department of Zoology, Oregon State University, Corvallis, Oregon 97331–2907, USA
Received 10 January 2004; accepted for publication 12 November 2004
We infer for the first time the phylogenetic relationships of genera and tribes in the ecologically and evolutionarilywell-studied subfamily Nymphalinae using DNA sequence data from three genes: 1450 bp of cytochrome oxidasesubunit I (COI) (in the mitochondrial genome), 1077 bp of elongation factor 1-alpha (EF1-
a
) and 400–403 bp of
wing-less
(both in the nuclear genome). We explore the influence of each gene region on the support given to each node ofthe most parsimonious tree derived from a combined analysis of all three genes using Partitioned Bremer Support.We also explore the influence of assuming equal weights for all characters in the combined analysis by investigatingthe stability of clades to different transition/transversion weighting schemes. We find many strongly supported andstable clades in the Nymphalinae. We are also able to identify ‘rogue’ taxa whose positions are weakly supported (thedifferent gene regions are in conflict with each other) and unstable. Our main conclusions are: (1) the tribe Coeinias currently constituted is untenable, and
Smyrna, Colobura
and
Tigridia
are part of Nymphalini; (2) ‘Kallimini’ isparaphyletic with regard to Melitaeini and should be split into three tribes: Kallimini
s.s.
, Junoniini and Victorinini;(3) Junoniini, Victorinini, Melitaeini and the newly circumscribed Nymphalini are strongly supported monophyleticgroups, and (4)
Precis
and
Junonia
are not synonymous or even sister groups. The species
Junonia coenia
, a modelsystem in developmental biology, clearly belongs in the genus
Junonia
. A dispersal-vicariance analysis suggests thatdispersal has had a major effect on the distributions of extant species, and three biotic regions are identified as beingcentres of diversification of three major clades: the Palaearctic for the
Nymphalis
-group, the Afrotropics for Junoniiniand the Nearctic for Melitaeini. © 2005 The Linnean Society of London,
Biological Journal of the Linnean Society
,2005,
86
, 227–251.
ADDITIONAL KEYWORDS:
DIVA – molecular phylogeny – Partitioned Bremer Support – sensitivity analysis.
INTRODUCTION
Butterflies belonging to the currently recognized sub-family Nymphalinae (Wahlberg, Weingartner & Nylin,2003b) have contributed extensively to our knowledgeof ecological and evolutionary processes, from hybridzones and ring-species (Forbes, 1928; Silberglied,1984; Dasmahaptra
et al
., 2002; Austin
et al
., 2003),metapopulation dynamics (Hanski, 1999) and evolu-tionary developmental biology (Carroll
et al
., 1994;Brakefield
et al
., 1996), to insect–plant interactions(Singer, 1971; Nylin, 1988; Janz, Nylin & Nyblom,
2001; Wahlberg, 2001). Despite the long-standinginterest in these butterflies, the phylogenetic relation-ships among the various tribes and genera haveremained remarkably obscure. Improving our under-standing of the phylogenetic resolution of such scien-tifically popular taxa should be a high priority, so thatthis abundance of knowledge can be placed in an evo-lutionary framework.
Since Nymphalinae is the type subfamily of thediverse family Nymphalidae, its delineation hasenjoyed a dynamic history, as various authors haveconsidered diverse subsets of tribes and genera to rep-resent ‘typical nymphalids’. This confusion has led toseveral competing classification schemes based on dif-ferent data sets (Ackery, 1988; Harvey, 1991; Kuznet-
228
N. WAHLBERG
ET AL
.
© 2005 The Linnean Society of London,
Biological Journal of the Linnean Society,
2005,
86
, 227–251
zov & Stekolnikov, 2001; Wahlberg
et al
., 2003b). Thehigher systematics of Nymphalidae is still in a state offlux and the delineation of Nymphalinae has not yetreached stability, though there is a growing consensusthat the classification of Harvey (1991) (based on theclassification of Müller, 1886), with the addition of thetribe Coeini (= Coloburini of authors), currently pro-vides the most natural definition of the subfamily(Brower, 2000; Wahlberg
et al
., 2003b; Freitas &Brown, 2004). This concept of Nymphalinae comprisesa monophyletic group that includes the supposedlywell-defined tribes Nymphalini, Coeini, Melitaeiniand Kallimini. It is this hypothesis of nymphalinerelationships that we take as our point of departure inour analysis and discussion.
Ehrlich’s (1958) influential paper on the classifica-tion of butterflies included a much broader concept ofNymphalinae, which was based mainly upon symple-siomorphic and homoplastic characters and appearedto comprise those taxa that do not fall into any of hismore restricted and homogeneous nymphalid subfam-ilies (Danainae, Ithomiinae, Satyrinae, Morphinae,Calinaginae, Charaxinae, Acraeinae). The taxa in Ehr-lich’s Nymphalinae are today considered to representseveral different subfamilies, including Heliconiinae,Limenitidinae, Biblidinae and Apaturinae. Severalsubsequent authors have used much the same groupof tribes and genera to represent the ‘core nymphalids’(e.g. Ackery, 1984; Scott, 1985; Scott & Wright, 1990),though often splitting off some components as distinctsubfamilies. At the other extreme, some authors havedelineated Nymphalinae in a very narrow sense, toinclude only the tribes Nymphalini and Kallimini(Ackery, 1988) or Nymphalini and Melitaeini (Clark,1948).
All of the above classifications have suffered from alack of clearly described morphological synapomor-phies that diagnose the various circumscriptions. Har-vey (1991) proposed a classification of the familyNymphalidae based on a set of larval characters, thatwas accepted by many authors (e.g. Ackery
et al.
,1999). He placed the tribes Nymphalini, Kallimini andMelitaeini together, based on the arrangement ofspines on the larvae, but was unable to resolve therelationships among them due to character conflictwithin the subfamily. Harvey (1991) placed the genus
Amnosia
in the tribe Kallimini, but Wahlberg
et al
.(2003b) showed that
Amnosia
does not belong inNymphalinae, but in Cyrestinae along with othermembers of Pseudergolini (with which
Amnosia
wasaffiliated prior to being moved to Kallimini (Ackery,1988).
Most historical classifications of the Nymphalidaehave been intuitive rather than the product of rigor-ous phylogenetic analysis of character state distribu-tions formalized in a data matrix. More recently,
however, several cladistic analyses have been pub-lished, based on either morphology (DeVries
et al.
,1985; de Jong
et al.
, 1996; Penz & Peggie, 2003; Freitas& Brown, 2004) or DNA sequences (Weller
et al.
, 1996;Brower, 2000; Wahlberg
et al
., 2003b). Three of thesestudies (Brower, 2000; Wahlberg
et al
., 2003b; Freitas& Brown, 2004) sampled enough representatives ofNymphalidae to provide evidence bearing upon themonophyly and circumscription of the Nymphalinae.In all three, sampled taxa belonging to Nymphalini,Coeini, Kallimini and Melitaeini form a monophyleticgroup, with Coeini as the sister group to Nymphalini.Freitas & Brown (2004) and Wahlberg
et al
. (2003b)suggest an association of Kallimini with Melitaeini,while Brower’s (2000) study has a basal, paraphyleticKallimini, with regard to Melitaeini and Nymphalini
+
Coeini. The sister group to the Nymphalinae remainsin doubt, with Freitas & Brown (2004) proposing Hel-iconiinae, Brower (2000) suggesting Biblidinae
+
Apa-turinae and Wahlberg
et al
.’s (2003b) data implyingApaturinae.
Within Nymphalinae, several taxa have receivedattention from systematists in recent years. The rela-tionships among genera and species groups have beeninvestigated in Melitaeini (Kons, 2000; Wahlberg &Zimmermann, 2000) and Nymphalini (Nylin
et al
.,2001; Wahlberg & Nylin, 2003) using both molecularand morphological data. In Melitaeini, it has becomeclear that Harvey’s (1991) proposed division into threesubtribes (Euphydryina, Melitaeina, Phyciodina) isnot satisfactory. Wahlberg & Zimmermann (2000),based on mtDNA, found that the
Euphydryas
-,
Meli-taea-, Chlosyne-
and
Phyciodes-
groups of species andgenera were of equal status. Kons (2000), using mor-phology, also found these four major groups andadditionally described a fifth group containing
Gna-thotriche
species. However, the relationships of thefour major groups are in conflict between these twostudies. Wahlberg and Zimmermann inferred the
Mel-itaea
-group to be sister to the
Chlosyne-
group,whereas Kons found Phyciodina and
Gnathotriche
tobe sister to the
Chlosyne
-group.The two studies on Nymphalini (Nylin
et al
., 2001;Wahlberg & Nylin, 2003) have established the mono-phyly of the the
Nymphalis
-group of genera (i.e.
Aglais, Nymphalis
and
Polygonia
) and the close rela-tionship between
Araschnia, Mynes
and
Symbrenthia
,but otherwise the relationships of genera withinNymphalini remain unclear. A further recent morpho-logical study of
Symbrenthia, Mynes
and
Araschnia
(Fric, Konvicka & Zrzav
y
, 2004) suggests that
Mynes
is nested within
Symbrenthia
.In addition to these tribal-level studies, species-
level phylogenetic studies have been done for the gen-era
Euphydryas
(Zimmermann, Wahlberg & Desci-mon, 2000),
Chlosyne
(Kons, 2000),
Hypanartia
PHYLOGENY OF NYMPHALINAE
229
© 2005 The Linnean Society of London,
Biological Journal of the Linnean Society,
2005,
86
, 227–251
(Willmott, Hall & Lamas, 2001),
Anartia
(Blum, Ber-mingham & Dasmahapatra, 2003),
Phyciodes
(Wahl-berg, Oliveira & Scott, 2003a) and
Symbrenthia
,
Mynes
and
Araschnia
(Fric
et al
., 2004). Of these, onlyKons (2000), Wahlberg
et al
. (2003a) and Fric
et al
.(2004) included sufficiently extensive outgroup sam-pling to test the monophyly of the genus being studied.
From this brief review of the current state ofnymphaline systematics, it is clear that many ques-tions remain unanswered. The relationships amonggenera in Kallimini and Coeini have never been inves-tigated, and relationships within Melitaeini andNymphalini are still contentious. In this study we testthe monophyly of the most recent definition of Nymph-alinae, comprising Nymphalini, Coeini, Kallimini andMelitaeini (Wahlberg
et al
., 2003b), and endeavour toresolve the relationships among the tribes and generawithin the subfamily. In the current circumscription,the subfamily Nymphalinae comprises about 496 spe-cies in 56 genera. We have studied the relationships ofrepresentative species belonging to the subfamily withDNA sequences from three gene regions. These arecytochrome oxidase subunit I (COI) from the mito-chondrial genome, and elongation factor-1
a
(EF1-
a
)and
wingless
from the nuclear genome. Based on ourresults, we investigate the biogeography of the groupusing dispersal-vicariance analysis (Ronquist, 1997).The investigation is intended to identify broad biogeo-graphical patterns for closer inspection at a later date.
MATERIAL AND METHODS
We sampled as many species as possible from almostall of the genera belonging to Nymphalinae, a total of161 species in 49 genera. Of the seven missing genera,six are in Melitaeini and one is in Coeini. In addition,we sampled 28 outgroup species, representing all sub-families of the nymphaline clade (
sensu
Wahlberg
et al
., 2003b), i.e. Cyrestinae, Biblidinae and Apaturi-nae, each of which may be the sister group to Nymph-alinae (see Wahlberg
et al
., 2003b). We also included aspecimen of Heliconiinae and Limenitidinae, whichbelong to the putative sister clade to the nymphalineclade. The species sampled and their collection locali-ties are listed in Appendix 1.
We extracted DNA mainly from one or two legs offreshly frozen or dried butterflies using QIAgen’sDNEasy extraction kit, although some specimenswere extracted following the protocol of Brower (1994).The spread voucher specimens can be viewed at http://www.zoologi.su.se/research/wahlberg/. For each spec-imen we sequenced 1450 bp of COI, 1077 bp of EF1-
a
and 400–403 bp of the
wingless
gene. Primers for COIwere taken from Wahlberg & Zimmermann (2000), forEF1-
a
from Monteiro & Pierce (2001) and for
wingless
from Brower & DeSalle (1998). We performed all PCRs
in a 20
m
L reaction volume. The cycling profile for COIand
wingless
was 95
∞
C for 5 min, 35 cycles of 94
∞
C for30 s, 47
∞
C for 30 s, 72
∞
C for 1 min 30 s and a finalextension period of 72
∞
C for 10 min. The cycling pro-file for EF1-
a
was 95
∞
C for 7 min, 35 cycles of 95
∞
Cfor 1 min, 55
∞
C for 1 min, 72
∞
C for 2 min and a finalextension period of 72
∞
C for 10 min. For all threegenes, the PCR primers were also used for sequencing.
In addition, we developed two internal primers forsequencing: (EFmid 5
¢-
CAA TAC CRC CRA TTTTGT
-
3
¢
) for EF1-
a
and (Patty 5
¢-
ACW GTW GGWGGA TTA ACW GG
-
3
¢
) for COI. Sequencing was donewith a Beckman-Coulter CEQ2000 or CEQ8000 capil-lary sequencer (Bromma, Sweden). We checked theresulting chromatograms using BioEdit (Hall, 1999)and aligned the sequences by eye. Clear heterozygouspositions in the nuclear genes (chromatogram peaksalmost or exactly equal) were coded according to theIUPAC ambiguity codes, but were treated as missingcharacters in further analyses. The sequences havebeen submitted to GenBank (Accession numbers inAppendix 1).
We searched for the most parsimonious cladogramsfrom the equally weighted and unordered data matrixconsisting of 189 taxa using a heuristic search algo-rithm in NONA 2.0 (Goloboff, 1998) via WINCLADA1.00.08 (Nixon, 2002). The heuristic searches wereconducted with 1000 random addition replicates usingTBR branch swapping with ten trees held during eachstep and a final swapping to completion. We did thisfor each gene separately and for all three genes com-bined. Trees were rooted with
Heliconius
for display.For the separate analyses, we evaluated clade supportusing bootstrap with 100 pseudoreplicates. It is nowwidely recognized that assessing incongruence amongdata partitions is much more complex than measuringwith a simple all-or-nothing significance test (Farris
et al
., 1994; DeSalle & Brower, 1997; Miller et al.,1997; Darlu & Lecointre, 2002). We have thus chosento analyse the three gene regions as a single data set,and have assessed the impact of each gene region onthe support values of each node using PartitionedBremer Support (PBS) analyses (Baker & DeSalle,1997; Baker et al., 1998).
We evaluated the character support for the clades inthe resulting cladograms using Bremer support (BS)(Bremer, 1988, 1994). We calculated BS and PBSvalues using anticonstraints in PAUP* 4.0b10 forWindows (Swofford, 2001). As in previous studies(Wahlberg & Nylin, 2003; Wahlberg et al., 2003b), werefer to the support values as giving weak, moderate,good or strong support when discussing our results.We define ‘weak support’ as BS values of 1–2 (gener-ally corresponding to bootstrap values <50%–63%),‘moderate support’ as values between 3 and 5 (boot-strap values 64%-75%), ‘good support’ as values
230 N. WAHLBERG ET AL.
© 2005 The Linnean Society of London, Biological Journal of the Linnean Society, 2005, 86, 227–251
between 6 and 10 (bootstrap values 76%-88%) and‘strong support’ as values >10 (bootstrap values 89%-100%). We strongly endorse BS values over bootstrapvalues because they are a parameter of the datarather than an estimate based on manipulated sub-samples of the data, and have no upper bound as sup-port for a given clade increases.
We evaluated the stability (sensu Wheeler, 1995;Judd, 1998; Giribet, 2003) of clades inferred for theequally weighted data set to different character-statetransformation weighting assumptions under parsi-mony searches using PAUP*. We weighted transver-sions 2, 3, 5, 7 and 10 times transitions for thecombined data set. Such sensitivity analyses may helpidentify potential instances of long branch attraction(Giribet, 2003), and can provide a valuable heuristictool to guide subsequent sampling strategies forrefinement of the current hypothesis. We refer toclades that are recovered under all the tested weight-ing schemes as stable.
We investigated the historical biogeography of thesubfamily using dispersal-vicariance analysis (Ron-quist, 1997) as implemented in DIVA (Ronquist,1996). Species distributions were recorded at the levelof zoological biomes, i.e. Nearctic, Neotropical,Afrotropical, Palaearctic, Oriental and Australasian.Clades for which component species had identical dis-tributions were collapsed into a single terminal. Themaximum number of ancestral areas was either notconstrained or restricted to two.
RESULTS
CHARACTERISTICS OF THE DATA SETS
The total combined data set consisted of 2935 nucle-otides, of which 12 positions were coded as gaps insome taxa. All inferred gaps were in the wingless dataset and represent indel events of whole codons. Theseinclude an inferred codon insertion in all Melitaeinisequences (as noted in Brower, 2000), an inferredcodon insertion in Tigridia acesta, an inferred codoninsertion in Rhinopalpa polynice and an inferredcodon deletion in the three species of Aglais (as notedin Nylin et al., 2001). All inferred indel events have
occurred between positions 91 and 137 in the Coloburadirce sequence (GenBank accession numberAY090162) and were easily detected when aligning byeye (flanking regions were relatively conserved). Gapswere treated as a ‘fifth base pair’ in all phylogeneticanalyses. The basic statistics for the three generegions are given in Table 1. All three gene regionsshowed an ample amount of variation, though EF1-awas less variable than the other two genes. The twonuclear genes show about equal base frequencies,while COI has a high AT bias, in accordance with allprevious publications on this gene in insects.
GENERAL PHYLOGENETIC PATTERNS
When analysed as separate data partitions, none ofthe three gene regions recovered the subfamilyNymphalinae as a monophyletic entity (data notshown). There are, however, several phylogenetic com-ponents common to the separate hypotheses: Meli-taeini and Nymphalini (as delineated below) aremonophyletic, the genera Junonia, Precis, Hypolim-nas, Yoma, Salamis and Protogoniomorpha togetherform a monophyletic group, and the Nymphalis-groupof species (see Wahlberg & Nylin, 2003) is monophyl-etic.
Combining the three data sets yields eight trees oflength 17547 steps (CI = 0.13, RI = 0.58), the strictconsensus of which gives a much clearer picture ofthe relationships among the various groups inNymphalinae (Figs 1–4). The subfamily as currentlydelimited is inferred to be polyphyletic, with thegenera Historis and Baeotus branching off close tothe base of the larger nymphaline clade (sensu Wahl-berg et al., 2003b). The monophyly of Nymphalinaewithout Historis and Baeotus receives moderate sup-port. Nymphalini and Melitaeini form monophyleticgroups, while Coeini and Kallimini are poly- andparaphyletic, respectively. Three genera tradition-ally placed in Coeini (Colobura, Tigridia andSmyrna) are associated with Nymphalini with goodsupport, while the two other sampled genera, His-toris and Baeotus, are outside Nymphalinae withweak support. Kallimini + Melitaeini has moderatesupport, but the kallimine taxa form a paraphyletic
Table 1. The basic statistics for the three molecular data sets in 174 species of Nymphalidae
Gene No. of sites No. variable No. informative
Empirical base frequencies (%)
T C A G
COI 1450 775 614 39.9 14.7 31.9 13.6EF1-a 1077 469 364 22.6 28.1 25.9 23.4wingless 412 238 192 21.0 26.2 24.9 27.9
PHYLOGENY OF NYMPHALINAE 231
© 2005 The Linnean Society of London, Biological Journal of the Linnean Society, 2005, 86, 227–251
Figure 1. Summary of strict consensus of eight trees found for the combined data set when all changes weighted equally(length = 17547, CI = 0.13, RI = 0.58), pruned to show only genera. For unpruned trees, see Figs 2–4. Numbers above thebranches are Bremer support values and numbers below are Partitioned Bremer Support values for the COI, EF1-a andwingless data partitions, respectively. Thickened branches are stable to changes in weighting schemes (transversionsweighted 1, 2, 3, 5, 7 and 10 times transitions).
HeliconiusAdelphaMarpesia chironMarpesia orsilochusChersonesiaCyrestisHistoris odiusHistoris acherontaBaeotus japetusBaeotus deucalionBaeotus aeilusBaeotus beotusDichorragiaStibochionaAmnosiaPseudergolisApaturaMimathymaTimelaeaEulaceuraAsterocampaChitoriaSeveniaDynamineHamadryasCatonepheleMysceliaPanaceaCallicoreNicaAriadneEurytelaBybliaMesoxanthaColoburaTigridiaSmyrnaAntanartiaSymbrenthiaAraschniaMynesHypanartiaVanessaAglaisNymphalisPolygoniaKallimoidesSiproetaNapeoclesMetamorphaAnartiaRhinopalpaVanessulaHypolimnasPrecisJunoniaSalamisYomaProtogoniomorphaDoleschalliaKallimaCatacropteraMallikaEuphydryasPoladryasMicrotiaChlosyneMelitaeaGnathotricheHigginsiusMaziaTegosaPhyciodesAnthanassaTelenassaCastiliaEresiaJanatella
Cyrestini
Coeini
Pseudergolini
Apaturinae
Biblidinae
Nymphalini
Victorinini
Junoniini
Kallimini
Melitaeini
3213.5,17.1,1.4
323.9,-6.0,-14.9
1
18.5,-2.6,-4.9
177.2,9.1,0.7
3127.7,11.4,-8.1
129.6,1.4,1 4
5.5,1.9,-3.4 2
176
20.5,1.3,-15.8
1028.2,0.8,-19.0
118.2,-5.7,-11.5
1314,3.4,-4.4
20.2,5.1,-3.3
2118.4,9.4,-6.8 2
-1.0,6.9,-3.9 20.4,6.0,-4.4
2-0.6,6.2,-3.6 12
1.5,4.3,6.2
1919.8,6.5,-7.4
27.9,-0.2,-5.7 1
22.1,-6.6,-14.5
4
14.1,-7.4,-2.7
9
518.6,-6.6,-7.0
122.2,-6.7,-14.5
121.9,-6.5,-14.4
3925.6,8.5,4.9
119.9,-5.4,-13.5
1721.3,2.8,-7.0 19
28.0,-0.1,-8.9 14
15.4,-0.2,-11.2
18.4,-2.6,-4.8
820.7,-9.9,-2.8 10
30
8.8,-2.8,-5
9.4,-2.2,-5.2
32.1,1.2,-16.3
22.5,-8.6,-4.9
8.7,-3.3,-4.4
16.8,-4.1,-2.7
12,10.9,7.1
232 N. WAHLBERG ET AL.
© 2005 The Linnean Society of London, Biological Journal of the Linnean Society, 2005, 86, 227–251
grade with respect to Melitaeini. The sister group ofNymphalinae (without Historis and Baeotus) is Bib-lidinae in the most parsimonious trees from the com-bined data set, although this clade has weak support(Fig. 1).
Results of the PBS analysis show that all three generegions are congruent at 52 of the 183 resolved nodes(Figs 1–4). These nodes generally have high BS values(range 3–63, mean 19) and tend to be concentrated inthe clades describing Nymphalini and Melitaeini. At
Figure 2. Relationships of sampled species from the tribe Nymphalini as delimited in Fig. 1. Tree statistics and branchthickness as in Fig. 1.
Aglais ioAglais milbertiAglais urticae
Antanartia abyssinica
Antanartia deliusAntanartia schaenia
Araschnia levana
Colobura dirce
Hypanartia bella
Hypanartia charonHypanartia kefersteini
Hypanartia letheHypanartia lindigii
Polygonia canace
Mynes geoffroyi
Nymphalis antiopaNymphalis californica
Nymphalis l albumNymphalis polychloros
Nymphalis xanthomelas
Polygonia c-album
Polygonia c-aureum
Polygonia comma
Polygonia egea
Polygonia faunus
Polygonia haroldi
Polygonia interrogationis
Polygonia oreas
Polygonia progne
Polygonia satyrus
Polygonia gracilis
Smyrna blomfildia
Symbrenthia lileaSymbrenthia hypatia
Symbrenthia hypselis
Tigridia acesta
Vanessa annabella
Vanessa atalanta
Vanessa braziliensis
Vanessa cardui
Vanessa gonerilla
Vanessa indica
Vanessa itea
Vanessa kershawi
Vanessa myrinna
Vanessa virginiensis
922.5,-8.6,-4.9
611.5,4.4,-9.9
1112.7,7.2,-8.9
421.5,-1.5,-16.0
419,-1.4,-13.5
821.2,-0.3,-12.9
165,0.9,10.1
269,11.4,
5.6
8
2,3.9,2.1
11,-1,1 6
4.5,-0.6,2.1 53.2,1.4,0.4 5
-4.5,5.4,4.1
23,-2.1,1.1
60.5,4.5,1.1
10.4,0.6,
011
5.6,4.2,1.2
79.3,-2.3,0 4
9.9,-5.2,-0.7 78,-1.1,0.1 1
7.3,-4.3,-2.0 17.2,-4.3,-1.9 4
5.4,-1.2,-0.2
3
2,1.9,-0.9
82.9,4.3,0.8
1-2.8,3.7,0.1
1-2.8,3.7,
0.17
2.9,4.6,-0.5
249.2,10.2,4.6 5
6, 2.1,1.1- 90.2,5.4,3.4
77.1,-1.1,
1.1
2511,6.9,
7.1
227,9.9,5.1
22,0.9,-0.9
1024.7,-4.4,
-10.3
6237,13.4,11.6
1723.9, 0.6,
-6.3- 2121.8,5.3,-6.2
93.6,2.7,
2.7
112.5,3.5,5.1 15
2.4,6.6,6.1
1933.6,-5.7,
-8.9
PHYLOGENY OF NYMPHALINAE 233
© 2005 The Linnean Society of London, Biological Journal of the Linnean Society, 2005, 86, 227–251
Figure 3. Relationships of sampled species from the tribes Junoniini and Victorinini, as delimited in Fig. 1, as well asspecies in the genera Kallimoides, Vanessula and Rhinopalpa. Tree statistics and branch thickness as in Fig. 1.
Anartia amatheaAnartia fatima
Anartia jatrophae
Hypolimnas alimena
Hypolimnas anthedon
Hypolimnas bolina
Hypolimnas misippus
Hypolimnas pandarus
Hypolimnas usambara
Junonia artaxia
Junonia coenia
Junonia erigoneJunonia iphita
Junonia natalica
Junonia oenoneJunonia sophia
Junonia terea
Junonia touhilimasa
Kallimoides rumia
Kamilla cymodoce
Metamorpha elissa
Napeocles jucunda
Precis andremiajaPrecis antilope
Precis archesiaPrecis ceryne
Precis cuama
Precis octaviaPrecis sinuata
Precis tugela
Protogoniomorpha anacardii
Protogoniomorpha cytora
Protogoniomorpha parhassus
Rhinopalpa polynice
Salamis antevaSalamis cacta
Siproeta epaphus
Siproeta stelenes
Vanessula milca
Yoma algina
1
22.2,-6.7,-14.5
319.9,-6.6,
-10.48
31.4,-6.9,-16.5
116.7,-5.6,-10.1 40
43.8,-0.7,-3.2 3340.6,0.1,-7.7
1627.5,-1.2,
-10.31
23.3,-8.0,-14.3
121.8,-6.4,-14.4
1927.2,1.1,-9.3
77.5,-0.3,
-0.2
179.7,-0.6,
7.9 115.5,4.3,1.2 2
2,-0.1,0.1 11,-0.1,0.1
726.8,-4.9,
-14.9
2531.3,1.7,-7.9
123.3,-7.4,-14.9
21.1,-0.2,1.1
23.2,-2.4,1.2 3
4,-1.1,0.1
23.0,-2.1,
1.1
423.6,-4.7,
-14.9
521.3,-5.6,-10.7
1314.5,-1.9,0.4
121.9,-6.5,-14.4
36.0,-2.1,
-0.916
29.2,0.3,-13.5
122,-6.5,-14.5
122.1,-6.6,-14.5
2335.2,-0.5,-11.7 4
20.6,-5.6,-11
2833.4,-1.5,
-3.9
2620,3.9,
2.1
3
6.9,0.5,-4.4 5
4.2,1.6,-0.8 3434,9.9,-9.9
6246.6,9.9,
5.5
418.4,-5.3,
-9.1
234 N. WAHLBERG ET AL.
© 2005 The Linnean Society of London, Biological Journal of the Linnean Society, 2005, 86, 227–251
Figure 4. Relationships of sampled species from the tribe Melitaeini and Kallimini as delimited in Fig. 1. Tree statisticsand branch thickness as in Fig. 1.
Anthanassa ardysAnthanassa drusilla
Anthanassa otanes
Anthanassa texanaAnthanassa tulcis
Castilia eranitesCastilia ofella
Catacroptera cloanthe
Chlosyne acastus
Chlosyne cyneas
Chlosyne gaudealis
Chlosyne gorgone
Chlosyne harrisii
Chlosyne janais
Chlosyne lacinia
Chlosyne narva
Chlosyne nycteis
Chlosyne palla
Chlosyne theona
Doleschallia bisaltide
Eresia clio
Eresia coela
Eresia eunice
Eresia letitia
Eresia peloniaEresia quintilla
Eresia sestia
Euphydryas aurinia
Euphydryas chalcedona
Euphydryas desfontainii
Euphydryas editha
Euphydryas gillettiiEuphydryas phaeton
Gnathotriche exclamationisHigginsius fasciata
Janatella leucodesma
Kallima paralektaKallima inachus
Mallika jacksoni
Mazia amazonica
Melitaea arduinna
Melitaea britomartis
Melitaea cinxia
Melitaea deione
Melitaea didymoidesMelitaea latonigena
Melitaea persea
Melitaea scotosiaMelitaea punicaMelitaea trivia
Melitaea varia
Dymasia dymasTexola elada
Microtia elva
Phyciodes batesii
Phyciodes cocytaPhyciodes graphica
Phyciodes mylittaPhyciodes orseis
Phyciodes pallescens
Phyciodes pallida
Phyciodes phaon
Phyciodes picta
Phyciodes pulchella
Phyciodes tharos
Poladryas arachne
Tegosa anietaTegosa tissoides
Telenassa trimaculata
121.9,-6.5,
-14.4
3925.6,8.5,4.9
2221.1,8.8,-7.9
69.3,2.1,-5.4
5-1,5.9,0.1
3312.9,15.5,4.6
815.2,-0.9,-6.3
28.3,-0.4,-6.0
67.7,1.4,-3.1
30.9,1.1,
1 93,3.9,2.1 3
1.2,-0.4,2.2
41.7,1.2,
1.1
34.3,-0.3,-1
32,-0.1,
1.115
4.9,8.6,1.5
55.8,0.6,-1.4
52.5,2.5,0 1
3.1,-1.3,-0.8 1012,-1.7,-0.3
3023.2,3.7,
3.2
156.9,5,3.1
42.7,2.2,
-0.9
155.4,11.7,
-2.128
14.4,11.2,2.4
36,-2.1,-0.9 6
3,2.9,0.1
2-2.9,7,-2.1
2-1.3,7.9,-4.6 2
1,4.5,-3.5 117.4,5.4,-1.8 4
4,0.9,-0.9
187.9,10.7,
-0.5
5531.5,17.7,
5.8
163.4,10,2.6
123.7,9.1,
-0.8 2-1.3,2.5,0.8
11.9,-1,0.1
2-1.3,2.3,1 25
6,13.9,5.1 1310,0.9,2.1
136,5.9,1.1
2211,8.9,2.1 9
7,1.9,0.1
187.6,7.4,
3
3-0.3,4.2,
-0.9 1-2.1,3.9,-0.8
179.1,8.2,-0.3
610.3,0.1,-4.3 1
-1.5,3.5,-1 66.2,-0.3,0.1 1
-1,2.9,-0.9 86.6,1.8,-0.4 4
0,4.9,-0.9
88.7,0.5,
-1.2
2910.1,8.6,
10.422
7.1,14.3,0.6
163,7,6 4
6.9,1.9,-4.8
3624.7,19.7,
-8.4 97.3,5.1,-3.3 30
20.2,6.7,3.1
2830.6,3.8,
-6.52
3,-1.1,0.1
119.9,-5.4,
-13.5 1615.7,2.2,-1.9 17
29,-0.6,-11.4
3737.8,4.3,
-5.5
PHYLOGENY OF NYMPHALINAE 235
© 2005 The Linnean Society of London, Biological Journal of the Linnean Society, 2005, 86, 227–251
131 nodes, one or two gene regions provide signal thatis incongruent with the weight of the combined evi-dence. The different patterns of incongruence are sum-marized in Table 2. Nodes which show incongruencetend to have lower BS values (range 1–19, mean 7.7).Most of the incongruent nodes lie in the basalbranches of the cladogram (Fig. 1) and in the kal-limine grade (Figs 3, 4).
The PBS results show that the data partitions arecontributing unequally to the phylogenetic patternsinferred in this study. The COI data set conflicts at 17nodes, of which one node has a PBS score of <-4. Thetotal PBS for the COI data partition is 2176.1 (mean11.9). The EF1-a data partition is incongruent at 70 ofthe 183 resolved nodes, 23 of which have a PBS scoreof <-4, and has a total PBS of 370.3 (mean 2.0). Thewingless data alone have a negative PBS score at 106of the 183 resolved nodes, 64 of which have a PBSscore of <-4, and the gene region has a total PBS of -533.5 (mean -2.9).
Many of the clades are very stable (sensu Giribet,2003) and are present in a strict consensus of all treesfound across the different weighting schemes from theequally weighted analysis to the analysis with 10 : 1TV/TI weighting (Figs 1–4). These include the largernymphaline clade (including Biblidinae, Cyrestinae,Apaturinae and Nymphalinae; see Wahlberg et al.,2003b), Apaturinae and Biblidinae, Pseudergolini,Cyrestini, Nymphalini and Melitaeini (Fig. 1), andmany of the well-defined genera (Figs 2–4). Weightingtransversions 2 times transitions is the only analysiswhich gives a monophyletic Nymphalinae, with His-toris + Baeotus coming out as sister to the rest ofNymphalinae.
PHYLOGENETIC PATTERNS IN COEINI
The tribe Coeini does not form a monophyletic groupin any of the analyses (Figs 1, 2). In the combinedanalysis, Smyrna is sister to Nymphalini with good
support, Colobura and Tigridia are sister genera withstrong support and they are sister to Nymphalini +Smyrna. The monophyly of Smyrna, Tigridia,Colobura and Nymphalini has good support and is sta-ble, though there is conflict from the nuclear genes(Fig. 1). Historis and Baeotus form a monophyleticgroup with strong support and both genera are mono-phyletic (Fig. 1). Historis odius and H. acheronta (thelatter sometimes placed in the genus Coea, from whichthe tribal name is derived) are sister species with onlyweak support and moderate conflict from the twonuclear genes. Baeotus, on the other hand, is astrongly supported monophyletic group with no con-flict from the different data partitions. Constrainingthe five coeine genera to form a monophyletic groupresults in trees that are 19 steps longer than the mostparsimonious trees found for the combined data sets.Constraining Colobura, Tigridia and Smyrna to forma monophyletic group results in trees seven stepslonger than the most parsimonious trees.
The association of Colobura, Tigridia and Smyrnawith Nymphalini is stable under all the differentweighting schemes, as is the clade containing Historisand Baeotus species. However, the position of Historis+ Baeotus changes as one increases the weight oftransversions to transitions. When transversions areweighted 2 times transitions, this clade is sister toNymphalinae. At higher tested weighting schemes(transversions weighted 3–10 times transitions), His-toris + Baeotus appears as the sister clade to Apaturi-nae.
PHYLOGENETIC PATTERNS IN NYMPHALINI
Within the monophyletic Nymphalini, relationshipswithin the Nymphalis-group of genera (i.e. Nymphalis,Polygonia and Aglais) are almost identical to thosehypothesized in a previous study (Wahlberg & Nylin,2003). The exception is Polygonia canace, which is sis-ter to Nymphalis in the present study, whereas it wassister to the rest of Polygonia in the 2003 study. Wahl-berg & Nylin (2003) was based on an additional genesequence and a morphological data set and thus weprefer the hypothesis supported in that study. Polygo-nia oreas and P. haroldi were not included in previousphylogenetic studies of the Nymphalis-group, and ourresults suggest that the former is related to P. gracilisand the latter is sister to a clade containing P. oreas,P. gracilis and P. satyrus.
Other genera in Nymphalini are here sampled morebroadly than in previous studies and we are now ableto identify three additional clades within Nymphalini.Two species of Antanartia (A. delius and A. schaenia)form the sister group to the rest of Nymphalini. Thenext most basal clade is formed by the genera Ara-schnia, Mynes and Symbrenthia, which group together
Table 2. Number of nodes in the ingroup at which differ-ent patterns of incongruence were observed. +, PBS scorepositive or 0; -, PBS score negative (magnitude not takeninto account)
Number of nodes COI EF1-a wingless
47 + + +46 + + –20 + – +
6 – + +56 + – –
0 – – +8 – + –
236 N. WAHLBERG ET AL.
© 2005 The Linnean Society of London, Biological Journal of the Linnean Society, 2005, 86, 227–251
with good support. The third major new clade com-prises the genera Vanessa and Hypanartia, thoughthis clade has only moderate support and is not stable.Vanessa + Hypanartia is sister to the Nymphalis-group clade. A surprising result is the position ofAntanartia abyssinica within Vanessa; it is not at allclosely related to the other Antanartia species. WithinVanessa, species often placed in the genus Cynthia donot form a monophyletic group, as one species,V. annabella (and presumably its putative sister spe-cies V. carye; Shapiro & Geiger, 1989), appears to besister to the rest of Vanessa.
Sensitivity analyses suggest that the relationshipswithin the Nymphalis-group are very stable; the onlynode which is unstable is the sister relationshipbetween Polygonia egea and P. c-album + P. faunus.Hypanartia is stable, as are the relationships of spe-cies within the genus. Within Vanessa, the sister rela-tionship of A. abyssinica to V. atalanta + V. indica isstable, as is the sister species relationship ofV. gonerilla + V. itea, and the V. cardui clade (V. carduito V. myrinna in Fig. 2). Other stable clades areAntanartia delius + A. schaenia and Symbrenthia.
PHYLOGENETIC PATTERNS IN KALLIMINI
The tribe Kallimini as currently circumscribed doesnot form a monophyletic group, but rather a paraphyl-etic grade with regard to Melitaeini. Constraining it tobe monophyletic results in trees that are four stepslonger than the most parsimonious trees found for thecombined data set. The strict consensus of the mostparsimonious trees shows two clades. The basal cladecontains the following subclades: (1) a largely Neotro-pical subclade including Anartia, Siproeta, Napeoclesand Metamorpha (termed the Anartia-clade), (2) theAfrican Kallimoides, a subclade with Vanessula andRhinopalpa (both monotypic genera that have not pre-viously been associated with each other or indeed withany other kallimine genera) and (3) a subclade thatcontains the largely African and Asian generaJunonia, Kamilla, Precis, Hypolimnas, Salamis, Pro-togoniomorpha and Yoma (termed the Junonia-clade).The second clade in the Kallimini grade, which is sis-ter to the tribe Melitaeini, includes the African Catac-roptera and Mallika, the Asian Kallima and theAustralasian Doleschallia (termed the Kallima-clade).
Weighting transversions more than transitionscauses Kallimoides and Vanessula to become sistertaxa. These two form the sister clade to Rhinopalpaand the Anartia- and Junonia-clades when transver-sions are weighted 2 or 3 times transitions. However,weighting schemes of 5 and above cause Kalliomoides,Vanessula, Rhinopalpa and the Anartia-clade tobecome the most basal clades in the entire tree, afterthe outgroups Heliconius and Adelpha.
Within the Anartia-clade, Napeocles and Siproetaform a strongly supported, stable subclade, as do thethree species of Anartia. The sister group of Anartiamay be either Siproeta/Napeocles or Siproeta/Napeo-cles + Metamorpha. It is clear that Metamorpha is anentity distinct from Siproeta (as suggested by Fox &Forbes, 1971), though historically Siproeta steleneshas occasionally been placed in Metamorpha. Theclade containing these neotropical genera is stable insensitivity analyses. The position of Kallimoides asthe most basal taxon of this clade has only weak sup-port and the node is not stable, suggesting that itsplacement requires further investigation.
The Junonia-clade exhibits some of the greatest sur-prises of this study. The first is that Junonia and Pre-cis are separate genera that are not even sistergroups. The genus Kamilla is clearly within Junonia,as was concluded by Shirôzu & Nakanashi (1984)based on morphology (but see Larsen, 1991). In addi-tion, the species Protogoniomorpha cytora is foundwithin Junonia, rather than with other speices of Pro-togoniomorpha or Salamis, with which it has alwaysbeen associated. The sister group relation betweenPrecis and Hypolimnas has good support and is stable,except when transversions are weighted 3 times tran-sitions, when it is sister to the Salamis + Yoma + Pro-togoniomorpha + Junonia clade. Another surprise isthat the African Protogoniomorpha, which has beenusually considered to be a synonym of Salamis, is thesister group to the Australasian Yoma, and that Sala-mis is the sister group to Protogoniomorpha + Yoma.Vári (1979) has argued that species of Protogoniomor-pha should not be considered to be congeneric withspecies of Salamis based on genitalic differences, aposition corroborated by our data. The sister grouprelationship of Yoma and Protogoniomorpha has goodsupport and is stable. To complete the surprise, Sala-mis, Yoma and Protogoniomorpha, usually consideredto be related to Hypolimnas, are placed in our resultsas sister group to Junonia, though this clade disap-pears when transversions are weighted 5 or moretimes transitions.
The implied position of the Kallima-clade as sisterto the Melitaeini is somewhat surprising given tradi-tional classifications, but this grouping has weak sup-port and may be due to long branch attraction. Indeedthe sister group relationship is not stable and disap-pears when transversions are weighted 2 or moretimes transitions. The relationships between Catac-roptera, Mallika and Kallima are strongly supportedand stable. The strongly supported, stable sister rela-tionship of Catacroptera and Mallika (both monotypic)has been suggested earlier by Shirôzu & Nakanishi(1984). Catacroptera and Mallika are usually associ-ated with Junonia rather than Kallima (Larsen,1991). The recently suggested (Parsons, 1999) associ-
PHYLOGENY OF NYMPHALINAE 237
© 2005 The Linnean Society of London, Biological Journal of the Linnean Society, 2005, 86, 227–251
ation of Doleschallia with Kallima is corroborated byour data, although support is not strong and there isstrong conflict between the nuclear genes and themitochondrial gene regions.
PHYLOGENETIC PATTERNS IN MELITAEINI
The monophyly of Melitaeini is very strongly sup-ported and stable, as are Euphydryas, Melitaea,Phyciodina and the sister relationship between Gna-thotriche and Higginsius (termed the Gnathotriche-group). The Chlosyne-group has moderate support butis unstable, with both the COI and wingless generegions conflicting with the EF1-a data. The genusChlosyne itself is a strongly supported, stable clade.The position of Euphydryas as the sister group to therest of the melitaeines is strongly supported and sta-ble, in agreement with previous studies (Kons, 2000;Wahlberg & Zimmermann, 2000). Relationshipsbetween the Chlosyne-, Melitaea-, Gnathotriche-groups and Phyciodina are not well-supported and areunstable. The most parsimonious trees from theequally weighted analysis suggests that the Gnathot-riche-group is sister to Phyciodina with moderate sup-port, and that the Melitaea-group is sister to these twowith good support (Fig. 4). This arrangement is stablewhen transversions are weighted 2 and 3 times tran-sitions, but breaks down at higher weights. The sisterrelationship of the Gnathotriche-group and Phyciod-ina is not stable and disappears after weights of 3times. This study presents a third novel arrangementof these four clades, with Kons (2000) having the Chlo-syne-group sister to Phyciodina + Gnathotriche-groupand Wahlberg & Zimmermann (2000) having theChlosyne-group sister to the Melitaea-group.
The positions of two genera (Poladryas and Higgin-sius) are not in agreement with previous studies. Incontrast to Wahlberg & Zimmermann (2000) but inagreement with Kons (2000), Poladryas is here relatedto the Chlosyne-group. Morphological evidence thatPoladryas is associated with Higginsius (Kons, 2000)is quite decisively contradicted by molecular evidence,which places Higginsius as sister to Gnathotriche withstrong support and stability. Within the Chlosyne-group, the genera Microtia, Texola and Dymasia forma strongly supported, stable monophyletic group, asfound by Kons (2000).
Relationships of the Neotropical Phyciodina (repre-sented by the genera Mazia, Tegosa, Eresia, Castilia,Telenassa, Janatella and Anthanassa) are generallynot well-supported or stable at the deeper nodes. Ere-sia, Castilia, Telenassa, Janatella and Anthanassa doform a stable monophyletic group to the exclusion ofTegosa and Mazia, even though the wingless data par-titions is in conflict with the COI and EF1-a data atthe node.
DISCUSSION
This is only the second comprehensive phylogeneticanalysis of the relationships within a nymphalidsubfamily, and the first to use molecular data. Thesole previous attempt to infer relationships within anymphalid subfamily is the recently published studyby Penz & Peggie (2003) on Heliconiinae. This is notaltogether surprising, given that the circumscrip-tions of the subfamilies have only recently stabi-lized (Harvey, 1991; Wahlberg et al., 2003b), andbecause the high degree of variation in morphologi-cal characters among species in Nymphalidae hasconfounded previous attempts to delineate naturalgroups (de Jong et al., 1996). In our study, we havebeen able to identify clades that are well-supportedby the three gene regions and stable to varied char-acter state transformation weights, as well as cladesthat are less robust and therefore likely to changewith the addition of more data. Based on ourresults, we are proposing a new classification of thesubfamily Nymphalinae, which is shown in Appen-dix 2.
THE CIRCUMSCRIPTION OF NYMPHALINAE
We have found that Nymphalinae as currentlycircumscribed is not monophyletic, although Nym-phalini, Kallimini and Melitaeini do form a mono-phyletic group with the inclusion of three generatraditionally placed in Coeini (i.e. Colobura,Tigridia and Smyrna). Two genera also tradition-ally placed in Coeini (Historis and Baeotus) areclearly not related to these three genera and indeeddo not appear to be closely related to Nymphali-nae. Our results are not unprecedented, as Muys-hondt & Muyshondt (1979) argued that based onlarval morphology Smyrna should be placed inNymphalini and that Colobura is closer to Nympha-lini than Historis and Baeotus. However, the sug-gestion by Muyshondt & Muyshondt (1979) thatCoeini is an unnatural group has not been followedby subsequent authors. On the other hand, Freitas& Brown (2004) found that, based on morphologi-cal data, Historis, Colobura and Smyrna formed amonophyletic group sister to Hypanartia and Van-essa. We were unable to sample the remainingostensibly coeine genus, Pycina, which has morpho-logical features in larvae and pupae that appear tobe intermediate between those in Historis/Baeotusand Colobura (Muyshondt & Muyshondt, 1979). Theunstable behaviour of the Historis + Baeotus cladein our study and its quite different position in thestudy by Freitas & Brown (2004) suggests thatmore data are needed to clarify its relationshipwithin Nymphalidae.
238 N. WAHLBERG ET AL.
© 2005 The Linnean Society of London, Biological Journal of the Linnean Society, 2005, 86, 227–251
THE CIRCUMSCRIPTION OF NYMPHALINI
The well-supported association of Colobura, Tigridiaand Smyrna with Nymphalini suggests that theyshould be incorporated into the tribe, though the tra-ditional delineation of the tribe is also well-supportedand, in addition, is stable.
The placement of Antanartia abyssinica within Van-essa is clear. Vanessa (including A. abyssinica) is awell-supported clade and the sister relationship ofA. abyssinica to V. atalanta + V. indica is stable.Indeed, larval morphology of A. abyssinica corrobo-rates our evidence that it is unrelated to otherAntanartia species and related to Vanessa (Nakanishi,1989). Antanartia has been divided into two species-groups, the delius-group, comprising A. delius,A. schaenia and the unsampled A. borbonica, and thehippomene-group, to which A. abyssinica, and theunsampled A. hippomene and A. dimorphica belong(Howarth, 1966). Whether A. hippomene andA. dimorphica should also be placed in Vanessa is notclear; Nakanishi (1989) noted that the larvae ofA. abyssinica differed greatly from those ofA. schaenia and A. hippomene. Clearly, the missingspecies need to be sampled.
The divergent position of Antanartia with respect toHypanartia is also surprising, as all species wereincluded in Hypanartia prior to the description ofAntanartia by Rothschild & Jordan (1903). Antanartiahas always been assumed to be the sister group ofHypanartia (e.g. Willmott et al., 2001). However, ourresults suggest that the sister group to Hypanartia isVanessa and that Antanartia is sister to all the rest ofthe traditionally recognized Nymphalini. These posi-tions are stable when transversions are weighted 2and 3 times transitions, but they break down at higherweights. Weighting 5 and 7 times suggests thatAntanartia is sister to the Symbrenthia/Mynes/Ara-schnia clade and that Hypanartia is sister to these. Inno analysis does Antanartia appear as the sister toHypanartia.
Vanessa is usually thought to be the sister group tothe Nymphalis-group of genera (Wahlberg & Nylin,2003). Our study suggests that Vanessa is sister toHypanartia and that these two clades make up the sis-ter group to the Nymphalis-group. This arrangementis stable when transversions are weighted 2 and 3times transitions, while at higher weights only Van-essa is sister to the Nymphalis-group. Thus, it is likelythat either Vanessa or Vanessa + Hypanartia is morerelated to the Nymphalis-group than to the other gen-era in Nymphalini, and further new characters(molecular and morphological) will help in refining therelationships further.
The relationships of Symbrenthia, Mynes and Ara-schnia are rather surprising, especially since the sis-
ter relationship between Mynes and Araschnia isstable up to a weighting of transversions 7 times tran-sitions. A recent morphological study of this groupfound that Mynes is within Symbrenthia (Fric et al.,2004). Normally, Symbrenthia has been associatedwith Mynes (Parsons, 1999; Nylin et al., 2001; Wahl-berg & Nylin, 2003). For our results to be congruentwith those of Fric et al. (2004), we would have to findthat Mynes geoffroyi is sister to Symbrenthia hypselis,which is clearly not the case. Obviously, more speciesof all three genera need to be sampled to resolve theconflicting results of the two studies.
This study contains a much more comprehensivesampling of species in Nymphalini than our two pre-vious studies (Nylin et al., 2001; Wahlberg & Nylin,2003). The Nymphalis-group continues to be a well-supported and stable clade, though the stable positionof Polygonia canace as sister to species in the genusNymphalis is in contrast to earlier results. Polygoniacanace is usually placed in its own genus Kaniska, andthe uncertainty of its position may warrant the use ofthat genus name. However, our study confirms the sta-ble relationship of Aglais io (usually placed in its owngenus Inachis) with other species of Aglais and thestable relationship of Nymphalis l-album with otherspecies of Nymphalis. Also stable and well-supportedis the sister relationship of Polygonia and Nymphalis,with Aglais as sister to these two. This study includestwo species of Polygonia that have not been includedin previous phylogenetic studies, P. oreas andP. haroldi. The position of P. oreas as sister toP. gracilis is surprising as it is often considered to be asubspecies of P. progne (Scott, 1984). Polygoniaharoldi is a little known species and morphologically itis somewhat intermediate between P. progne andP. satyrus. The relationships of species in Polygoniaare being investigated in more detail currently (E.Weingartner, N. Wahlberg & S. Nylin, unpubl. data).
THE FATE OF KALLIMINI
The monophyly of Kallimini appears to be doubtful.The tribe, as previously circumscribed, never formed amonophyletic group in our analyses. Species belongingto Kallimini have occasionally been placed in highertaxa of their own, and our results show that some ofthese groups are strongly supported and stable. Thesewell supported groups of genera should be recognizedat the tribal level, unless one cares to resort to the (inour opinion, unacceptable) synonymization of Kal-limini with Melitaeini. Names are available for twotribes: Victorinini Scudder, 1893 for the Anartia-cladeand Junoniini Reuter, 1896 for the Junonia-clade. TheKallima-clade constitutes the newly circumscribedtribe Kallimini. The three remaining genera (Kalli-moides, Vanessula and Rhinopalpa) have to remain
PHYLOGENY OF NYMPHALINAE 239
© 2005 The Linnean Society of London, Biological Journal of the Linnean Society, 2005, 86, 227–251
incertae sedis until their positions are stabilized byfurther investigation.
There are 11 species in four genera in the newly cir-cumscribed tribe Victorinini. The close relationship ofNapeocles and Siproeta has never been considered,though Metamorpha has been considered to be aderived Siproeta (Fox & Forbes, 1971). Perhaps thesuperficial similarity of Metamorpha to Siproeta andthe highly distinctive wing patterns of Napeocles hasled previous investigators to ignore the latter. How-ever, weighting transversions higher than transitionscauses Napeocles to become sister to S. stelenes, whichis also recovered in some of the most parsimonioustrees of the equally weighted analysis. All weightingschemes other than equal weighting also recover a sis-ter relationship between Metamorpha and Anartia,suggesting that Metamorpha is not as closely relatedto Siproeta as previously thought. The relationships ofthe three species of Anartia included in this study arein concordance with a previous study on the genus(Blum et al., 2003). It is clear from our study that Vic-torinini is a well-defined entity that deserves the rankof tribe in the Nymphalinae.
The Junoniini, as circumscribed here, is also a well-defined and stable group. The relationships of the gen-era in Junoniini found in the equally weighted analy-sis are stable under fairly severe weighting schemes(transversions weighted up to 5 times transitions). Athigher weights, the Yoma + Protogoniomorpha cladebecomes sister to the Precis + Hypolimnas clade, andSalamis is sister to the rest of Junoniini. The closerelationships of Protogoniomorpha and Yoma and Pre-cis and Hypolimnas are stable under all weightingschemes. The clean separation of Precis and Junoniain our study is both surprising and gratifying. Thesetwo genera have long been conflated in the literature,despite their biogeographical disjunction and the factthat de Lesse (1952) showed clear genitalic differencesbetween them. de Lesse’s (1952) delimitations of thetwo genera are corroborated here. For North Americanresearchers it is important to emphasize that all NewWorld species (including the well-studied J. coenia)belong to Junonia and that Precis is restricted toAfrica.
The once common concept of Kallima (i.e. containingKallimoides, Kamilla and Mallika) is clearly untena-ble, as firmly concluded by Shirôzu & Nakanishi(1984). The Kallimini, as delimited here, contains onlyfour genera: Kallima, Catacroptera, Mallika and Dole-schallia.
THE DELINEATION OF MELITAEINI
The monophyly of Melitaeini is beyond doubt. Allthree gene regions support the clade and it is stable ina variety of weighting schemes. The sister group to
Melitaeini appears to be the newly circumscribed Kal-limini and/or Junoniini. Within Melitaeini, the sisterrelationship of Euphydryas to the rest of Melitaeini isin agreement with all previous studies and has a clear,well-supported, stable position in this study. The restof Melitaeini appears to be divided up into four dis-tinct groups: Phyciodina, Melitaeina (including onlyMelitaea), the Gnathotriche-group and the Chlosyne-group. Of these, only the Chlosyne-group is not stable,though Poladryas remains associated with Chlosynewith transversions weighted up to 7 times transitions.The Microtia + Texola + Dymasia clade becomes thesister to the rest of Melitaeini (excluding Euphydryas)when transversions are weighted 5 and 7 times tran-sitions, though it returns to being sister to Chlosyne atthe highest weighting scheme. The close relationshipof Microtia, Texola and Dymasia was also found byKons (2000), who suggested that Texola and Dymasiashould be synonymized with Microtia.
The close relationship of Higginsius to Gnathotrichewas suggested by Higgins (1981), but Kons (2000)found Higginsius to be associated with Poladryas. Ourresults strongly corroborate Higgins’ (1981) hypothe-sis. The two genera are very curious members of theAndean fauna and appear to be always rare (Higgins,1981). They comprise a total of six species and, withthe unsampled Caribbean Atlantea and Antillea (alsovery species-poor and always rare), create somethingof a biogeographical mystery. We find that the sisterrelationship of Gnathotriche and Higginsius to Phycio-dina is stable with transversions weighted up to 7times transitions. When transversions are weighted10 times transitions, the Gnathotriche-group becomessister to Melitaea. Since Phyciodina is largely a Neo-tropical subtribe, its close relationship to the Gnathot-riche-group is perhaps not surprising, but the possibleorigin of both groups in South America requires fur-ther investigation.
Of the six missing genera, four (Tisona, Dagon, Orti-lia and Phystis) putatively belong to the subtribe Phy-ciodina (Higgins, 1981). The remaining two genera,Antillea and Atlantea, are restricted to the GreaterAntilles in the Caribbean and have until recently notbeen associated with other melitaeine genera. Kons(2000) placed both in the Chlosyne-group, withAtlantea being sister to Higginsius. It is clear thatthese two genera need to be sampled for moleculardata.
A NOTE ON PHYLOGENETIC PATTERNS IN THE OUTGROUPS
Although our study has concentrated on Nymphali-nae, we have extensively sampled several potentialsister groups to the subfamily. In particular, we havesampled all genera in the tribes Cyrestini and Pseud-
240
N. WAHLBERG
ET AL
.
© 2005 The Linnean Society of London,
Biological Journal of the Linnean Society,
2005,
86
, 227–251
ergolini, which constitute the Cyrestinae in Wahlberg
et al
. (2003b). However, in the present study the twotribes do not form a monophyletic group; rather,Cyrestini appears as the most basal group in thenymphaline clade and Pseudergolini is sister to Apa-turinae (Fig. 1). There are four clades in the outgroupappearing with good to strong support that correspondto the Apaturinae, Biblidinae, Cyrestini and Pseuder-golini. However, the relationships among these fourtaxa and Nymphalinae are very poorly supported, areunstable and show much conflict among partitions.Clearly, the nymphaline clade (
sensu
Wahlberg
et al
.,2003b) needs to be more extensively sampled andmore data need to be added before any robustconclusions emerge about relationships of the majorclades.
B
IOGEOGRAPHICAL
HISTORY
OF
N
YMPHALINAE
Despite the limitations of our current phylogenetichypothesis, some strong patterns emerge from our dis-persal-vicariance analysis (Fig. 5). First of all, manydispersal events are required to explain the distribu-tion patterns seen today: 44 when the maximum num-ber of ancestral areas is not constrained and 46 whenthey are constrained to two. This should not be sur-prising, since Nymphalinae contains some of the mostmobile butterflies known, such as
Vanessa cardui
,which is famous for being able to disperse over thou-sands of kilometers in one generation. Indeed, thegenus
Vanessa
, which is present in all the major zoo-logical biomes of the world, has a highly ambiguousreconstruction of its ancestral distribution. Perhapsthe ancestor of
Vanessa
was a widespread species,much like
V. cardui
today, that subsequently speciatedduring intervals of isolation due to climatic or otherfactors.
The distribution of the most recent common ances-tor of Nymphalinae (excluding
Historis
and
Baeotus
)appears to be widespread (Fig. 5), though this may bean artefact of the program used and might be fixed byincluding outgroups with known ancestral distribu-tions (Ronquist, 1996). We refrained from doing so, asour sampling of the outgroups is not sufficient toresolve their ancestral distributions; also, we are notconfident about the current hypothesis of relation-ships among major groups in the nymphaline clade.
The tribe Nymphalini may have originated in SouthAmerica, one of the seven possible ancestral areas inthe unconstrained analysis (Fig. 5) and one of two inthe constrained searches. Africa may have been colo-nized from South America by the common ancestor of
Smyrna
and the rest of Nymphalini. Once the lineageleading to
Smyrna
split off, it appears that the Palae-arctic was colonized from Africa by the common ances-tor of
Antanartia
and the rest of Nymphalini
(excluding
Colobura, Tigridia
and
Smyrna
). An alter-native scenario would have a widespread commonancestor in South America, Africa and the Palaearcticthat then speciated through a series of vicarianceevents (Fig. 5). The Palaearctic has, however, been avery important area for the diversification of the gen-era
Aglais, Nymphalis, Polygonia
and possibly
Van-essa
. There have clearly been several independentcolonizations of the Nearctic from the Palaearctic byspecies in the the former three genera.
The clade that includes the tribes Junoniini and Vic-torinini, as well as
Kallimoides, Rhinopalpa
and
Van-essula
, appears to have originated in Africa (Fig. 5).There is a colonization of South America from Africaby the ancestor of
Kallimoides
and Victorinini. Also,the Oriental region appears to have been colonizedindependently several times. The patterns in thisclade are obscured somewhat by the weak support forthe current hypothesis of relationships; in addition,many missing species in the genera
Hypolimnas
and
Junonia
are likely to have had a decisive effect on theresults due to their presence in the African, Orientaland Australasian areas. However, it is clear that theAfrican region has been instrumental in the diversifi-cation of taxa in the clade.
The tribe Kallimini appears to have originated inthe Oriental region, with a colonization of Africa bythe common ancestor of
Kallima
and
Catacroptera/Mallika
(Fig. 5). This hypothesis needs to be tested byincluding the several species of
Doleschallia
foundexclusively in the Australasian region.
The Nearctic region has been an important area forthe diversification of the tribe Melitaeini (Fig. 5). Oneinterpretation of the patterns recovered by DIVA isthat the group originated in the Nearctic and subse-quently colonized the Palaearctic (the ancestor of
Melitaea
) and the Neotropics (ancestor of the
Gnathot-riche
-group and Phyciodina). Interestingly, the cur-rent phylogenetic hypothesis suggests a disjunctdistribution for the ancestor of
Melitaea
+
Gnathot-riche/Higginsius
+
Phyciodina being found in thePalaearctic and the Neotropics. Sampling the twomissing Caribbean genera will be important to under-standing the historical biogeography of Melitaeini.
In summary, it is clear that several regions havebeen important areas of diversification of clades inNymphalinae, viz. the Palaearctic for the
Nymphalis
-group of genera, Africa for Junoniini and related gen-era and the Nearctic for Melitaeini. Our DIVA analy-sis indicates that dispersal has had a major effect onthe distributions of extant species in Nymphalinae.Whether dispersal or vicariance has been the moreimportant process in generating the deeper diver-gences in Nymphalinae can be tested by dating diver-gences using molecular clocks calibrated with fossils.Such a study is in preparation and the implications of
PHYLOGENY OF NYMPHALINAE 241
© 2005 The Linnean Society of London, Biological Journal of the Linnean Society, 2005, 86, 227–251
Figure 5. Reconstructed ancestral distributions according to a dispersal-vicariance analysis. Letters after the names oftaxa give the current distribution of those taxa.
Colobura/Tigridia BSmyrna B
Antanartia C
Araschnia DMynes F
Symbrenthia E
Hypanartia BVanessa annabella A
Vanessa abyssinica CVanessa atalanta ADVanessa indica DE
Vanessa gonerilla/itea F
Vanessa cardui ACDEVanessa kershawi F
Vanessa virginiensis ABVanessa braziliensis/myrinna B
Aglais io DAglais milberti AAglais urticae D
Polygonia canace DENymphalis l-album AD
Nymphalis polychloros DNymphalis xanthomelas D
Nymphalis californica ANymphalis antiopa AD
Polygonia c-aureum D
Polygonia egea DPolygonia faunus APolygonia c-album D
Polygonia North American clade A
Kallimoides rumia CVictorinini B
Rhinopalpa EVanessula C
Hypolimnas anthedon/usambara CHypolimnas misippus CEF
Hypolimnas alimena FHypolimnas bolina CEFHypolimnas pandarus F
Precis C
Salamis CYoma EFProtogoniomorpha CJunonia erigone FJunonia iphita EJunonia artaxia/touhilimasa/cytora C
Junonia cymodoce CJunonia terea/natalica C
Junonia coenia AJunonia sophia/oenone C
Doleschallia EFCatacroptera/Mallika CKallima E
Euphydryas gilllettii AEuphydryas aurinia/desfontainii D
Euphydryas North American clade A
Poladryas AMicrotia elva ABTexola/Dymasia AChlosyne Central American clade AB
Chlosyne cyneas AChlosyne theona A
Chlosyne nycteis A
Chlosyne gorgone AChlosyne lacinia AB
Chlosyne acastus/palla A
Melitaea DGnathotriche/Higginsius B
Mazia BTegosa B
Phyciodes A
Anthanassa texana AAnthanassa tulcis B
Anthanassa rest BEresia/Castilia/Telenassa B
DFDEEF
DEF
DCD
B
DDF
CDF
ADACDADF
ACDF
ADD
ADD
DD
D
ADD
ADD
D
D
D
DDF
DEF
CDCDF
CDEF
BCBD
BCDBDF
BCDFBDEF
BCDEF
BBD
BCDBDF
BCDFBDEF
BCDEF
BC
CE
FF
FCF
CEF
CCF
C
CEC
EF
ACC
CC
CECF
CEF
C
C
C
C
CEE
ADA
A
AA
AA
AA
A
A
ABB
BAB
BB
BBD
ABAD
ABD
AAD
ABD
AEADE
ABDE
ACACE
ACDEABCDE
ABCEACDE
ABCDEACDEF
ABCDEF
A = NearcticB = NeotropicsC = AfrotropicsD = PalaearcticE = OrientalF = Australasia
AF CF ACF DFADF CDF ACDF EF
AEF CEF ACEF DEFADEF CDEF ACDEF
BD ABD BCD ABCDBDE ABDE BCDE
ABCDE BF ABF BCFABCF BDF ABDF BCDFABCDF BEF ABEF BCEF
ABCEF BDEF ABDEFBCDEF ABCDEF
F DF ADF BDF ABDF CDF ACDFBCDF ABCDF DEF ADEF BDEF ABDEF
CDEF ACDEF BCDEF ABCDEF
BDABDBCD
ABCDBDF
ABDFBCDF
ABCDF
242 N. WAHLBERG ET AL.
© 2005 The Linnean Society of London, Biological Journal of the Linnean Society, 2005, 86, 227–251
it will be discussed elsewhere (N. Wahlberg, unpubl.data).
PBS AND SENSITIVITY ANALYSIS
The ease with which molecular data can be generatedhas led to an ever-increasing number of publishedphylogenetic studies. Common to almost all thesestudies is the lack of sensitivity analyses and, in thecase of multiple independent data sets, evaluation ofcongruence at given nodes (but see Reed & Sperling,1999). An exception to the former are studies based ondirect optimization of ribosomal DNA sequences(Wheeler, 1995; Giribet, 2003). Congruence amongdata partitions is usually evaluated at the level ofentire data sets through tests such as the ILD (Farriset al., 1994), rather than through evaluation at everyresolved node. Again, a few exceptions are notable (e.g.Gatesy et al., 1999; Cognato & Vogler, 2001; Lambkinet al., 2002; Damgaard & Cognato, 2003; Wahlberg &Nylin, 2003). Both the PBS and sensitivity analysisallow detailed evaluation of which nodes are likely tobe robust and stable to the addition of new data, aswell as which nodes require further investigation. Weemphasize that these are heuristic tools to assess thequality of the most parsimonious hypothesis and guidefuture sampling of characters and taxa, rather thanalternate analytical approaches that challenge thephilosophical basis of our preferred optimalitycriterion.
CONCLUSION
The main conclusions of this study are: (1) Smyrna,Colobura and Tigridia are associated with Nympha-lini, (2) ‘Kallimini’ is more closely related to Melitaeinithan to Nymphalini, (3) ‘Kallimini’ is not monophyl-etic, but is paraphyletic with regard to Melitaeini, (4)Melitaeini and Nymphalini plus the three coeine gen-era are strongly supported monophyletic groups, and(5) Precis and Junonia are not synonymous or evensister groups.
Major unanswered questions are: (1) the positions ofHistoris + Baeotus in the nymphaline clade, and (2)the sister group of Nymphalinae. Increased samplingof species in the subfamilies Biblidinae and Apaturi-nae, as well as increased character data (molecularand morphological), will help to resolve thesequestions.
ACKNOWLEDGEMENTS
As always, this study would not have been possiblewithout the truly wonderful, altruistic help of bothamateur and professional lepidopterists. We are eter-nally grateful to Jun-ichi Asahi, Alexei Belik, Anton
Chichvarkhin, Tim Davenport, Phil DeVries, AndréFreitas, Daniel Janzen, Chris Jiggins, Tjeerd Jonge-ling, Darrell Kemp, Norbert Kondla, Runar Krogen,Gerardo Lamas, Torben B. Larsen, Yi-Hsin Lee, JorgeLeón-Cortés, Mike Leski, Debra Murray, NaomiPierce, James Scott, Mike Soukop, Constant’ Stefa-nescu, Martin Steinbauer, Man-Wah Tan, BruceWalsh, Andy Warren, and Keith Willmott for helpingus to obtain specimens. We thank Tiina Berg and Sim-ina Vintila for expert help in the lab. We also thankAndy Warren, Keith Willmott, Carla Penz and TorbenB. Larsen for useful comments on a draft version of themanuscript.
REFERENCES
Ackery PR. 1984. Systematic and faunistic studies on but-terflies. In: Vane-Wright RI, Ackery PR, eds. The biology ofbutterflies. Princeton, NJ: Princeton University Press, 9–21.
Ackery PR. 1988. Hostplants and classification: a review ofnymphalid butterflies. Biological Journal of the LinneanSociety 33: 95–203.
Ackery PR, de Jong R, Vane-Wright RI. 1999. The butter-flies: Hedyloidea, Hesperoidea and Papilionoidea. In: Kris-tensen NP, ed. Lepidoptera, moths and butterflies. 1.Evolution, systematics and biogeography. Handbook of Zool-ogy 4 (35), Lepidoptera. Berlin: de Gruyter, 263–300.
Austin GT, Murphy DD, Baughman JF, Launer AE,Fleishman E. 2003. Hybridization of checkerspot butter-flies in the Great Basin. Journal of the Lepidopterists’ Society57: 176–192.
Baker RH, DeSalle R. 1997. Multiple sources of characterinformation and the phylogeny of Hawaiian drosophilids.Systematic Biology 46: 654–673.
Baker RH, Yu X, DeSalle R. 1998. Assessing the relative con-tribution of molecular and morphological characters insimultaneous analysis trees. Molecular Phylogenetics andEvolution 9: 427–436.
Blum MJ, Bermingham E, Dasmahapatra K. 2003. Amolecular phylogeny of the neotropical butterfly genus Anar-tia (Lepidoptera: Nymphalidae). Molecular Phylogeneticsand Evolution 26: 46–55.
Brakefield PM, Gates J, Keys D, Kesbeke F, WijngaardenPJ, Monteiro A, French V, Carroll SB. 1996. Develop-ment, plasticity and evolution of butterfly eyespot patterns.Nature 384: 236–242.
Bremer K. 1988. The limits of amino acid sequence data inangiosperm phylogenetic reconstruction. Evolution 42: 795–803.
Bremer K. 1994. Branch support and tree stability. Cladistics10: 295–304.
Brower AVZ. 1994. Phylogeny of Heliconius butterfliesinferred from mitochondrial DNA sequences. Molecular Phy-logenetics and Evolution 3: 159–174.
Brower AVZ. 2000. Phylogenetic relationships among theNymphalidae (Lepidoptera), inferred from partial sequences
PHYLOGENY OF NYMPHALINAE 243
© 2005 The Linnean Society of London, Biological Journal of the Linnean Society, 2005, 86, 227–251
of the wingless gene. Proceedings of the Royal Society of Lon-don B 267: 1201–1211.
Brower AVZ, DeSalle R. 1998. Patterns of mitochondrial ver-sus nuclear DNA sequence divergence among nymphalidbutterflies: the utility of wingless as a source of charactersfor phylogenetic inference. Insect Molecular Biology 7: 73–82.
Carroll SB, Gates J, Keys DN, Paddock SW, PanganibanGEF, Selegue JE, Williams JA. 1994. Pattern-formationand eyespot determination in butterfly wings. Science 265:109–114.
Clark AH. 1948. Classification of the butterflies with the allo-cation of the genera occurring in America north of Mexico.Proceedings of the Biological Society of Washington 61: 77–84.
Cognato AI, Vogler AP. 2001. Exploring data interaction andnucleotide alignment in a multiple gene analysis of Ips(Coleoptera: Scolytinae). Systematic Biology 50: 758–780.
Damgaard J, Cognato AI. 2003. Sources of character conflictin a clade of water striders (Heteroptera: Gerridae). Cladis-tics 19: 512–526.
Darlu P, Lecointre G. 2002. When does the IncongruenceLength Difference test fail? Molecular Biology and Evolution19: 432–437.
Dasmahaptra KK, Blum MJ, Aiello A, Hackwell S, DaviesN, Bermingham EP, Mallet J. 2002. Inferences from arapidly moving hybrid zone. Evolution 56: 741–753.
DeSalle R, Brower AVZ. 1997. Process partitions, congru-ence, and the independence of characters: inferring relation-ships among closely related Hawaiian Drosophila frommultiple gene regions. Systematic Biology 46: 751–764.
DeVries PJ, Kitching IJ, Vane-Wright RI. 1985. The sys-tematic position of Antirrhea and Caerois, with comments onthe classification of the Nymphalidae (Lepidoptera). System-atic Entomology 10: 11–32.
Ehrlich PR. 1958. The comparative morphology, phylogenyand higher classification of the butterflies (Lepidoptera:Papilionoidea). University of Kansas Science Bulletin 39:305–370.
Farris JS, Källersjö M, Kluge AG, Bult C. 1994. Testing thesignificance of incongruence. Cladistics 10: 315–319.
Forbes WTM. 1928. Variation in Junonia lavinia (Lepi-doptera: Nymphalidae). Journal of the New York Entomolog-ical Society 36: 305–320.
Fox RM, Forbes AC. 1971. The butterflies of the generaSiproeta and Metamorpha (Lepidoptera: Nymphalidae).Annals of the Carnegie Museum 43: 223–247.
Freitas AVL, Brown KSJ. 2004. Phylogeny of the Nympha-lidae (Lepidoptera: Papilionoidea). Systematic Biology 53:363–383.
Fric Z, Konvicka M, Zrzavyyyy J. 2004. Red & black or black &white? Phylogeny of the Araschnia butterflies (Lepidoptera:Nymphalidae) and evolution of seasonal polyphenism. Jour-nal of Evolutionary Biology 17: 265–278.
Gatesy J, O’Grady P, Baker RH. 1999. Corroborationamong data sets in simultaneous analysis: hidden supportfor phylogenetic relationships among higher level artiodactyltaxa. Cladistics 15: 271–313.
Giribet G. 2003. Stability in phylogenetic formulations and itsrelation to nodal support. Systematic Biology 52: 554–564.
Goloboff PA. 1998. NONA. 2.0. Published by the author.Hall TA. 1999. BioEdit: a user-friendly biological sequence
alignment, editing and analysis program for Windows 95/98/NT. Nucleic Acids Symposium Series 41: 95–98.
Hanski I. 1999. Metapopulation ecology. Oxford: Oxford Uni-versity Press.
Harvey DJ. 1991. Higher classification of the Nymphalidae,Appendix B. In: Nijhout HF, ed. The development and evo-lution of butterfly wing patterns. Washington, DC: Smithso-nian Institution Press, 255–273.
Higgins LG. 1981. A revision of Phyciodes Hübner and relatedgenera, with a review of the classification of the Melitaeinae(Lepidoptera: Nymphalidae). Bulletin of the British Museumof Natural History 43: 77–243.
Howarth TG. 1966. Revisional notes on the genus Antanartia(Lepidoptera: Nymphalidae). Bulletin of the British Museumof Natural History 18: 21–43.
Janz N, Nylin S, Nyblom K. 2001. Evolutionary dynamics ofhost plant specialization: a case study of the tribe Nympha-lini. Evolution 55: 783–796.
de Jong R, Vane-Wright RI, Ackery PR. 1996. The higherclassification of butterflies (Lepidoptera): problems and pros-pects. Entomologica scandinavica 27: 65–101.
Judd DD. 1998. Exploring component stability using life-stage concordance in sabethine mosquitoes (Diptera: Culi-cidae). Cladistics 14: 64–94.
Kons HLJ. 2000. Phylogenetic studies of the Melitaeini (Lepi-doptera: Nymphalidae: Nymphalinae) and a revision of thegenus Chlosyne. Gainesville: Butler Department of Entomol-ogy & Nematology, University of Florida, 799.
Kuznetzov VI, Stekolnikov AA. 2001. New approaches to thesystem of Lepidoptera of world fauna. St. Petersburg: Nauka.
Lambkin CL, Lee MSY, Winterton SL, Yeates DK. 2002.Partitioned Bremer support and multiple trees. Cladistics18: 436–444.
Larsen TB. 1991. The butterflies of Kenya and their naturalhistory. Oxford: Oxford University Press.
de Lesse H. 1952. Note sur les genres Precis Hb. et JunoniaHb. (Lep. Nymphalidae). Bulletin de la Société Ento-mologique de France 57: 74–77.
Miller JS, Brower AVZ, DeSalle R. 1997. Phylogeny of theneotropical moth tribe Josiini (Notodontidae: Dioptinae):comparing and combining evidence from DNA sequences andmorphology. Biological Journal of the Linnean Society 60:297–316.
Monteiro A, Pierce NE. 2001. Phylogeny of Bicyclus (Lepi-doptera: Nymphalidae) inferred from COI, COII, and EF-1alpha gene sequences. Molecular Phylogenetics and Evolu-tion 18: 264–281.
Müller W. 1886. Südamerikanische Nymphalidaeraupen: Ver-such eines natürlichen Systems der Nymphaliden. Zoolo-gische Jahrbücher 1: 417–678.
Muyshondt AJ, Muyshondt A. 1979. Notes on the life cycleand natural history of butterflies of El Salvador. III C. His-toris odius and Coea acheronta (Nymphalidae-Coloburinae).Journal of the Lepidopterists’ Society 33: 112–123.
244 N. WAHLBERG ET AL.
© 2005 The Linnean Society of London, Biological Journal of the Linnean Society, 2005, 86, 227–251
Nakanishi A. 1989. Immature stages of Antanartia abyssin-ica (Felder) (Lepidoptera, Nymphalidae). Japanese Journalof Entomology 57: 712–719.
Nixon KC. 2002. WINCLADA, 1.00.08. Published by theauthor.
Nylin S. 1988. Host plant specialization and seasonality in apolyphagous butterfly, Polygonia c-album (Nymphalidae).Oikos 53: 381–386.
Nylin S, Nyblom K, Ronquist F, Janz N, Belicek J,Källersjö M. 2001. Phylogeny of Polygonia, Nymphalis andrelated butterflies (Lepidoptera: Nymphalidae): a total-evi-dence analysis. Zoological Journal of the Linnean Society132: 441–468.
Parsons M. 1999. The butterflies of Papua New Guinea: Theirsystematics and biology. London: Academic Press.
Penz C, Peggie D. 2003. Phylogenetic relationships amongHeliconiinae genera based on morphology (Lepidoptera:Nymphalidae). Systematic Entomology 28: 451–479.
Reed RD, Sperling FAH. 1999. Interaction of process parti-tions in phylogenetic analysis: an example from the swallow-tail butterfly genus Papilio. Molecular Biology and Evolution16: 286–297.
Ronquist F. 1996. DIVA, 1.1. Program and manual availablefrom Uppsala University (ftp.uu.se or ftp.systbot.uu.se).
Ronquist F. 1997. Dispersal-vicariance analysis: a newapproach to the quantification of historical biogeography.Systematic Biology 46: 195–203.
Rothschild W, Jordan K. 1903. Lepidoptera collected byOscar Neumann in north-east Africa. Novitates Zoologica 10:491–542.
Scott JA. 1984. A review of Polygonia progne (oreas) and P.gracilis (zephyrus) (Nymphalidae), including a new subspe-cies from the southern Rocky Mountains. Journal ofResearch on the Lepidoptera 23: 197–210.
Scott JA. 1985. The phylogeny of butterflies (Papilionoideaand Hesperoidea). Journal of Research on the Lepidoptera23: 241–281.
Scott JA, Wright DM. 1990. Butterfly phylogeny and fossils.In: Kudrna O, ed. Butterflies of Europe. 2. Wiesbaden, 152–208.
Shapiro AM, Geiger H. 1989. Electrophoretic comparisons ofvicariant Vanessa: genetic differentiation between V. anna-bella and V. carye (Nymphalidae) since the Great AmericanInterchange. Journal of the Lepidopterists’ Society 43: 81–92.
Shirôzu T, NakanashiNakanishi A. 1984. A revision of thegenus Kallima Doubleday (Lepidoptera, Nymphalidae). Tyôto Ga 34: 97–110.
Silberglied RE. 1984. Visual communication and sexualselection among butterflies. In: Vane-Wright RI, Ackery PR,eds. The biology of butterflies. Princeton, NJ: Princeton Uni-versity Press, 204–223.
Singer MC. 1971. Evolution of food-plant preference in thebutterfly Euphydryas editha. Evolution 25: 383–389.
Swofford DL. 2001. PAUP*: phylogenetic analysis using par-simony (*and other methods), Version 4.0b10. Sunderland,MA: Sinauer Associates.
Vári L. 1979. The butterflies of southern Africa. Part IV:Nymphalidae; Nymphalinae. Pretoria, South Africa: Trans-vaal Museum.
Wahlberg N. 2001. The phylogenetics and biochemistry ofhost plant specialization in melitaeine butterflies (Lepi-doptera: Nymphalidae). Evolution 55: 522–537.
Wahlberg N, Nylin S. 2003. Morphology versus molecules:resolution of the positions of Nymphalis, Polygonia andrelated genera (Lepidoptera: Nymphalidae). Cladistics 19:213–223.
Wahlberg N, Oliveira R, Scott JA. 2003a. Phylogenetic rela-tionships of Phyciodes butterfly species (Lepidoptera:Nymphalidae): complex mtDNA variation and species delim-itations. Systematic Entomology 28: 257–273.
Wahlberg N, Weingartner E, Nylin S. 2003b. Towards a bet-ter understanding of the higher systematics of Nymphalidae(Lepidoptera: Papilionoidea). Molecular Phylogenetics andEvolution 28: 473–484.
Wahlberg N, Zimmermann M. 2000. Pattern of phylogeneticrelationships among members of the tribe Melitaeini (Lepi-doptera: Nymphalidae) inferred from mtDNA sequences.Cladistics 16: 347–363.
Weller SJ, Pashley DP, Martin JA. 1996. Reassessment ofbutterfly family relationships using independent genes andmorphology. Annals of the Entomological Society of America89: 184–192.
Wheeler WC. 1995. Sequence alignment, parameter sensitiv-ity, and the phylogenetic analysis of molecular data. System-atic Biology 44: 321–331.
Willmott KR, Hall JPW, Lamas G. 2001. Systematics ofHypanartia (Lepidoptera: Nymphalidae: Nymphalinae),with a test for geographical speciation mechanisms in theAndes. Systematic Entomology 26: 369–399.
Zimmermann M, Wahlberg N, Descimon H. 2000. Phy-logeny of Euphydryas checkerspot butterflies (Lepi-doptera: Nymphalidae) based on mitochondrial DNAsequence data. Annals of the Entomological Society ofAmerica 93: 347–355.
PHYLOGENY OF NYMPHALINAE 245
© 2005 The Linnean Society of London, Biological Journal of the Linnean Society, 2005, 86, 227–251
AP
PE
ND
IX 1
Lis
t of
spe
cies
sam
pled
in
th
is s
tudy
alo
ng
wit
h t
he
Gen
Ban
k ac
cess
ion
nu
mbe
rs o
f th
e th
ree
gen
es s
equ
ence
d. F
or i
mag
es o
f vo
uch
er s
peci
men
s, s
ee h
ttp:
//w
ww
.zoo
logi
.su
.se/
rese
arch
/wah
lber
g
Hig
her
tax
onS
peci
esV
ouch
er c
ode
Loc
alit
yC
OI
EF
1-a
win
gles
s
HE
LIC
ON
IIN
AE
Hel
icon
ius
hec
ale
NW
70-6
CO
ST
A R
ICA
: Bu
tter
fly
farm
su
ppli
erA
Y09
0202
AY
0901
68A
Y09
0135
LIM
EN
ITID
INA
EA
del
pha
bred
owi
NW
107-
16U
SA
: Ore
gon
AY
7885
91A
Y78
8693
AY
7884
57C
YR
ES
TIN
AE
Cyr
esti
s th
yod
amas
NW
100-
11B
AN
GL
AD
ES
H: S
ylh
et D
ivis
ion
AY
2182
40A
Y21
8260
AY
2182
78C
her
son
esia
rah
ria
NW
111-
3IN
DO
NE
SIA
: Su
law
esi,
Sek
oA
Y78
8601
AY
7887
03A
Y78
8465
Mar
pesi
a or
silo
chu
sR
B25
0B
RA
ZIL
: Ron
don
ia, A
riqu
emes
AY
7886
04A
Y78
8706
AF
2465
32M
arpe
sia
chir
onR
B22
7B
RA
ZIL
: Ron
don
ia, A
riqu
emes
AY
7886
03A
Y78
8705
AY
7884
67A
mn
osia
dec
ora
NW
101-
1IN
DO
NE
SIA
: W. S
um
atra
, Bra
stag
iA
Y21
8235
AY
2182
54A
Y21
8273
Dic
hor
ragi
a n
esim
ach
us
NW
111-
10P
HIL
IPP
INE
S: S
. Ley
te, H
inu
nan
gan
AY
7886
02A
Y78
8704
AY
7884
66P
seu
der
goli
s w
edah
NW
118-
1V
IET
NA
M: L
ao C
aiA
Y78
8605
AY
7887
07A
Y78
8468
Sti
boch
ion
a n
icea
NW
100-
10B
AN
GL
AD
ES
H: S
ylh
et D
ivis
ion
AY
2182
49A
Y21
8269
AY
2182
87B
IBL
IDIN
AE
Ari
adn
e en
otre
aN
W82
-10
UG
AN
DA
: Kib
ale
For
est
AY
2182
37A
Y21
8256
AY
2182
74B
ybli
a an
vata
raN
W88
-14
ZIM
BA
BW
E: M
aron
dera
AY
7885
95A
Y78
8697
AY
7884
60C
alli
core
pac
ifica
NW
119-
3C
OS
TA
RIC
A: A
C G
uan
acas
te,
03-s
rnp-
1174
4A
Y78
8596
AY
7886
98A
Y78
8461
Cat
onep
hel
e n
um
ilia
NW
62-5
CO
ST
A R
ICA
: Bu
tter
fly
farm
su
ppli
erA
Y09
0215
AY
0901
81A
Y09
0148
Dyn
amin
e m
aeon
RB
249
BR
AZ
IL: R
ondo
nia
, Ari
quem
esA
Y78
8597
AY
7886
99A
F24
6581
Eu
ryte
la d
ryop
eN
W82
-6T
AN
ZA
NIA
: Am
ani
AY
2182
42A
Y21
8262
AY
2182
80H
amad
ryas
feb
rua
NW
62-3
CO
ST
A R
ICA
: Bu
tter
fly
farm
su
ppli
erA
Y09
0216
AY
0901
82A
Y09
0149
Mes
oxan
tha
esot
hea
NW
83-5
UG
AN
DA
: Kib
ale
For
est
AY
7885
98A
Y78
8700
AY
7884
62M
ysce
lia
cape
nsi
sN
W10
9-4
EC
UA
DO
RA
Y78
8599
AY
7887
01A
Y78
8463
Nic
a fl
avil
laN
W85
-11
PE
RU
: Roa
d to
Yu
rim
agu
asA
Y21
8245
AY
2182
65A
Y21
8283
Pan
acea
reg
ina
NW
109-
8E
CU
AD
OR
AY
7886
00A
Y78
8702
AY
7884
64S
even
ia b
oisd
uva
liN
W88
-15
ZIM
BA
BW
E: H
arar
eA
Y21
8247
AY
2182
67A
Y21
8285
AP
AT
UR
INA
EA
patu
ra i
ris
NW
69-6
Bu
tter
fly
farm
su
ppli
erA
Y09
0199
AY
0901
65A
Y09
0132
Ast
eroc
ampa
lei
lia
NW
82-1
5U
SA
: Ari
zon
aA
F18
7734
AY
2182
57A
Y21
8275
Eu
lace
ura
ost
eria
NP
95-Y
227
MA
LA
YS
IA (
MC
Z v
ouch
er)
AY
7885
93A
Y78
8695
AF
2465
88M
imat
hym
a sc
hre
nck
iiN
W84
-7R
US
SIA
: Pri
mor
yeA
Y78
8594
AY
7886
96A
Y78
8459
Tim
elae
a m
acu
lata
NW
97-8
TA
IWA
N: T
aitu
ng
Cou
nty
AY
2182
51A
Y21
8271
AY
2182
89C
hit
oria
ch
ryso
lora
NW
97-1
1T
AIW
AN
: Tai
tun
g C
oun
tyA
Y78
8592
AY
7886
94A
Y78
8458
NY
MP
HA
LIN
AE
‘Coe
ini’
Bae
otu
s ja
petu
sN
W13
0-16
EC
UA
DO
R: P
asta
za, L
oroc
ach
iA
Y78
8613
AY
7887
18A
Y78
8479
Bae
otu
s d
euca
lion
NW
130-
13E
CU
AD
OR
: Nap
o, R
io Y
utu
riA
Y78
8616
AY
7887
21A
Y78
8482
Bae
otu
s ae
ilu
sN
W13
0-14
EC
UA
DO
R: S
ucu
mbi
os, C
erro
Lu
mba
qui
AY
7886
14A
Y78
8719
AY
7884
80B
aeot
us
beot
us
NW
130-
15E
CU
AD
OR
: Im
babu
ra, R
io C
ach
aco
AY
7886
15A
Y78
8720
AY
7884
81H
isto
ris
odiu
sN
W81
-7C
OS
TA
RIC
A: B
utt
erfl
y fa
rm s
upp
lier
AY
7886
32A
Y78
8751
AY
7885
12H
isto
ris
ach
eron
taP
E-1
8–16
PE
RU
: Mad
re d
e D
ios,
Tam
bopa
ta P
res.
AY
7886
31A
Y78
8750
AY
7885
11C
olob
ura
dir
ceN
W68
-11
CO
ST
A R
ICA
: Bu
tter
fly
farm
su
ppli
erA
Y09
0228
AY
0901
96A
Y09
0162
246 N. WAHLBERG ET AL.
© 2005 The Linnean Society of London, Biological Journal of the Linnean Society, 2005, 86, 227–251
Tig
rid
ia a
cest
aR
B34
9B
RA
ZIL
: Ron
don
ia, A
riqu
emes
AY
7886
84A
Y78
8822
AY
7885
82S
myr
na
blom
fild
iaN
W85
-2B
RA
ZIL
: São
Pau
loA
Y78
8678
AY
7888
16A
Y78
8576
Nym
phal
ini
An
tan
arti
a ab
yssi
nic
aN
W86
-7T
AN
ZA
NIA
: Kit
um
bein
eA
Y78
8609
AY
7887
11A
Y78
8472
An
tan
arti
a d
eliu
sN
W82
-4U
GA
ND
A: K
ibal
e F
ores
tA
Y78
8610
AY
7887
12A
Y78
8473
An
tan
arti
a sc
hae
nia
NW
65-5
CA
ME
RO
ON
AY
2182
36A
Y21
8255
AF
4127
80A
rasc
hn
ia l
evan
aN
W39
-2E
ST
ON
IAA
Y24
8780
AY
2488
05A
F41
2762
Myn
es g
eoff
royi
NW
63-2
0A
US
TR
AL
IA: Q
uee
nsl
and
AY
2487
78A
Y24
8803
AF
4127
60S
ymbr
enth
ia h
ypat
iaN
W65
-2M
AL
AY
SIA
: Sel
ango
rA
Y24
8779
AY
2488
04A
F41
2784
Sym
bren
thia
hyp
seli
sN
W97
-2T
AIW
AN
: Kau
hsi
un
g C
oun
tyA
Y78
8680
AY
7888
18A
Y78
8578
Sym
bren
thia
lil
eaN
W97
-3T
AIW
AN
: Kau
hsi
un
g C
oun
tyA
Y78
8679
AY
7888
17A
Y78
8577
Hyp
anar
tia
bell
aP
E-1
0–7
PE
RU
: Cu
zco,
Qu
ebra
da C
hap
imay
oA
Y78
8638
AY
7887
57A
F24
6590
Hyp
anar
tia
char
onN
W89
-11
EC
UA
DO
R: S
ucu
mbi
osA
Y78
8639
AY
7887
58A
Y78
8518
Hyp
anar
tia
kefe
rste
inii
NW
89-6
EC
UA
DO
R: S
ucu
mbi
osA
Y78
8640
AY
7887
59A
Y78
8519
Hyp
anar
tia
leth
eN
W36
-6B
RA
ZIL
: Min
as G
erai
sA
F18
7774
AY
7887
60A
Y78
8520
Hyp
anar
tia
lin
dig
iiP
E-0
5 A
-15
PE
RU
: Cu
zco,
Qu
ebra
da S
an L
uis
AY
2487
81A
Y24
8806
AF
4127
59V
anes
sa a
nn
abel
laN
W74
-4U
SA
: Wyo
min
gA
Y78
8685
AY
7888
23A
Y78
8583
Van
essa
ata
lan
taN
W63
-21
SW
ED
EN
: Sto
ckh
olm
AY
0902
21A
Y09
0187
AF
4127
72V
anes
sa b
razi
lien
sis
NW
89-4
EC
UA
DO
R: E
smer
alda
sA
Y78
8686
AY
7888
24A
Y78
8584
Van
essa
car
du
iN
W63
-3U
SA
: Mis
sou
riA
Y24
8782
AY
2488
07A
F41
2770
Van
essa
gon
eril
laN
W63
-4N
EW
ZE
AL
AN
DA
Y24
8784
AY
2488
09A
F41
2782
Van
essa
in
dic
aN
W63
-9JA
PA
NA
Y78
8687
AY
7888
25A
Y78
8585
Van
essa
ite
aN
W63
-14
NE
W Z
EA
LA
ND
AY
7886
88A
Y78
8826
AY
7885
86V
anes
sa k
ersh
awi
NW
77-3
AU
ST
RA
LIA
: Sou
th A
ust
rali
aA
Y78
8689
AY
7888
27A
Y78
8587
Van
essa
myr
inn
aN
W36
-5B
RA
ZIL
: Min
as G
erai
sA
Y78
8690
AY
7888
28A
Y78
8588
Van
essa
vir
gin
ien
sis
NW
77-1
6U
SA
: Ten
nes
see
AY
2487
83A
Y24
8808
AY
2488
27A
glai
s io
NW
63-1
6S
WE
DE
N: S
tock
hol
mA
Y24
8785
AY
2488
10A
F41
2766
Agl
ais
mil
bert
iN
W77
-14
US
A: W
ash
ingt
onA
Y24
8787
AY
2488
12A
Y24
8828
Agl
ais
urt
icae
NW
63-3
SW
ED
EN
: Sto
ckh
olm
AY
2487
86A
Y24
8811
AF
4127
77N
ymph
alis
an
tiop
aN
W70
-2S
WE
DE
N: S
tock
hol
mA
Y21
8246
AY
2182
66A
Y21
8284
Nym
phal
is c
alif
orn
ica
NW
74-1
4U
SA
: Ore
gon
AY
2487
89A
Y24
8814
AY
2488
30N
ymph
alis
l-a
lbu
mN
W78
-1C
AN
AD
A: B
riti
sh C
olu
mbi
aA
Y24
8791
AY
2488
16A
Y24
8832
Nym
phal
is p
olyc
hlo
ros
NW
62-2
SW
ED
EN
: Öla
nd
AY
2487
88A
Y24
8813
AY
2488
29N
ymph
alis
xan
thom
elas
NW
84-1
RU
SS
IA: Y
aku
tia
AY
2487
90A
Y24
8815
AY
2488
31P
olyg
onia
can
ace
EW
19-1
1JA
PA
NA
Y24
8792
AY
2488
17A
Y24
8833
Pol
ygon
ia c
-alb
um
NW
70-3
SW
ED
EN
: Sto
ckh
olm
AY
0902
22A
Y09
0188
AY
0901
54P
olyg
onia
c-a
ure
um
NW
65-8
JAP
AN
AY
2487
99A
Y24
8824
AF
4127
86P
olyg
onia
com
ma
NW
65-6
US
A: T
enn
esse
eA
Y24
8794
AY
2488
19A
F41
2781
Pol
ygon
ia e
gea
NW
77-1
5G
RE
EC
EA
Y24
8800
AY
2488
25A
Y24
8838
Pol
ygon
ia f
aun
us
NW
74-1
2U
SA
: Ore
gon
AY
2487
98A
Y24
8823
AY
2488
37
Hig
her
tax
onS
peci
esV
ouch
er c
ode
Loc
alit
yC
OI
EF
1-a
win
gles
s
AP
PE
ND
IX 1
Con
tin
ued
PHYLOGENY OF NYMPHALINAE 247
© 2005 The Linnean Society of London, Biological Journal of the Linnean Society, 2005, 86, 227–251
Pol
ygon
ia h
arol
di
NW
112-
3M
EX
ICO
: Son
ora,
Yec
ora
AY
7886
62A
Y78
8800
AY
7885
60P
olyg
onia
in
terr
ogat
ion
isN
W77
-12
US
A: T
enn
esse
eA
Y24
8793
AY
2488
18A
Y24
8834
Pol
ygon
ia o
reas
NW
74-1
0U
SA
: Ore
gon
AY
7886
63A
Y78
8801
AY
7885
61P
olyg
onia
sat
yru
sN
W74
-9U
SA
: Ore
gon
AY
2487
96A
Y24
8821
AY
2488
35P
olyg
onia
gra
cili
sN
W74
-6U
SA
: Ore
gon
AY
2487
97A
Y24
8822
AY
2488
36‘K
alli
min
i’A
nar
tia
amat
hea
NW
68-5
CO
ST
A R
ICA
: Bu
tter
fly
farm
su
ppli
erA
Y78
8606
AY
7887
08A
Y78
8469
An
arti
a fa
tim
aN
W66
-5C
OS
TA
RIC
A: B
utt
erfl
y fa
rm s
upp
lier
AY
7886
07A
Y78
8709
AY
7884
70A
nar
tia
jatr
oph
aeN
W36
-6B
RA
ZIL
: Min
as G
erai
sA
Y78
8608
AY
7887
10A
Y78
8471
Sip
roet
a ep
aph
us
NW
64-7
CO
ST
A R
ICA
: Bu
tter
fly
farm
su
ppli
erA
Y78
8677
AY
7888
15A
Y78
8575
Sip
roet
a st
elen
esN
W69
-5C
OS
TA
RIC
A: B
utt
erfl
y fa
rm s
upp
lier
AY
2182
48A
Y21
8268
AY
2182
86N
apeo
cles
ju
cun
da
NW
85-1
BR
AZ
IL: M
ato
Gro
sso
AY
7886
61A
Y78
8788
AY
7885
48M
etam
orph
a el
issa
PE
-10–
10P
ER
U: C
uzc
o, Q
ueb
rada
Ch
aupi
may
oA
Y78
8658
AY
7887
84A
Y78
8544
Hyp
olim
nas
ali
men
aN
W81
-1A
US
TR
AL
IA: Q
uee
nsl
and
AY
7886
33A
Y78
8752
AY
7885
13H
ypol
imn
as a
nth
edon
NW
68-8
KE
NY
A: B
utt
erfl
y fa
rm s
upp
lier
AY
7886
34A
Y78
8753
AY
7885
14H
ypol
imn
as b
olin
aN
W62
-6A
US
TR
AL
IA: Q
uee
nsl
and
AY
0902
24A
Y09
0190
AY
0901
56H
ypol
imn
as m
isip
pus
NW
68-3
KE
NY
A: B
utt
erfl
y fa
rm s
upp
lier
AY
7886
35A
Y78
8754
AY
7885
15H
ypol
imn
as p
and
aru
sN
W80
-11
IND
ON
ES
IA: C
eram
Isl
and
AY
7886
36A
Y78
8755
AY
7885
16H
ypol
imn
as u
sam
bara
NW
66-4
KE
NY
A: B
utt
erfl
y fa
rm s
upp
lier
AY
7886
37A
Y78
8756
AY
7885
17P
reci
s an
dre
mia
jaN
W11
1-6
MA
DA
GA
SC
AR
: Man
drak
aA
Y78
8664
AY
7888
02A
Y78
8562
Pre
cis
anti
lope
NW
88-4
ZIM
BA
BW
E: M
aron
dera
AY
7886
65A
Y78
8803
AY
7885
63P
reci
s ar
ches
iaN
W88
-6Z
IMB
AB
WE
: Mar
onde
raA
Y78
8666
AY
7888
04A
Y78
8564
Pre
cis
cery
ne
NW
88-3
ZIM
BA
BW
E: M
aron
dera
AY
7886
67A
Y78
8805
AY
7885
65P
reci
s cu
ama
NW
83-1
3Z
IMB
AB
WE
: Har
are
AY
7886
68A
Y78
8806
AY
7885
66P
reci
s oc
tavi
aN
W68
-9K
EN
YA
: Bu
tter
fly
farm
su
ppli
erA
Y78
8669
AY
7888
07A
Y78
8567
Pre
cis
sin
uat
aN
W83
-1U
GA
ND
A: K
ibal
e F
ores
tA
Y78
8670
AY
7888
08A
Y78
8568
Pre
cis
tuge
laN
W11
4-15
ZA
MB
IA: N
of
Mw
inil
un
ga, L
esom
bo R
AY
7886
71A
Y78
8809
AY
7885
69R
hin
opal
pa p
olyn
ice
NW
81-5
MA
LA
YS
IAA
Y78
8674
AY
7888
12A
Y78
8572
Sal
amis
an
teva
NW
111-
9M
AD
AG
AS
CA
R: M
andr
aka
AY
7886
75A
Y78
8813
AY
7885
73S
alam
is c
acta
NW
82-3
UG
AN
DA
: Kib
ale
For
est
AY
7886
76A
Y78
8814
AY
7885
74Yo
ma
algi
na
NW
80-1
3P
NG
: Mor
obe
Pro
v., W
au V
alle
yA
Y78
8692
AY
7888
30A
Y78
8590
Pro
togo
nio
mor
pha
anac
ard
iiN
W73
-15
KE
NY
A: B
utt
erfl
y fa
rm s
upp
lier
AY
0902
23A
Y09
0189
AY
0901
55P
r. pa
rhas
sus
NW
82-7
UG
AN
DA
: Kib
ale
For
est
AY
7886
73A
Y78
8811
AY
7885
71P
r. cy
tora
NW
123-
23G
HA
NA
AY
7886
72A
Y78
8810
AY
7885
70Ju
non
ia a
rtax
iaN
W11
4-11
ZA
MB
IA: S
of
Mw
inil
un
gaA
Y78
8642
AY
7887
62A
Y78
8522
Jun
onia
coe
nia
NW
85-1
3U
SA
: Ten
nes
see
AY
7886
43A
Y24
8801
AY
2488
26Ju
non
ia i
phit
aN
W68
-17
Bu
tter
fly
farm
su
ppli
erA
Y09
0225
AY
0901
91A
Y09
0157
Jun
onia
eri
gon
eN
W81
-2P
NG
: Mor
obe
Pro
vin
ce, B
ulo
loA
Y78
8644
AY
7887
63A
Y78
8523
Jun
onia
nat
alic
aN
W68
-13
KE
NY
A: B
utt
erfl
y fa
rm s
upp
lier
AY
7886
45A
Y78
8764
AY
7885
24Ju
non
ia o
enon
eN
W68
-1K
EN
YA
: Bu
tter
fly
farm
su
ppli
erA
Y78
8646
AY
7887
65A
Y78
8525
Jun
onia
sop
hia
NW
83-1
0U
GA
ND
A: K
ibal
e F
ores
tA
Y78
8647
AY
7887
66A
Y78
8526
Jun
onia
ter
eaN
W68
-15
KE
NY
A: B
utt
erfl
y fa
rm s
upp
lier
AY
7886
48A
Y78
8767
AY
7885
27Ju
non
ia t
ouh
ilim
asa
NW
95-1
5Z
AM
BIA
: Kal
ene
Hil
lA
Y78
8649
AY
7887
68A
Y78
8528
Hig
her
tax
onS
peci
esV
ouch
er c
ode
Loc
alit
yC
OI
EF
1-a
win
gles
s
248 N. WAHLBERG ET AL.
© 2005 The Linnean Society of London, Biological Journal of the Linnean Society, 2005, 86, 227–251
Kam
illa
cym
odoc
eN
W11
4-8
ZA
MB
IA: N
of
Mw
inil
un
ga, L
esom
bo R
AY
7886
52A
Y78
8771
AY
7885
31K
alli
ma
inac
hu
sN
W85
-15
Bu
tter
fly
farm
su
ppli
erA
Y78
8650
AY
7887
69A
Y78
8529
Kal
lim
a pa
rale
kta
NW
62-8
Bu
tter
fly
farm
su
ppli
erA
Y09
0229
AY
0901
97A
Y09
0163
Kal
lim
oid
es r
um
iaN
W96
-8G
HA
NA
: Ash
anti
Reg
ion
AY
7886
51A
Y78
8770
AY
7885
30C
atac
ropt
era
cloa
nth
eN
W88
-1Z
IMB
AB
WE
: Mar
onde
raA
Y78
8619
AY
7887
24A
Y78
8485
Mal
lika
jac
kson
iN
W12
2-6
TA
NZ
AN
IAA
Y78
8653
AY
7887
72A
Y78
8532
Dol
esch
alli
a bi
salt
ide
NW
64-5
Bu
tter
fly
farm
su
ppli
erA
Y78
8621
AY
7887
35A
Y78
8496
Van
essu
la m
ilca
NW
96-5
GH
AN
A: A
shan
ti R
egio
nA
Y78
8691
AY
7888
29A
Y78
8589
Mel
itae
ini
Eu
phyd
ryas
au
rin
iaN
W6-
4F
RA
NC
E: C
ervi
ères
AF
1877
46A
Y78
8743
AY
7885
04E
uph
ydry
as c
hal
ced
ona
NW
14-4
US
A: C
alif
orn
iaA
F18
7752
AY
7887
44A
Y78
8505
Eu
phyd
ryas
des
fon
tain
iiN
W70
-4S
PA
IN: E
l G
uix
AY
0902
26A
Y09
0193
AY
0901
59E
uph
ydry
as e
dit
ha
NW
5-8
US
A: C
alif
orn
iaA
F18
7765
AY
7887
45A
Y78
8506
Eu
phyd
ryas
gil
lett
iiN
W24
-6U
SA
: Mon
tan
aA
F18
7771
AY
7887
46A
Y78
8507
Eu
phyd
ryas
ph
aeto
nN
W13
-3U
SA
: Mar
ylan
dA
F18
7797
AY
7887
47A
Y78
8508
Ch
losy
ne
acas
tus
NW
35-1
5U
SA
: Col
orad
oA
F18
7735
AY
7887
25A
Y78
8486
Ch
losy
ne
cyn
eas
NW
38-1
7U
SA
: Ari
zon
aA
F18
7757
AY
7887
26A
Y78
8487
Ch
losy
ne
gau
dea
lis
NW
37-2
CO
ST
A R
ICA
: La
Sel
vaA
F18
7770
AY
7887
27A
Y78
8488
Ch
losy
ne
gorg
one
NW
34-4
US
A: C
olor
ado
AF
1877
72A
Y78
8728
AY
7884
89C
hlo
syn
e h
arri
sii
NW
35-1
0U
SA
: New
Yor
kA
F18
7773
AY
7887
29A
Y78
8490
Ch
losy
ne
jan
ais
NW
62-1
CO
ST
A R
ICA
: Bu
tter
fly
farm
su
ppli
erA
Y78
8620
AY
7887
30A
Y78
8491
Ch
losy
ne
laci
nia
NW
62-4
CO
ST
A R
ICA
: Bu
tter
fly
farm
su
ppli
erA
Y09
0227
AY
0901
95A
Y09
0161
Ch
losy
ne
nar
vaN
W37
-3C
OS
TA
RIC
A: L
a S
elva
AF
1877
86A
Y78
8731
AY
7884
92C
hlo
syn
e n
ycte
isN
W34
-5U
SA
: Col
orad
oA
F18
7788
AY
7887
32A
Y78
8493
Ch
losy
ne
pall
aN
W20
-4U
SA
: Cal
ifor
nia
AF
1877
91A
Y78
8733
AY
7884
94C
hlo
syn
e th
eon
aN
W27
-6U
SA
: Ari
zon
aA
F18
7808
AY
7887
34A
Y78
8495
Dym
asia
dym
asN
W27
-7U
SA
: Ari
zon
aA
F18
7764
AY
7887
85A
Y78
8545
Mic
roti
a el
vaN
W61
-1M
EX
ICO
: Ch
iapa
sA
Y78
8660
AY
7887
87A
Y78
8547
Tex
ola
elad
aN
W7-
1U
SA
: Tex
asA
Y78
8659
AY
7887
86A
Y78
8546
Pol
adry
as a
rach
ne
NW
27-4
US
A: C
alif
orn
iaA
F18
7740
AY
7887
99A
Y78
8559
Mel
itae
a ar
du
inn
aN
W23
-5G
RE
EC
E: P
isso
deri
AF
1877
42A
Y78
8774
AY
7885
34M
elit
aea
brit
omar
tis
NW
69-8
SW
ED
EN
AY
7886
55A
Y78
8775
AY
7885
35M
elit
aea
cin
xia
NW
73-1
4S
WE
DE
N: S
tock
hol
mA
Y78
8656
AY
7887
76A
Y78
8536
Mel
itae
a d
eion
eN
W95
-5F
RA
NC
E: A
ude
AY
7886
57A
Y78
8777
AY
7885
37M
elit
aea
did
ymoi
des
NW
26-1
RU
SS
IA: B
ury
atia
AF
1877
62A
Y09
0194
AY
0901
60M
elit
aea
lato
nig
ena
NW
25-3
RU
SS
IA: B
ury
atia
AF
1877
80A
Y78
8778
AY
7885
38
Hig
her
tax
onS
peci
esV
ouch
er c
ode
Loc
alit
yC
OI
EF
1-a
win
gles
s
AP
PE
ND
IX 1
Con
tin
ued
PHYLOGENY OF NYMPHALINAE 249
© 2005 The Linnean Society of London, Biological Journal of the Linnean Society, 2005, 86, 227–251
Mel
itae
a pe
rsea
NW
34-1
0L
EB
AN
ON
: Moh
afaz
at B
ehar
réA
F18
7796
AY
7887
79A
Y78
8539
Mel
itae
a pu
nic
aN
W34
-11
LE
BA
NO
N: M
ohaf
azat
Kes
ron
anA
F18
7803
AY
7887
81A
Y78
8541
Mel
itae
a sc
otos
iaN
W27
-11
CH
INA
: Heb
ei P
rovi
nce
AF
1878
04A
Y78
8780
AY
7885
40M
elit
aea
triv
iaN
W23
-6G
RE
EC
E: P
isso
deri
AF
1878
10A
Y78
8782
AY
7885
42M
elit
aea
vari
aN
W24
-13
FR
AN
CE
: Lau
s de
Cer
vièr
esA
F18
7812
AY
7887
83A
Y78
8543
Gn
ath
otri
che
excl
amat
ion
isN
W89
-9E
CU
AD
OR
: Su
cum
bios
AY
7886
29A
Y78
8748
AY
7885
09H
iggi
nsi
us
fasc
iatu
sP
E-1
0–20
PE
RU
: Cu
zco,
Qu
ebra
da C
hau
pim
ayo
AY
7886
30A
Y78
8749
AY
7885
10A
nth
anas
sa d
rusi
lla
NW
76-6
EC
UA
DO
R: E
smer
alda
sA
Y78
8611
AY
7887
14A
Y78
8475
An
than
assa
tex
ana
NW
12-6
US
A: T
exas
AF
1878
06A
Y78
8716
AY
7884
77A
nth
anas
sa a
rdys
NW
22-4
CO
ST
A R
ICA
: Mon
teve
rde
AF
1877
43A
Y78
8713
AY
7884
74A
nth
anas
sa o
tan
esN
W24
-4C
OS
TA
RIC
A: M
onte
verd
eA
F18
7790
AY
7887
15A
Y78
8476
An
than
assa
tu
lcis
NW
104-
12P
AN
AM
A: G
ambo
aA
Y78
8612
AY
7887
17A
Y78
8478
Cas
tili
a er
anit
esN
W76
-2E
CU
AD
OR
: Pic
hin
cha
AY
7886
17A
Y78
8722
AY
7884
83C
asti
lia
ofel
laN
W10
5-3
PA
NA
MA
: Ach
iote
Roa
dA
Y78
8618
AY
7887
23A
Y78
8484
Ere
sia
clio
NW
76-5
EC
UA
DO
R: E
smer
alda
sA
Y78
8622
AY
7887
36A
Y78
8497
Ere
sia
coel
aN
W10
4-3
PA
NA
MA
: Pat
h t
o G
lori
a A
lta
AY
7886
23A
Y78
8737
AY
7884
98E
resi
a eu
nic
eN
W92
-5B
RA
ZIL
: São
Pau
loA
Y78
8624
AY
7887
38A
Y78
8499
Ere
sia
leti
tia
NW
91-9
EC
UA
DO
R: S
ucu
mbi
osA
Y78
8625
AY
7887
39A
Y78
8500
Ere
sia
quin
till
aN
W76
-3E
CU
AD
OR
: Esm
eral
das
AY
7886
27A
Y78
8741
AY
7885
02E
resi
a pe
lon
iaN
W10
8-11
PE
RU
AY
7886
26A
Y78
8740
AY
7885
01E
resi
a se
stia
NW
76-8
EC
UA
DO
R: E
smer
alda
sA
Y78
8628
AY
7887
42A
Y78
8503
Jan
atel
la l
euco
des
ma
NW
85-1
6P
AN
AM
A: G
ambo
aA
Y78
8641
AY
7887
61A
Y78
8521
Maz
ia a
maz
onic
aN
W76
-6E
CU
AD
OR
AY
7886
54A
Y78
8773
AY
7885
33P
hyc
iod
es b
ates
iiN
W72
-4C
AN
AD
A: O
nta
rio
AF
1877
47A
Y78
8789
AY
7885
49P
hyc
iod
es c
ocyt
aN
W11
-4C
AN
AD
A: B
riti
sh C
olu
mbi
aA
F18
7755
AY
0901
92A
Y09
0158
Ph
ycio
des
myl
itta
NW
11-1
0C
AN
AD
A: B
riti
sh C
olu
mbi
aA
F18
7785
AY
7887
91A
Y78
8551
Ph
ycio
des
ors
eis
NW
67-3
US
A: C
alif
orn
iaA
Y15
6631
AY
7887
92A
Y78
8552
Ph
ycio
des
pal
lesc
ens
NW
64-2
ME
XIC
O: M
ich
oacá
nA
Y15
6640
AY
7887
93A
Y78
8553
Ph
ycio
des
pal
lid
aN
W34
-6U
SA
: Col
orad
oA
F18
7792
AY
7887
94A
Y78
8554
Ph
ycio
des
ph
aon
NW
35-1
1M
EX
ICO
: Maz
atla
nA
F18
7798
AY
7887
95A
Y78
8555
Ph
ycio
des
pic
taN
W34
-7U
SA
: Col
orad
oA
F18
7800
AY
7887
96A
Y78
8556
Ph
ycio
des
pu
lch
ella
NW
67-1
4U
SA
: Ore
gon
AY
1566
62A
Y78
8797
AY
7885
57P
hyc
iod
es t
har
osN
W34
-2U
SA
: Min
nes
ota
AF
1878
07A
Y78
8798
AY
7885
58P
hyc
iod
es g
raph
ica
NW
67-9
ME
XIC
O: M
exic
o S
tate
AY
1566
84A
Y78
8790
AY
7885
50T
elen
assa
tri
mac
ula
taN
W91
-6E
CU
AD
OR
: Su
cum
bios
AY
7886
83A
Y78
8821
AY
7885
81T
egos
a an
ieta
NW
91-1
1E
CU
AD
OR
: Su
cum
bios
AY
7886
81A
Y78
8819
AY
7885
79T
egos
a ti
ssoi
des
NW
76-4
EC
UA
DO
R: E
smer
alda
sA
Y78
8682
AY
7888
20A
Y78
8580
Hig
her
tax
onS
peci
esV
ouch
er c
ode
Loc
alit
yC
OI
EF
1-a
win
gles
s
250 N. WAHLBERG ET AL.
© 2005 The Linnean Society of London, Biological Journal of the Linnean Society, 2005, 86, 227–251
APPENDIX 2
Proposed higher classification of the subfamilyNymphalinae based on results presented in this paper.Genera marked with asterisks were not sampled forthis study. For a full synonymic list of the speciesplease see http://www.zoologi.su.se/research/wahlberg
Subfamily NYMPHALINAE Rafinesque, 1815incertae sedis
Pycina* Doubleday, 1849Kallimoides Shirôzu & Nakanishi, 1984Vanessula Dewitz, 1887Rhinopalpa Felder & Felder, 1860
Tribe Coeini Scudder, 1893Historis Hübner, 1819
= Coea Hübner, 1819= Aganisthos Boisduval & Le Conte, 1835= Megistanis Doubleday, 1845= Megistanis Boisduval, 1870
Baeotus Hemming, 1939Tribe Nymphalini Rafinesque, 1815
Colobura Billberg, 1820= Gynoecia Doubleday, 1845
Tigridia Hübner, 1819= Callizona Doubleday, 1848
Smyrna Hübner, 1823Antanartia Rothschild & Jordan, 1903Araschnia Hübner, 1819Symbrenthia Hübner, 1819
= Laogona Boisduval, 1836= Brensymthia Huang 2001
Mynes Boisduval, 1832Hypanartia Hübner, 1821
= Eurema Doubleday, 1845Vanessa Fabricius, 1807
= Nymphalis Latreille, 1804= Cynthia Fabricius, 1807= Pyrameis Hübner, 1819= Bassaris Hübner, 1821= Ammiralis Rennie, 1832= Neopyrameis Scudder, 1889= Fieldia Niculescu, 1979
Aglais Dalman, 1816= Ichnusa Reuss, 1939= Inachis Hübner, 1819= Hamadryas Hübner, 1806
Nymphalis Kluk, 1780= Scudderia Grote, 1873= Euvanessa Scudder, 1889= Roddia Korshunov, 1996
Polygonia Hübner, 1819= Kaniska Moore, 1899= Eugonia Hübner, 1819= Comma Rennie, 1832= Grapta Kirby, 1837
Tribe Victorinini Scudder, 1893Anartia Hübner, 1819
= Celaena Doubleday, 1849= Celoena Boisduval, 1870= Anartiella Fruhstorfer, 1907
Siproeta Hübner, 1823= Victorina Blanchard, 1840= Amphirene Doubleday, 1845= Amphirene Boisduval, 1870= Aphnaea Capronnier, 1881
Napeocles Bates, 1864Metamorpha Hübner, 1819
Tribe Junoniini Reuter, 1896Junonia Hübner, 1819
= Alcyoneis Hübner, 1819= Aresta Billberg, 1820= Kamilla Collins & Larsen, 1991
Salamis Boisduval, 1833Yoma Doherty, 1886
= Yoma de Nicéville, 1886Protogoniomorpha Wallengren,
1857Precis Hübner, 1819
= Coryphaeola Butler, 1878= Kallimula Holland, 1920
Hypolimnas Hübner, 1819= Esoptria Hübner, 1819= Diadema Boisduval, 1832= Euralia Westwood, 1850= Eucalia Felder, 1861
Tribe Kallimini Doherty, 1886Kallima Doubleday, 1849Doleschallia Felder & Felder, 1860
= Apatura Hübner, 1819Catacroptera Karsch, 1894Mallika Collins & Larsen, 1991
Tribe Melitaeini Newman, 1870incertae sedis
Antillea* Higgins, 1959Atlantea* Higgins, 1959Gnathotriche Felder & Felder, 1862
= Gnathotrusia Higgins, 1981Higginsius Hemming, 1964
= Fulvia Higgins, 1959Chlosyne Butler, 1870
= Morpheis Geyer, 1833= Synchloe Doubleday, 1845= Anemeca Kirby, 1871= Coatlantona Kirby, 1871= Limnaecia Scudder, 1872= Charidryas Scudder, 1872= Thessalia Scudder, 1875
Microtia Bates, 1864Texola Higgins, 1959Dymasia Higgins, 1960Poladryas Bauer, 1961
PHYLOGENY OF NYMPHALINAE 251
© 2005 The Linnean Society of London, Biological Journal of the Linnean Society, 2005, 86, 227–251
Subtribe Euphydryina Higgins, 1978Euphydryas Scudder, 1872
= Lemonias Hübner, 1806= Occidryas Higgins, 1978= Eurodryas Higgins, 1978= Hypodryas Higgins, 1978
Subtribe Melitaeina Newman, 1870Melitaea Fabricius, 1807
= Lucina Rafinesque, 1815= Schoenis Hübner, 1819= Cinclidia Hübner, 1819
= Mellicta Billberg, 1820= Didymaeformia Verity, 1950
= Athaliaeformia Verity, 1950
Subtribe Phyciodina Higgins, 1981Phyciodes Hübner, 1819Phystis* Higgins, 1981Anthanassa Scudder, 1875
= Tritanassa Forbes, 1945Dagon* Higgins, 1981Telenassa Higgins, 1981Ortilia* Higgins, 1981Tisona* Higgins, 1981Tegosa Higgins, 1981Eresia Boisduval, 1836Castilia Higgins, 1981Janatella Higgins, 1981Mazia Higgins, 1981