s p ecies d iversiw ca tion pa ttern s in the p o lyn esian ju mpi n g...

Molecular Phylogenetics and Evolution 41 (2006) 472–495www.elsevier.com/locate/ympev

1055-7903/$ - see front matter 2006 Elsevier Inc. All rights reserved.doi:10.1016/j.ympev.2006.05.012

Species diversiWcation patterns in the Polynesian jumping spider genus Havaika Prószyjski, 2001 (Araneae, Salticidae)

Miquel A. Arnedo ¤, Rosemary G. Gillespie 1

Division of Insect Biology, University of California-Berkeley, ESPM 201 Wellman Hall, Berkeley, CA 94720-3112, USA

Received 30 March 2006; revised 10 May 2006; accepted 13 May 2006Available online 20 May 2006

Abstract

Hotspot archipelagoes provide exceptional models for the study of the evolutionary process, due to the eVects of isolation and topograph-ical diversity in inducing the formation of unique biotic assemblages. In this paper, we examine the evolutionary patterns exhibited by thejumping spider genus Havaika Prószyjski, 2001 in the Polynesian islands of the Hawaiian and Marquesas chains. To date, systematicresearch on Havaika has been seriously limited by the poor taxonomic knowledge on the group, which was based on a handful of specimensthat showed continuous variability and lacked clear-cut diagnostic characters. Here, we circumvent this problem by inferring a phylogenybased on DNA sequences of several fragments including both mitochondrial (protein coding cytochrome oxidase I, NAD1 dehydrogenase,ribosomal 16S, and tRNA leu) and nuclear (internal transcribed spacer 2) genes, and a statistical morphological analyses of a large sample ofspecimens. Results suggest that the Marquesan and Hawaiian Havaika may be the result of independent colonizations. Furthermore, dataprovide little support for the standard “progression rule” (evolution in the direction of older to younger islands) in Hawaiian Islands. Thismay be explained by a recent arrival of the group: age estimates of the diVerent lineages suggest that Havaika colonized the Hawaiian Islandsafter most of the extant islands were already formed. The lack of clear-cut diagnostic characters among species may also be explained by therecent origin of the group since molecular data do not provide any evidence of hybridization among lineages. Quantitative morphologicaldata coupled with the phylogenetic information allow us to reevaluate the current limitation of Havaika taxonomy. Molecular data supportthe existence of at least four diVerent evolutionary lineages that are further morphologically diagnosable. However, genealogical relationshipsare better predicted by geographical aYnity (i.e. island) than by morphological characters used in the original descriptions of the species. Apattern of size segregation linked to largely overlapping distributions of some of the species hints at a potential involvement of competition ingenerating morphological diversity. This study contributes to our understanding on the origin and shaping of the biodiversity of oceanicislands and sets the stage for more detailed studies on particular aspects of these previously overlooked spiders. 2006 Elsevier Inc. All rights reserved.

Keywords: Salticidae; Hawaiian islands; Marquesas; Competition; Island evolution; Phylogeny

1. Introduction

The Hawaiian Islands are often considered a naturallaboratory for evolution, allowing studies of patterns ofdiversiWcation and species formation (Roderick and Gilles-

pie, 1998; Simon, 1987), and development of communities(Gillespie, 2004) and ecosystems (Vitousek et al., 1998). Inthe context of evolutionary biology, studies have shown theimportance of isolation that the islands provide in allowingecological exploration and adaptive radiation (Carson andSato, 1969; Gillespie, 2005; Vandergast et al., 2004).

It has recently been shown that the proportion of speciesendemic to an oceanic island is related linearly to its speciesrichness, from which it was inferred that species diversitymay drive diversiWcation (Emerson and Kolm, 2005). Thisrelationship suggests that competition not only plays a keyrole in structuring biological communities (Gillespie, 2004),

* Corresponding author. Present address: Departament de Biologia Ani-mal, Universitat de Barcelona, Av. Diagonal 645, 08028 Barcelona, Spain.Fax: +34 93 403 5740.

E-mail addresses: [email protected] (M.A. Arnedo), [email protected] (R.G. Gillespie).

1 Fax: +1 510 642 7428.

M.A. Arnedo, R.G. Gillespie / Molecular Phylogenetics and Evolution 41 (2006) 472–495 473

but also may act as a selective agent triggering speciesdiVerentiation. Indeed, character displacement, an outcomeof interspeciWc competition due to limiting resources,appears to be more widespread in nature than previouslythought (Losos, 2000), and its key role has been implicatedin the evolution of diversity in many adaptive radiations(Schluter, 2000a).

Several spider groups with Hawaiian endemic lineageshave been used as models for the study of diVerent aspectsof the evolutionary process. Ecological shifts have played amajor role in the diversiWcation of the spider genus Tetrag-natha (Blackledge and Gillespie, 2004; Gillespie, 2004),while speciation in the endemic spider Orsonwelles has beenmostly driven by inter-island colonization (Hormiga et al.,2003). In all spider groups examined to date, species onyounger islands appear to have been derived from ances-tors on older islands, which is in accordance with prevailing“progression rule” found in the large majority of lineageswithin the Hawaiian Islands (Funk and Wagner, 1995).

The current study uses a unique system of jumping spi-ders: the genus Havaika Prószyjski, 2001 (Araneae: Saltici-dae) in the Hawaiian Islands. Havaika is one of the mostspecies-rich salticid genera in the PaciWc region, only sur-passed by Sobasina Simon, 1898 from the western PaciWc(Berry et al., 1998). It currently comprises three species inthe Marquesas (Berland, 1933, 1934) and nine species in theHawaiian Islands (Prószyjski, 2002; Simon, 1900) althoughgiven the status of knowledge of other spiders groups (e.g.Tetragnatha) prior to recent research, the possibility ofmany more undescribed species is evident (Gillespie, 1999).Recent research using molecular characters for the entirefamily Salticidae, which included 81 genera broadly scat-tered throughout the more than 500 known genera accord-ing to previous notions of the phylogenetic diversity of thefamily, has identiWed the continental genera PellenesSimon, 1876 and Habronattus F. O. P.-Cambridge, 1901 asthe closest relatives of Havaika (Maddison and Hedin,2003b). These results are further supported by somatic andgenitalic morphological similarities (Prószyjski, 2002). Thegenus Habronattus (Masta and Maddison, 2002) is wellknown for elaborate male secondary sexual characteristics,with sexual selection having been demonstrated to play aprominent role in species diversiWcation (Masta and Madd-ison, 2002). However, Havaika males, very much like Pel-lenes males, have only mild ornamentation that showsmoderate diversity.

Hawaiian Havaika species are remarkably variable insize, ranging from 2 to 10 mm in overall length. Althoughabout half of the species of Havaika are single-islandendemics, islands are generally inhabited by more than onespecies that largely overlap in distribution. These observa-tions allow us to hypothesize a possible role of intraspeciWccompetition in species diversiWcation.

Despite the recent taxonomic treatment by Prószyjski(2002), Havaika remains poorly deWned, with the authordescribing the group as a “cluster of similar looking species,diVering by inconspicuous and intergrading characters”

(Prószyjski, 2002). Diagnostic characters are restricted to thebulb of the male palp (Prószyjski, 2003) and the pattern, dis-tribution, and color of setae around the chelicerae and eyes(Simon, 1900). Although these characters cannot be used forspecies identiWcation, they do serve to separate Hawaiianspecies into three phenetic groups: species with reddish andwhite setae covering the proximal anterior part of the chelic-erae and with long male palpal tibia [H. albociliata (Simon,1900), H. canosa (Simon, 1900), H. pubens (Simon, 1900), andH. valida (Simon, 1900)] and species with chelicerae coveredwith long white bristles forming lines, and with a short malepalpal tibia. The latter group is further divided into specieswith a long male bulb embolus [H. cruciata (Simon, 1900), H.jamiesoni Prószyjski, 2001, H. navata (Simon, 1900)] andspecies with a short embolous [(H. senicula (Simon, 1900)and H. verecunda (Simon, 1900)]. The absence of clear-cutlimits separating species in Hawaiian Havaika could beexplained by a recent diversiWcation of the group with insuY-

cient time for Wxation of diagnostic characters. Alternatively,species limits may have been secondarily obscured as a resultof recurrent hybridization events. Indeed, occurrence of nat-ural hybridization has been documented in several Hawaiianarthropod taxa, including Drosophila (Carson, 1989), Lau-pala crickets (Shaw, 1996), and Megalagrion damselXies (Jor-dan et al., 2003).

An additional biogeographic puzzle presented by thegenus Havaika is the inclusion of the Marquesas islands, inaddition to Hawaii, in its distribution (Berland, 1933, 1934).Biological similarities across PaciWc islands, in particularamong snails (Pilsbry, 1900), certain insects (Meyrick,1935a,b), spiders (Berland, 1942), and plants (Brown, 1921;Campbell, 1933; Guillaumin, 1928) led scientists in the ear-lier part of the 20th century to propose the existence of anextensive land mass (submerged around the early Tertiary)in the area currently occupied by the PaciWc depression(Gregory, 1930). However, considerable evidence nowexists to support the hypothesis that the islands within thePaciWc depression are of volcanic origin and acquired theirfaunas by overseas dispersal (Gregory, 1928): The Hawai-ian archipelago originated from a volcanic hotspot (Wil-son, 1963), and the islands of the Marquesas were formed ina similar fashion, though from diVerent hotspots (Nunn,1994). Nevertheless, molecular evidence has recently indi-cated biogeographic connections between the remotePaciWc islands. The Hawaiian plant genus Bidens is sister toa radiation from the Marquesas Islands, and together thisclade is derived from continental America (Ganders et al.,2000). Similarly, Ilex anomala occurs in Hawaii and Tahitiand is related to continental American species. Crab spiders(Thomisidae) appear to form a monophyletic Polynesiangroup (Hawaii + Marquesas + Society Islands) that wouldalso include American representatives (Garb and Gillespie,2006). However, this pattern of phylogenetic aYnity acrossthe far-Xung islands of Polynesia is not universal. Forinstance, endemic lineages of the spider genus Tetragnathawithin the diVerent archipelagoes appear to have arisenfrom independent sources (Gillespie, 2002).

474 M.A. Arnedo, R.G. Gillespie / Molecular Phylogenetics and Evolution 41 (2006) 472–495

Unfortunately, taxonomic understanding of the Mar-quesan species is even worse than that of Hawaiian repre-sentatives, with poor descriptions based on few specimens(in two species only females are known). One of theendemic species originally included in the group (H. rufes-cens Berland, 1934 from Nuku Hiva) has recently beentransferred to the genus Habronattus (Prószyjski, 2002).However, this species does not share any of the synapomor-phies currently assigned to Habronattus, namely a uniqueelbow on the tegular apophysis (TA) of the male palpus(according to drawings, H. rufescens does not even have awell-developed TA) and the male third legs longer than theWrst (Maddison and Hedin, 2003a).

In this paper, we use the information provided by amolecular phylogeny inferred from mitochondrial andnuclear markers, and morphological analyses of a largesample of specimens, to address some of the evolutionaryquestions posed by the spider genus Havaika. SpeciWcally,we examine the relationship between Hawaiian and Mar-quesan representatives, the relationship between morpho-logical variability and species boundaries (and hencedetermine the basis for the diYculty in diagnosing species),and the potential role played by interspeciWc competition inthe diversiWcation of Hawaiian Havaika.

2. Materials and methods

2.1. Taxonomic sampling

2.1.1. IngroupA total of 250 specimens of Hawaiian Havaika were ana-

lyzed for morphology, and for 29 of these we also obtainedDNA sequences (Table 1, localities shown in Fig. 1). Speci-mens were obtained through Weld collections by the authorsor loaned from public institutions (American Museum ofNatural History, Bishop Museum, Muséum national d’His-toire naturelle, and the Natural History Museum in London)and private collections (J. Berry). Type material of HawaiianHavaika was loaned from the Bishop Museum (Honolulu),the Muséum national d’Histoire naturelle (Paris), and theBritish Natural History Museum (London) to help in identi-Wcation. Specimens for sequencing were selected on the basisof their variation in morphological characters used in classi-cal Havaika taxonomy (Prószyjski, 2002; Simon, 1900), theirsex, and island location. Marquesan representatives includedHabronattus rufescens and an undescribed species from HivaOa with a tegular apophysis on the male bulb, hence inferredto belong to the genus Habronattus (referred to as Habronat-tus?). In addition to allowing assessment of phylogeneticaYnities, Marquesan species provided an external calibrationpoint to estimate clade ages in Hawaiian Havaika.

2.1.2. OutgroupOutgroup selection relied heavily on a recently published

study on the molecular phylogeny of jumping spiders(Maddison and Hedin, 2003b). Representatives of clades atsuccessive levels of exclusivity in relation to Havaika were

included in the analyses (see Table 1). The genus Thiodina, amember of the amycoids, was used to root the trees, accord-ing to the results of Maddison and Hedin (2003a).

2.1.3. Molecular phylogenetic analysesTotal DNA was extracted from one or two legs of

freshly collected specimens Wxed in 95% ethanol. Extrac-tion, ampliWcation, and sequencing followed the protocolsdescribed in Arnedo et al. (2004). Partial fragments of themitochondrial genes Cytochrome c oxidase 1 (CO1), thelarge ribosomal subunit (16S), the complete tRNA leucine(tRANleu), and the NADH dehydrogenase subunit 1(ND1), as well as the complete nuclear intron ITS-2 (Inter-nal transcribed Spacer 2) were ampliWed and sequencedusing the following primer pairs: C1-J-1718 and C1-N-2191(Simon et al., 1994) (CO1, 472 bp), LR-N-13398 (Simonet al., 1994), and N1-J-12261 (Hedin, 1997) or N1-J-12334(CCIATTAIAAGAATTGAATATGCTG, this study)(16S-tRNAleu-ND1, »980 bp); in some cases, the formerfragment had to be ampliWed and sequence in two frag-ments with the primer pairs LR-N-13398 and LR-J-12864(Hsiao, pers. comm.) (16S, »430 bp), and N1-N-12945 andN1-J-12261 or N1-J-12334 (tRNAleu-ND1, »550 bp); andITS-5.8S and ITS-28S (White et al., 1990) (ITS-2, »400 bp).ITS-2 sequences were only obtained for the Marquesan andHawaiian taxa.

Sequences including fragments of the 16S, tRNAleu, andITS-2 showed diVerences in length suggesting insertion ordeletion events (indels) during the evolutionary history ofthese sequences. The assignment of positional homology (i.e.alignment) in these situations is not trivial. The most widelyused strategy to accommodate indel events in phylogeneticinference involves the alignment of sequences of diVerentlength by adding gap characters to keep positional similarity.The aligned matrix is then subject to phylogenetic analysesusing any of the available inference methods. Alternatively,the assignment of positional homology can be viewed as partof the phylogenetic inference problem (Mindell, 1991; but seeSimmons, 2004; Simmons and Ochoterena, 2000 for criticalreappraisal of this approach). The direct optimizationmethod (Wheeler, 1996) overcomes the alignment construc-tion step by incorporating indel events as one of the possibletransformations during the tree evaluation process. Thismethod has been considered to be superior because it oVers amore consistent treatment of indels and is less dependent oninitial conditions (Brocchieri, 2001; Thorne et al., 1992; Ving-ron and Von Haeseler, 1997; Wheeler, 1996). However, mostof the tests and inference methods of common use in system-atics require a Wxed alignment or have not yet been imple-mented for a dynamic optimization framework. For thisreason, the present study used both static alignment basedanalyses and direct optimization analyses.

2.1.4. Dynamic optimizationAnalyses under dynamic optimization were performed

with the computer program POY (Wheeler et al., 1996–2003). Although current versions of the program include


settings for maximum likelihood analyses, both the noveltyof the method (results of this option have not yet been fullyexplored) and the elevated computational time led us tolimit the analyses to parsimony.

The 16S and tRNAleu mitochondrial gene fragmentswere considered a single partition that was analyzed individ-ually and in combination with CO1, ND1, and ITS-2. In allthe analyses, the protein coding genes were considered as pre-aligned, which avoids the use of indel transformations duringtree length calculations of these particular partitions. Addi-tional computation time can be saved by dividing gene frag-ments into small putative homologous pieces, without

compromising results (Giribet, 2001). For these reasons, apreliminary static alignment of the 16S–tRNAleu partitionwas built with the aid of the automatic alignment programClustalX (Thompson et al., 1997) using default options. Thefragment was then spliced into eight regions of about 100-bpXanked by a series of 10 identical nucleotides to guaranteethe homology of these regions across all the taxa. ITS-2 wasanalyzed as a single fragment because internal homologousregions were diYcult to identify.

Sensitivity of the results to particular assumptions of theanalyses were investigated using combinations of diVerentgap opening, gap extension, and transversion (tv)/transition

Table 1Taxonomic and locality information of the specimens included in the molecular analyses

GenBank accession number for each gene fragment sequenced. New sequences generated during the present study in bold.

Clade Genus Species/Morph Locality Co1 16s-ND1 ITS

Amycoids Thiodina sp. USA: AZ: Tucson AF327930 AF327958/AF328017 —Ballinae Attidops youngi USA: MO: Valley View Galde AF327990 AF327961/AF327990 —Dendryphantinae Messua limbata USA: AZ: Pinaleno Mts. AF328045 AF327986/AF328045 —Euophryinae Zenodorus microphtalmus HI: Oahu: Wai’anae DQ531816 DQ532097 —Unassigned Heratemita alboplagiata Philippines: Luzon AF327991 AF327962/AF328021 —Synagelinae Peckhamia sp. USA: AZ: Atascosa Peak AF327995 AF327966/AF328025 —Marpissoids Itata sp. Ecuador: Machalilla, NP AF327989 AF327960/AF328019 —Misc. Wx. emb. Hasarius adansoni HI: Oahu: Wai’anae DQ531785 DQ532067 —Marpissoids Platycryptus undatus USA: FL: Newnan’s Lake AF327992 AF327963/AF328022 —Marpissines Marpissa pikei USA: AZ: W of Nogales AF327993 AF327964/AF328023 —Heliophaninae Phintella versicolor HI: Oahu: Wai’anae DQ531815 DQ532096 —Pelleninae Sibianor aemulus Canada: Ontario — AY296675/AY297318 —Pelleninae Pellenes shoshonensis USA: CA: White Mts. AY297383 AF477252 —Pelleninae Pellenes cf. apacheus USA: AZ: Huachuca Mts. — AF477250 —Pelleninae Pellenes cf. longimanus USA: TX: Rio Grande V. — AF477251 —Pelleninae Habronattus sp. MACHAL Ecuador: Manabi, Salaite AY297380 AF477276 —Pelleninae Habronattus mexicanus USA: TX: Pecos River AY297381 AF477353 —Pelleninae Habronattus rufescens Marquesas: Nuku Hiva DQ531803 DQ532084 DQ531817Pelleninae Habronattus? sp. Marquesas: Hiva Oa DQ531801 DQ532082 DQ531818Pelleninae Havaika sp. “pubens” HI: Kauai, Kumuwela DQ531795 DQ532077 —Pelleninae Havaika sp. “pubens” HI: Kauai, Kumuwela DQ531796 DQ532078 DQ531820Pelleninae Havaika sp. “pubens” HI: Oahu: Ko’olau, Aiea DQ531804 DQ532085 DQ531835Pelleninae Havaika sp. “pubens” HI: Oahu: Ko’olau, Waimano DQ531806 DQ532087 DQ531841Pelleninae Havaika sp. “pubens” HI: Oahu: Wai’anae, Palikea DQ531809 DQ532090 DQ531837Pelleninae Havaika sp. “pubens” HI: Oahu: Wai’anae, Palikea DQ531808 DQ532089 (16S) DQ531836Pelleninae Havaika sp. “pubens” HI: West Maui, Pu’u Kukui DQ531812 DQ532093 DQ531831Pelleninae Havaika sp. “pubens” HI: West Maui, Pu’u Kukui DQ531813 DQ532094 DQ531830Pelleninae Havaika sp. “pubens” HI: Maui: Haleakala, Waikamoi DQ531786 DQ532068 DQ531843Pelleninae Havaika sp. “pubens” HI: Maui: Haleakala, Hanawai DQ531787 DQ532069 (16S) DQ531819Pelleninae Havaika sp. “pubens” HI: Big Island, Kahaualea DQ531791 DQ532073 DQ531832Pelleninae Havaika sp. “pubens” HI: Big Island, Waiakea DQ531793 DQ532075 DQ531834Pelleninae Havaika sp. “pubens” HI: Big Island, Laupahoehoe DQ531794 DQ532076 DQ531833Pelleninae Havaika morphotype D HI: Maui: Haleakala, Hana DQ531789 DQ532071 DQ531840Pelleninae Havaika cruciata HI: Big Island, Kipukas Saddle Rd. DQ531790 DQ532072 DQ531838Pelleninae Havaika cruciata HI: Big Island, Kipukas Saddle Rd. DQ531792 DQ532074 DQ531839Pelleninae Havaika sp. “verecunda” HI: Kauai, Kauluaha’ula DQ531797 — —Pelleninae Havaika sp. “verecunda” HI: Kauai, La’au Ridge DQ531798 DQ532079 DQ531828Pelleninae Havaika sp. “verecunda” HI: Kauai, Mt. Kahili DQ531799 DQ532080 DQ531827Pelleninae Havaika sp. “verecunda” HI: Kauai, Makalehas Mts. DQ531800 DQ532081 DQ531826Pelleninae Havaika sp. “verecunda” HI: Oahu: Ko’olau, Aiea DQ531805 DQ532086 DQ531825Pelleninae Havaika sp. “verecunda” HI: Oahu: Ko’olau, Waimano DQ531807 DQ532088 DQ531823Pelleninae Havaika sp. “verecunda” HI: Oahu: Wai’anae, Mt. Ka’ala DQ531810 DQ532091 DQ531842Pelleninae Havaika sp. “verecunda” HI: Oahu: Wai’anae, Mt. Ka’ala DQ531811 DQ532092 DQ531824Pelleninae Havaika sp. “verecunda” HI: Molokai, Kamakou DQ531802 DQ532083 DQ531821Pelleninae Havaika sp. “verecunda” HI: West Maui, Pu’u Kukui DQ531814 DQ532095 DQ531829Pelleninae Havaika sp. “verecunda” HI: Maui: Haleakala, Auwahi DQ531788 DQ532070 DQ531822


(ts) costs (Wheeler, 1995). Maximum congruence amongdata partitions as measured by the ILD (Mickevich andFarris, 1981) was used to select across the diVerent parame-ter cost combinations assayed (Wheeler and Hayashi,1998). The actual parameter combinations investigated arelisted in Table 2.

Analyses were run at either (1) the BioinformaticsCentre of the University of Copenhagen (http://www.binf.ku.dk), using a cluster of 236 P4 processors at 2.4 GHz

connected in parallel with PVM software and the parallelversion of POY (commands–parallel–controllers 3 ineVect); or (2) the Centre de Supercomputació de Catalu-nya (CESCA, http://www.cesca.es), run sequentially oneither an 8-node Compac HPC320 at 833 MHz or an 8-node Compaq Beowulf 600 MHz. The heuristic searchstrategy involved “quick” building of 10 trees by randomaddition of taxa (-buildsperreplicate 10 -buildspr -build-tbr -approxbuild -buildmaxtrees 2), followed by spr and

Fig. 1. (A) Map showing the location of the Marquesas and the Hawaiian Islands in the PaciWc region. (B) Map of the Marquesas with geological agesafter Craig et al. (2001), and collection localities of specimens included in the molecular analyses. (C) Map of the Hawaiian Islands with volcano geologicalages after Carson and Clague (1995) and Price and Clague (2002), and collection localities of specimens included in the molecular analyses.

Waikamoi

Hana

Kahaualea

Waiakea

Laupahoehoe

KumuwelaAiea

Waimano

Palikea

Pu'u Kukui

Saddle Rd.

Hanawai

Auwahi

Kauluaha´ula

La'au

Makalehas

Mt. Kahili

Kamakou

Mt. Ka'ala

HatutuEiao

Nuku Hiva

Ua Huka

Ua Pou

Hiva Oa

TahuataFatu Hiva

20 km.0

KAUAI

OAHU

MOLOKAI

MAUI

HAWAIIA

B

C

Mt. Tekoa

4.75 My

5.40 My

4.00 My

2.65 My

3.65 My

2.01 My

2.35 My

1.55 My

4.70 My

3.00 My2.60 My

2.20 My

1.28 My

1.32 My

1.2 My

0.50 My

LANAI

Table 2Summary of the analyses performed using direct optimization with diVerent parameter combinations

ILD: diVerence between the length of the combined analyses and the sum of the partial analyses divided by the length of the combined analysis; Combinedmt: all mitochondrial genes combined; Combined g: all genes combined. Lowest ILD values in bold. Gap opening, gap extension, and Tv/Ts (transversiontransition ratio) refer to the actual costs used in each particular analyses.

Gap opening Gap extension Tv/Ts Combined mt 16S+tRNAleu CO1 ND1 ILD ITS-2 Combined g Combined mt ILD

Length #Trees Length Length Length Length #Trees Length #Trees Length #Trees

1 1 1 2668 8 899 830 888 0.0191 182 14 963 3 760 4 0.02182 1 1 2731 16 966 830 888 0.0172 213 1 1003 3 768 2 0.02192 2 1 2742 16 974 830 888 0.0182 243 6 1052 6 769 2 0.03802 1 2 4013 18 1440 1205 1282 0.0214 262 6 1322 1 1037 1 0.01742 2 2 4171 2 1473 1205 1282 0.0506 301 50 1376 2 1038 1 0.02694 1 1 2811 16 1041 830 888 0.0185 261 8 1071 8 784 2 0.02434 2 1 2824 16 1051 830 888 0.0195 295 7 1128 6 785 2 0.04264 4 1 2839 8 1061 830 888 0.0211 357 50 1224 6 787 2 0.06544 1 2 4152 8 1587 1205 1282 0.0188 318 20 1398 1 1053 1 0.01934 2 2 4159 4 1591 1205 1282 0.0195 357 3 1454 1 1054 1 0.02964 4 2 4363 2 1603 1205 1282 0.0626 456 4 1548 1 1056 1 0.02334 1 4 6546 2 2409 1935 2057 0.0222 411 27 2028 1 1581 1 0.01784 2 4 6674 2 2538 1935 2057 0.0216 417 4 2087 1 1584 1 0.04124 4 4 6732 2 2591 1935 2057 0.0221 529 18 2190 1 1586 1 0.0342

http://www.binf.ku.dk



http://www.cesca.es

http://www.cesca.es


tbr branch swapping holding one cladogram per round(-sprmaxtrees 1 -tbrmaxtrees 1). Two rounds of tree fusing(-treefuse -fuselimit 10 -fusemingroup 5) and tree drifting(-numdriftchanges 30 -driftspr -numdriftspr 10 -drifttbr -numdrifttbr 10) were added to increase eYciency, holdingup to Wve trees per round (-maxtrees 5) and using the com-mand -Wchtrees, which saves the most diverse cladogramsfound for each island. Additionally, all cladograms foundwithin 0.5% of the minimum tree length (-slop 5 -check-slop 10) were examined to avoid Wnding of suboptimaltrees due to tree length miscalculations (POY uses someshortcuts to speed-up tree evaluation). This strategy wasrepeated 100 times (-random 100) and a maximum of 50trees was retained (-holdmaxtrees 50). Clade support wasassessed by means of Jackknife proportions using 100randomly resampled matrices, with a probability of char-acter deletion of 1/e (default option). Individual searchstrategies involved taking the best tree from Wve rounds ofrandom additions of taxa.

2.1.5. Static alignmentsAlignments with Wxed homology statements, hereafter

referred to as static alignments, of the ribosomal genes andthe ITS-2 were constructed following the method of Hedinand Maddison (2001), building multiple automatic align-ments that correspond to diVerent combinations of gapopening and gap extension costs. The resulting alignmentsrange from relatively gappy alignments to more com-pressed ones. A particular gap opening/extension costalignment is chosen based on topological congruence to theelision matrix (Wheeler et al., 1995), which results fromappending all the alignments constructed for a given genefragment. The method was implemented by building multi-ple alignments with ClustalX (Thompson et al., 1997) underthe following gap opening/extension cost values: 8/2, 8/4,20/2, 24/4, and 24/6. In all cases, transition weight was Wxedat 0.5. Topological congruence was measured by the parti-tion, triplet, and quartet symmetric diVerences as imple-mented in the computer program Component 2.0 (Page,1993). All the former analyses were run with gaps consid-ered as absence/presence data as suggested by Simmonsand Ochoterena (2000). The program GapCoder (Youngand Healy, 2002) facilitates the automatic recoding of gapsusing the simple indel coding version of Simmons and Och-oterena’s method. A mitochondrial combined static matrixwas then constructed by adding the best alignment selectedfor the ribosomal and tRNA genes plus the COI and ND1fragments. The same alignment selection protocol was fol-lowed for the nuclear ITS-2. The combined static matriceswere analyzed using three diVerent gap treatments (gaps asmissing data, gaps as 5th character-state and gaps asabsence/presence characters) to assess their eVect on treeselection.

Analyses of the static alignment matrices under parsi-mony were run with the computer programs NONA v.2(GoloboV, 1993) and TNT v.1 (GoloboV et al., 2003) andthe matrices and tree were manipulated with the program

WINCLADA v.1.00.08 (Nixon, 2002). Heuristic searchesinvolved 100–500 rounds of random addition of taxa, hold-ing Wve trees per round and a total maximum of 1000. Bestoverall trees were further subjected to a new round of tbrbranch swapping. Clade support was assessed using jack-kniWng, with 1000 jackknife pseudoreplicates. Individualsearches consisted of 10 rounds of random addition of taxa,holding two trees per replicate and an overall maximum of1000.

The program Modeltest v.3.06 (Posada and Crandall,1998) was used to select the model of evolution thatexplained the data signiWcantly better with fewer parame-ters using the AIC criterion (Buckley et al., 2002). Maxi-mum likelihood analyses of the static alignments wereperformed with the computer program PAUP* v.4.0 (Swo-

Vord, 2001) using the model of evolution and parametervalues suggested by Modeltest and a heuristic search with10 random stepwise additions of taxa and TBR branchswapping. Additional ML trees and bootstrap proportions(500 pseudoreplicates) were calculated with the aid of thecomputer program PHYML v.2.4 (Guindon and Gascuel,2003). Bayesian inference was implemented with the com-puter program Mr Bayes v.3.0 (Ronquist and Huelsenbeck,2003). Four simultaneous MCMCMC chains (one cold andthree heated) were run for 1,500,000 iterations. Three inde-pendent runs were performed to ensure convergence of theresults. Plots of the negative log likelihood against the num-ber of generations were used to identify the point of sta-tionarity, and trees obtained from generations before thispoint were discarded as burn in. All analyses were run withspeciWc and unlinked GTR + I + ! models for each genefragment, although all mitochondrial genes were combinedin a single data set.

2.2. Data partition congruence and alternative hypothesis testing

Congruence among mitochondrial and nuclear parti-tions was assessed on the static alignments under the parsi-mony criterion using the ILD test (Farris et al., 1994), asimplemented in the program Winclada. Heuristic searcheson 1000 pseudoreplicates consisted of Wve random additionof taxa holding two trees per iteration and a maximum of10 trees overall. In all cases, uninformative characters andincomplete taxa were removed from the matrices beforeperforming the test, as suggested in the literature (Arnedoet al., 2001; Cunningham, 1997).

The ILD test has been shown to be excessively prone totype 1 error (i.e. it may indicate signiWcant incongruenceamong partition when they are actually not) (Barker andLutzoni, 2002). For this reason, incongruence among parti-tions was investigated further using topology tests. TheTempleton (Templeton, 1983) and Winning-site (Pragerand Wilson, 1988) tests were used for parsimony analysesand the Shimodaira–Hasegawa (Shimodaira and Hase-gawa, 1999) test for maximum-likelihood analyses. Thesame tests were also applied to elucidate the diVerences


between alternative topological hypotheses. All tests wererun in PAUP 4.0. Finally, local incongruence was examinedby looking for well-supported but irreconcilable clades inthe trees derived from separate analyses of the partitions(Wiens, 1998). This is the only method to assess incongru-ence among partitions currently available for analysesunder direct optimization.

2.3. Estimate of lineage divergence time

A likelihood ratio test rejected the presence of a strictmolecular clock in the molecular phylogeny of Havaikaspecies (p D 0.001, !2 D 56.89, 28 d.f.). Several methodshave been proposed to estimate divergence times in theabsence of a molecular clock (Huelsenbeck et al., 2000;Sanderson, 1997, 2002; Thorne and Kishino, 2002; Yoderand Yang, 2000). Here we use a semi-parametricapproach, the penalized likelihood method (Sanderson,2002), which has been proven to be superior to both fulland non-parametric approaches (Sanderson, 2002). Penal-ized likelihood (PL) requires choice of a smoothingparameter value ("), which determines the variation in therates of evolution across branches. In his proposition ofthe PL, Sanderson suggested cross-validation, a widelyused model-selection method based on comparing param-eter estimates after sequential branch removal, as anobjective criterion for choosing "’s value or even for con-trasting other estimators. Time of divergence and cross-validation calculations were carried out with the Linuxversion of the computer program R8S v.1.5 (Sanderson,2003). At least one calibration point is required for esti-mating absolute divergence times, regardless of themethod of choice. Fossil data are rarely available whenworking at the species level, necessitating the use of bio-geographic information which must be treated with cau-tion since association between cladogenesis andgeological events is always diYcult to prove (Cacconeet al., 1997). However, volcanic “hotspot” archipelagoes,such as the Hawaiian Islands or Marquesas provide excel-lent opportunities for calibrating rates of molecularchange based on the geological age of the islands (Fig. 1).Estimates of the date of an island’s formation can then beused as a maximum age for taxa inhabiting the island.Nevertheless, this assumption is based on certainpremises, the most important ones being that (1) thebranching pattern in the phylogeny parallels the timing ofisland formation and (2) that the divergence between sis-ter taxa does not greatly predate the formation of the col-onized younger island (Fleischer et al., 1998). Baldwin andSanderson (1998) have cautioned against the use of inter-nal calibration points for dating evolutionary events inhotspot archipelagoes because errors in the placement ofisland age are greatly magniWed deeper within the tree.Additional sources of error can arise from extinction oflineages and inclusion of the divergence that was alreadypresent in the common ancestral population (Emersonet al., 2000).

ConWdence intervals for each clade age were estimatedwith PAUP* from 100 bootstrap replicates of the originaldata matrix and the preferred tree constrained. Mean andstandard deviations of the bootstrapped branch lengths ofeach clade were calculated with r8s and used to construct95% conWdence intervals.

2.4. Analysis of morphological variation

Characters traditionally used for diagnosing HawaiianHavaika species were scored for available specimens. Thesecharacters refer to the face (clypeus and frontal side of thechelicerae) ornaments and the structure of the male copu-latory bulb. Three types of face patterns have beenreported in Hawaiian Havaika (Simon, 1900): (1) Yellow-ish, scale-like hairs on the frontal part of the basal segmentof the chelicerae (Fig. 2A and F) in both male and femalespecimens; (2) several rows of white bristles instead ofscales on the chelicerae and a necked clypeus (Fig. 2D andG–J) in both males and females; and (3) female clypeusand basal-most anterior part of the chelicerae densely cov-ered with white bristles (Fig. 2B, C, and E). The male pedi-palp show clear diVerences in at least two characters(Prószyjski, 2002): some bulbs have a long and slenderembolus emerging from the basal half of the tegulum(Fig. 2K, L, and N), while in others the embolus is shortand stout and emerges from the distal half of the tegulum(Fig. 2M and O). The relative length of the palpal tibia,compared with the length of the cymbium, also showdiVerences across individuals. In some cases, the tibia isclearly shorter than the cymbium (Fig. 2L–N), while inothers it is equal in size or even longer (Fig. 2K and O).The diVerent kind of face ornaments were recorded forevery specimen examined and in male individuals also theratio between the length of the palpal tibia and the cym-bium, the ratio between the distance from the base of thetegulum to the base of the embolus and the total length ofthe tegulum. Finally, the length of the carapace, a surro-gate for body length, was measured in all specimens toaccount for the large variation in size reported in thegenus. All measurements were taken in millimeters using adissection scope either Leica MZ16A or Leica MZAPOequipped with an ocular measuring graticule.

2.5. Statistical procedures

2.5.1. Morphotype discrimination analysisA canonical correspondence analysis (CCA) was per-

formed to assess the discriminating power of morphologi-cal characters based on measurements. Specimens wereassigned to lineages as suggested by face ornament patternsand island (see Results). Analyses were run with the com-puter program CANOCO for Windows v.4.5 (Hill’s scalingoption on, Monte Carlo permutation test enabled using 999permutations). Results were visualized by plotting the Wrsttwo discriminant axes and adding information on discrimi-nating characteristics, the centroids of classes, and individ-


ual’s canonical scores. Biplots and scatter plots wereconstructed with the program CANOCO DRAW for Win-dows v.4.0.

2.5.2. Size diVerentiationVariation in size and its association with lineages, sex,

and/or island population was investigated by means ofstandard statistical techniques. Individuals from eachisland were grouped into diVerent lineages based onsomatic characters discussed above coupled with the infor-mation derived form the molecular phylogeny. For eachisland, a two-way analysis of variance was performed withfactors (1) lineage and (2) sex, and dependent variable “car-apace length.” When applicable, post hoc multiple compar-isons were run using the Student–Newman–Keuls (SNK)test (Zar, 1984). For some islands, representatives of bothsexes were not available for all lineages. In these cases, sig-niWcant diVerences between lineages were assessed using a ttest. Additionally, diVerences in size among island popula-tions of each lineage were investigated by means of one-way ANOVA. Normality and homogeneity of variance ofthe data were assessed by Kolmogorov–Smirnov and Bart-lett tests, respectively (Zar, 1984). Only data from Oahu didnot pass the homogeneity test and data from pubens (seebelow) within islands did not pass normality test. Thesedata were subject to rank transformation prior to the anal-ysis (Potvin and RoV, 1993).

SigniWcant size diVerences among lineages from a partic-ular volcano were also investigated for Oahu and the MauiNui complex. The last name refers to the group of islandscomprising Molokai + Lanai + Maui, which constituted asingle land mass until about 1.2 million years ago. About

0.6 my ago, Molokai split from Maui Nui, followed about0.4 my ago by Lanai (Price and Elliott-Fisk, 2004).

3. Results

3.1. Phylogenetic analyses

3.1.1. Dynamic optimizationResults of the multiple analyses performed under diVer-

ent parameter values are summarized in Table 2. For themitochondrial data set, the parameter combination withgap opening twice the value of the uniformly weighted gapextension and transition/transversion ratio maximizedcharacter congruence across partitions as shown by theILD (Fig. 3). Pellenine monophyly is supported with lowjackknife values. Hawaiian Havaika are shown as mono-phyletic with 100 jackknife support, and its inclusion into aclade containing mainland Habronattus and Pellenes spe-cies as well as the Marquesan representatives is also wellsupported. Hawaiian Havaika are divided into several well-supported clades that correspond to species groups diag-nosed by face ornaments and male bulb characters. TheWrst clade includes male and female specimens from Kauai,Oahu, Maui, and Hawaii, all of which bear yellowish scaleson the chelicerae and males have long palpal tibia (cym-bium/tibia length < 1.3). Island populations of this clade areshown as monophyletic. We will refer to this monophyleticgroup as the pubens clade. In the second clade, which is sup-ported by very high jackknife values, males bear rows oflong white bristles on the chelicerae and female specimenshave a hairy clypeus. All males in this clade have short pal-pal tibia (cymbium/tibia length > 1.3). Island populations

Fig. 2. Morphological traits of Hawaiian Havaika. Female face ornamentation: (A) pubens lineage; (B) H. cruciata; (C) verecunda lineage; (D) morphotypeD; (E) morphotype Necker and Nihoa. Male face ornamentation: (F) pubens lineage; (G) H. cruciata; (H) verecunda lineage; (I) morphotype D; (J) mor-photype Necker and Nihoa. Copulatory bulb: (K) pubens lineage; (L) H. cruciata: (M) verecunda lineage; (N) morphotype D; (O) morphotype Necker andNihoa; (P) Habronattus? sp. Hiva Oa; (Q) Habornattus rufescens Nuku Hiva. Overall dorsal view, female: (R) H. cruciata; (S) morphotype D.


also are shown as monophyletic (Molokai and Maui con-sidered a single unit) with high jackknife values with a sin-gle exception: one haplotype from East Maui joins thehaplotypes from the Big Island (i.e. Hawaii Island) in awell-supported clade. The single male included in the EastMaui-Big Island clade shows a long, slender embolus grow-ing below the half part of the tegulum, allowing identiWca-tion of haplotypes from the Big Island as H. cruciata. TheMaui female shows some somatic diVerences from BigIsland females, including a sparsely haired clypeus and anoverall dark brownish color pattern; we will refer to thisspecimen as morphotype D. Finally, remaining Kauai,Oahu, and Maui Nui haplotypes are joined in a clade,which is characterized by males having a short, stout bulbembolus that grows from above the half distal part of thetegulum. We will refer to this clade as the verecunda clade.Relationships among island lineages in clades pubens andverecunda are mostly unresolved.

Pellenine monophyly and monophyly of the mainlandHabronattus and the Marquesan species is recovered in half

of the parameter cost combinations implemented. Hawai-ian Havaika and each of the named clades are alwaysmonophyletic regardless of the parameter cost combina-tion. Conversely, the internal topologies of pubens and vere-cunda clades are very sensitive to changes in the cost of theexplored parameters.

The lack of additional partitions in the nuclear data(ITS-2) precluded the use of congruence across data sets asa criterion for choosing the best parameter combination.Fig. 4 shows the results for the same parameter combina-tion selected for the mitochondrial data. Monophyly ofeach Hawaiian Havaika named clades and the sister-grouprelationship between (morphotype D + H. cruciata) and theverecunda clade is recovered but not supported by jackknifevalues. In several cases, island sequence types did not formmonophyletic groups, for instance, the sequence-typeOW111 from Oahu always clustered with sequence typesfrom Hawaii, while the sequence-type OK9 from Oahuclustered with Maui Nui sequence types (only some treesunder cost scheme 441 support Oahu monophyly). In some

Fig. 3. Strict consensus tree of 16 trees of length 2731 steps resulting for the direct optimization analysis with parsimony of all mitochondrial gene frag-ment combined, using the following parameter values: gap opening 2, extension gap 1, transversion/transition ratio 1. Jackknife support values >50 areshown below branches.

Haw

aiian IslandsZenodurus microphtalmusSibianor aemulus

Pellenes cf. longimanusPellenes cf. apacheus

Habronattus rufescensPellenes shoshonensis

Habronattus sp.Habronattus? sp. Hiva Oa

Habronattus mexicanus

Havaika OW111Havaika OW158

Havaika OK23Havaika OK8

Havaika WM159Havaika WM88

Havaika EM128Havaika EM130

Havaika K146Havaika K112

Havaika H83Havaika H109

Havaika H137

Havaika K38


Havaika K85

Havaika OW28Havaika OW29

Havaika OK9Havaika OK24

Havaika MK82Havaika WM89

Havaika EM81


Havaika EM90

Phintella versicolorHasarius adansoni Heratemita alboplagiataItata sp.

Platycryptus undatusPeckhamia sp.Marpisaa pikei

Messua limbataAttidops youngi

Thiodina sp.

PE

LLEN

INE

S

Marquesas

Havaika

100

53

90

78

100

99100

99

100

10099

100

54

80

100

100

100

98

100

100

97

10097

100

91

10089

100

pubenslineage

H. cruciata

verecundalineage

OAHU

KAUAI

MAUI NUI

HAWAII

Morph D


other cases, island sequence-type monophyly seemed to beparameter value-dependent. The verecunda clade is alwayssupported as monophyletic except in the parameter combi-nation 414. Monophyly of pubens and (morphotype D + H.cruciata) clades is supported in combinations where the gapextension value was smaller than gap opening. However,none of the topological inconsistencies with the mitochon-drial analyses received signiWcant (>50) jackknife support.

Simultaneous analyses of the combined data set underdiVerent parameter combination are summarized inTable 2. Only taxa with information for both mitochon-drial and the nuclear partitions were included in this

analysis (Marquesan species and Hawaiian Havaika).Trees obtained with gap opening twice gap extension costand transversions twice the cost of transitions maximizedcongruence between the mitochondrial and nuclear parti-tions and were selected as reference (Fig. 5). For theHawaiian taxa, the results strongly resemble those of themitochondrial data set alone, which is not surprisinggiven the overall higher support values for the branchesin the mitochondrial tree: clades pubens, (morphotypeD + H. cruciata), vercunda, and (morphotype D + H.cruciata + vercunda) and the monophyly of each islandpopulation are well supported, but the relationships

Fig. 4. Single tree of length 213 steps resulting for the direct optimization analysis with parsimony of the ITS-2 using the following parameter values: gapopening 2, extension gap 1, transversion/transition ratio 1. Jackknife support values >50 are shown below branches.

Habronattus sp.Habronattus? sp. Hiva Oa

Havaika EM130Havaika K146

Havaika MK82Havaika EM81

Havaika OK9

Havaika OW29Havaika OK24


Havaika K85Havaika WM89

Havaika WM88Havaika WM159

Havaika H109Havaika H83Havaika H137

Havaika OK23

Havaika OW111

Havaika OW158


Havaika EM90

Havaika OK8

Havaika OW28

Havaika EM128

Haw

aiian Islands

Marquesas

pubenslineage

H. cruciata

verecundalineage

100

7163

55

61

90

6451OAHU

KAUAI

MAUI NUI

HAWAII

Morph D

Fig. 5. Single tree of length 1037 steps resulting for the direct optimization analysis with parsimony of the combined mitochondrial and nuclear partitionsusing the following parameter values: gap opening 2, gap extension 1, transversion/transition ratio 2. Refer to Table 7 for clade support values.


between island populations in pubens and verecunda arenot. The position of the specimen HavEM130 in thepubens clade is sensitive to changes in the parameters ofthe analyses, with Wve out of 14 combinations (gap open-ing: gap extension: tv/ts D 212, 222, 414, 424, 444) sup-porting Maui Nui monophyly. Similarly, twocombinations (441, 442) put specimen HavOW111 at thebase of the Big Island populations. These observationssuggest that the topological incongruence between themitochondrial and nuclear partitions may not be simplynoise but rather the result of diVerent molecular mecha-nisms acting on each partition.

3.2. Static alignments

Results of the exploratory analyses of the multiple staticalignments obtained under diVerent gap opening/extensioncosts are summarized in Table 3. Alignments of the 16SrRNA + tRNAleu partition generated under costs 20/2 and26/4 showed the smaller topological distances with the eli-sion tree, according to the topology metrics calculated. ITS-2 alignments under costs 24/4 and 24/6 yielded the mostsimilar trees to the elision matrix. Since topology metricsshowed no diVerence between the two best costs, one of thetwo best alignments of each partition was chosen arbi-trarily: 20/2 for the 16S rRNA + tRNAleu partition and 24/4 for the ITS-2.

Analyses of the mitochondrial data sets under uni-formly weighted parsimony, treating gaps as a 5th state,gaps as an absence/presence character and as missingdata, yielded eight trees of 2849 steps, four trees of 2832steps, and 12 trees of 2762 steps, respectively (results notshown). The AIC criterion as implemented in Modeltestselected HKY85 with invariants and gamma distribution(four categories) as the most appropriate evolutionarymodel to analyze the data. A heuristic search using maxi-mum likelihood resulted in one tree (¡log L 13,938.06)that is shown in Fig. 6. Table 4 summarizes clade supportfor the diVerent analyses performed on the static mito-

chondrial matrix, including those clades obtained in aBayesian inference analyses with speciWc GTR + I + !models of evolution for each of the mitochondrial parti-tions (16S rRNA + tRNAleu, CO1, ND1), averagedacross three independent iterations with 1,500,000 gener-ations and discarding 15,000, 17,000, and 22,000 trees asburn in, respectively.

In general, results from static analyses, both in termsof topology and clade support, are noticeably congruentwith trees obtained through dynamic alignment. Staticparsimony failed to Wnd support for the pellenines, butthe clade is supported in model-based analyses. Mono-phyly of mainland Habronattus and Pellenes, and Mar-quesan species, is strongly supported in all analyses,while their interrelationships are not. The putative mono-phyly of the Marquesan and Hawaiian endemics is nei-ther supported nor rejected by any of the testsimplemented, regardless of the inference method and thegap treatment (Table 5).

Monophyly of Hawaiian Havaika, monophyly of eachof the pubens, (morphotype D + H. cruciata) and vere-cunda lineages and their interrelationship, as well asmonophyly of island populations of each clade are recov-ered in every analysis, with high clade support values.Conversely, the relationships among the island popula-tions of clades pubens and verecunda, respectively, areanalysis-dependent, and never receive signiWcant supportvalues, with the only exception being the basal positionof the Kauai populations of clade III relative to the MauiNui and Oahu populations, supported by the model-based analyses.

Analyses of the ITS-2 data sets under uniformlyweighted parsimony treating gaps as a 5th state, gaps asabsence/presence characters and gaps as missing data,yielded six trees of 210 steps, one tree of 169 steps, andone tree of 137 steps, respectively (results not shown).The TVM with invariants and gamma distribution (fourcategories) was selected as the best model for the data bythe AIC criterion, as implemented in Modeltest. A

Table 3Summary of the results of the parsimony analyses with gaps as absence/presence characters of the static alignments produced by Clustal under diVerentvalues of gap opening and extension

8/2: gap opening 8, gap extension 2 (remaining costs follow same convention); A/P characters: number of new characters added to the matrix after recod-ing the presence of speciWc gaps; SD: symmetric diVerence measurements.

Costs length #trees A/P chars. Partition SD Triplets SD Quartets SD

16S+tRNAleu8/2 1039 10 72 10 0.070 0.0928/4 1053 24 58 6 0.030 0.06120/2 1050 22 39 4 0.052 0.03124/4 1049 7 38 5 0.061 0.03524/6 1050 19 39 4 0.052 0.031

ITS-28/2 165 1 32 4 0.017 0.0518/4 165 1 32 5 0.022 0.05120/2 163 10 28 3 0.018 0.05624/4 169 1 26 3 0.009 0.02524/6 169 1 26 3 0.009 0.025


heuristic search using maximum likelihood resulted intwo trees (¡log L 1362.99) (strict consensus shown inFig. 7). Table 6 summarizes clade support for the diVer-ent analyses performed on the static ITS-2 data set,including clades obtained in a Bayesian inference analy-ses under the GTR + I + ! model of evolution, with1,500,000 generations and discarding 25,000 trees as burnin. The eVect of the inclusion of gaps in the model-basedapproximations was studied by running additionalBayesian analyses with gaps coded as presence/absence inan independent partition and with a particular stochasticmodel. Results were almost identical to the analyses per-formed with gaps as missing data.

All analyses run on the static alignment of the ITS-2data set support monophyly of pubens and (morphotypeD + H. cruciata) lineages, while the verecunda clade is alsorecovered but with marginal support only. The sister-grouprelationship of the verecunda clade to the clade (morpho-type D + H. cruciata) is supported under parsimony treat-ing gaps either as 5th state or absence/presence characters.In some cases, island populations of each lineage were notrecovered as monophyletic, as was also reported for thedynamic optimization analyses. Similarly, the major area ofconXict with regard to the mitochondrial analyses involvessingle haplotypes of pubens from Oahu and Maui that clus-ter with the haplotype from Kauai instead of with others

Fig. 6. Maximum likelihood single tree of score ¡ln D 13,938.06472 resulting from analysis of all mitochondrial gene fragments combined using theHKY + I + ! model with following parameter values: transition/transversion ration D 2.156; nucleotide frequencies: A D 0.39750, C D 0.10550,G D 0.07270, T D 0.42430; among-site rate variation with invariant sites I D 0.4348 and gamma shape parameter # D 0.7497. Refer to Table 4 for clade sup-port values.


from the same island. These results are only (very weakly)supported under maximum likelihood. Statistical tests onthe topologies obtained by constraining the island sequencetypes of each lineage to monophyly oVer mixed results(Table 5). The monophyly of the Oahu sequence types ofverecunda is never rejected but the monophyly of the othergroups is dependent on the test and the gap treatment.Monophyly of the Maui Nui sequence types of pubens isonly marginally rejected by the Templeton test in some

trees with gaps as missing data or coded as presence/absence. However, signiWcance of the test is lost after Bon-ferroni corrections for multiple comparisons. The sameapplies for the Oahu population, although in this case low-est p values are obtained with gaps as 5th state. Templetonand Winning-site tests agree in rejecting simultaneousmonophyly of the former groups, regardless of the gaptreatment, although in some cases signiWcance is lost afterBonferroni corrections. The Shimodaira–Hasegawa test ofthe maximum likelihood trees agrees in rejecting the simul-taneous monophyly of the Maui Nui and Oahu popula-tions of pubens and the Oahu populations of verecunda,although it fails to reject any of them when constrainedindependently. The ILD test reveals signiWcant incongru-ence (p < 0.001) between the mitochondrial and ITS-2 datasets, regardless of the gap treatment enforced.

The mitochondrial and nuclear data sets were merged in asingle matrix and incomplete taxa for one of the two parti-tions removed. Analyses of the combined data set under par-simony analyses with gaps as 5th state, as absence/presencecharacters and as missing data, resulted in 10 trees of 1052steps, eight trees of 1004 steps, and 10 trees of 955 steps,respectively. The AIC criterion implemented in Modeltestselected K81uf+ I+ G as the best model for the combineddata set. ML analysis resulted in one tree (¡logL7734.19040). Results using Bayesian inference were obtainedby averaging through three independent runs of the MCM-CMC and discarding in each run 35,000 trees as burn in.

Table 7 summarizes clade support across all the analysesperformed on the combined nuclear and mitochondrialmatrix. Results nearly mirror those obtained for the mito-chondrial data set alone, except for the lower support forthe pubens and (morphotype D + H. cruciata + verecunda)lineages in the maximum likelihood analysis. The similarityof the result is not surprising given that the conXictingclades reported in the partial analyses of the ITS data setwere only marginally supported.

3.3. Lineage divergence time

Divergence times were estimated on the topologyobtained in the direct optimization analysis of all gene frag-ments (parameter combination gap opening 2, gap exten-sion 1, tv/ts 2). The same topology was obtained in theBayesian analyses of the static matrices and it is not in con-Xict with any of the remaining analyses. Only mitochon-drial genetic divergences were used since there was evidencethat some ITS-2 relationships may not reXect historical pat-terns (see Discussion for details). Age estimation wasrestricted to the Marquesan species and Hawaiian Havaikarepresentatives. Bianor was included in the analyses as adistant outgroup to asses the root node, but it was prunedbefore calculation of divergence times. Maximum likeli-hood branch lengths were obtained with PAUP, using themodel selected by AIC in Modeltest with the preferredtopology constrained (K81uf + I + G). The split between theMaui and Big Island populations of morphotype D + H.

Table 4Clade support values of all the diVerent analyses performed on staticalignment of the mitochondrial data set

Numbers refer to jackknife support (parsimony analyses), bootstrap pro-portions (maximum likelihood), and posterior probabilities (BayesianInference). Clade numbers as shown in Fig. 6. —: clade supported belowthreshold value (50% for jackknife and bootstrap, 95% posterior probabil-ity); NS: clade not supported; Wfth: parsimony analyses with gaps as 5thstate; A/P: parsimony analysis with gasp recoded as absence/presencedata; ?: parsimony analysis with gaps as missing data; ML: maximumlikelihood analysis; BI: Bayesian inference analysis.

Clade # Clade description 5th A/P ? ML BI

1 NS NS NS — 962 NS NS — 62 993 NS NS — — 994 — NS — — 975 NS NS — — —6 68 60 — 58 977 62 58 NS — 988 NS NS NS — —9 72 75 66 — —

10 Pellenines NS NS NS 84 9711 89 76 86 81 9512 — — — 51 —13 Pellenes 100 100 100 100 10014 100 100 100 100 10015 NS NS NS — —16 Habronattus 78 79 78 83 9517 Marquesas sp. 100 99 99 100 10018 Hawaiian Havaika 100 100 100 100 10019 pubens 99 98 99 100 10020 pubens Maui Nui 72 73 74 81 9921 — — — 52 —22 100 100 100 100 10023 NS NS NS — NS24 pubens Big Island 100 100 100 100 10025 — — — 65 9826 NS NS NS NS —27 pubens Kauai 100 100 100 100 10028 pubens Oahu 100 100 100 100 9929 99 99 99 94 9930 100 100 100 100 10031 verecunda + D + H. cruciata 95 95 96 100 10032 D + H. cruciata 100 100 100 100 10033 H. cruciata Big Island 99 100 100 100 10034 verecunda 99 99 99 100 10035 verecunda Kauai 100 100 100 100 10036 — — — 50 —37 90 88 88 90 —38 — — — 67 9539 verecunda Maui Nui 99 100 100 100 10040 99 99 99 100 10041 verecunda Oahu 100 100 100 100 10042 87 88 87 83 —43 100 100 100 100 100


cruciata is the only well-supported node available for cali-bration inside Hawaiian Havaika. The age of this node wasWxed to 0.5 million years (my), assuming this age of originof the Big Island (Price and Clague, 2002). Other branchingpatterns in pubens and verecunda clades that also followedthe timing of island formation were initially not used as cal-ibration points because they were poorly supported by thedata (low bootstrap or jackknife values). The split betweenthe two Marquesan species provide an external calibrationpoint. However, in this case the age of the node was notWxed but constrained to a minimum age of 2.01 my and amaximum of 4 my, which corresponds to the age of the old-est and youngest islands inhabited by these species (Craiget al., 2001). Age estimation was performed using the penal-ized likelihood method and the truncated Newton algo-rithm, with the smoothing parameter set to 1000. Thesesettings were selected after a preliminary cross-validationanalyses that included other values of the smoothingparameter as well as the NPRS and Langley–Fitch meth-ods. Problems with local optima in the calculation of diver-gence times were addressed by running three analyses withrandom starting values. Results are shown in Table 7. Theaverage rate of sequence evolution observed was 0.02649per site per million years, which is about twice the ratereported for arthropod mitochondrial DNA (Brower,1994). Most of the age estimates are in accordance with thegeological setting, as far as they suggest time of origins oflineages that are younger than the age of the islands theyinhabit. However, there is one major exception: The splitbetween the Maui Nui and the Big Island populations ofpubens is inferred to be about three times older than the BigIsland. A new set of divergence estimate analyses were runwith alternative constrained clade ages. Both removal ofinternal age constraints (i.e. only Marquesan species diver-

gence constrained) and simultaneous constraint of the timeof divergence of the two Maui Nui-Big Island splits in thetree (set to 0.5 my) resulted in divergence time estimatesthat were far older than the inhabited islands. In both cases,cross-validation selected the NPRS method over the PLmethod. However, when only the split between the MauiNui and the Big Islands populations of pubens was Wxed(0.5 my), the PL method was preferred (smoothingparameter D 1), and the estimated ages were again compati-ble with island ages. Moreover, the estimated time of diver-gence between morphotype D and H. cruciata was onlyslightly higher than the age of the youngest islands (Table7), although the resulting average rate of sequence evolu-tion was higher than in previous analyses (0.03843).

3.4. Morphological analyses

As it currently stands, the taxonomy of Hawaiian Hava-ika is of little use for species identiWcation. However, phylo-genetic analyses provide strong support for the separationof Hawaiian species into lineages characterized by faceornament patterns. The pubens lineage is characterized bythe presence of yellowish scales on the chelicerae in bothmales and females. The verecunda lineage shows sexualdimorphism with males bearing rows of white bristles onthe chelicerae and females having few bristles on the chelic-erae but dense white hairs on the clypeus. A single femalespecimen from Maui Nui shows sparse clypeal white scalesrestricted to the lower margin of the clypeus (Fig. 2), aunique pattern not observed in any other female specimenand, therefore, it is assigned to a new lineage: morphotypeD. A fourth lineage, H. cruciata, includes specimens fromthe Big Island in which face ornamentation patterns areundistinguishable from the verecunda lineage. However,

Table 5Summary of the results of the statistical tests of alternative topologies

Constraints: None: no constraint enforced; pubens Maui Nui: monophyly of all ITS-2 sequence types of the pubens lineage from Maui Nui; pubens Oahu:monophyly of all ITS-2 sequence types of the pubens lineage from Oahu; verecunda Oahu: monophyly of all ITS-2 sequence types of the verecunda lineagefrom Oahu; All: all former constraints enforced; MA + HI: monophyly of haplotypes from Marquesas and Hawaiian Havaika. 5th: parsimony analyseswith gaps as 5th state; A/P: parsimony analysis with gaps recoded as absence/presence data; ?: parsimony analysis with gaps as missing data; ML: maxi-mum likelihood analysis. #t: number of trees; T: p values of the Templeton test; WS: p values of the Winning-site test; SH: p values of the Shimodaira–Hasegawa test.

¤ SigniWcance at p < 0.05.

Constraint 5th A/P ? ML

#t T WS #t T WS #t T WS #t SH

ITS-2None 14 — — 1 — — 1 — — 2pubensMaui Nui 72 0.0833–0.2568 0.25–0.4531 3 0.0455¤–0.1025 0.125–0.2188 234 0.0455¤–0.1025 0.125–0.2188 2 0.162pubensOahu 24 0.0008¤–0.0093¤ 0.0010¤–0.0169¤ 3 0.0143¤–0.0339¤ 0.0313¤–0.0703 393 0.0143¤–0.0578 0.0313¤–0.125 1 0.067verecundaOahu 32 0.3173–0.763 1.0000 1 0.3173 1.0000 64 – — 2 0.855

All 40 0.0001¤–0.0017¤ <0.0001¤–0.0026¤ 3 0.0067¤–0.0209¤ 0.0078¤–0.0391¤ 242 0.0114¤–0.0335¤ 0.0156¤–0.0703 2 0.041¤

MitochondrialNone 8 4 14 1MA + HI 8 0.5930–0.9434 0.4334–0.8036 8 0.8185–0.8709 0.8784-0.9087 4 0.8982–0.9014 1.0000 1 0.445


bulb characters provide further discrimination in thisgroup. The verecunda clade has a short palpal tibia (palpcymbium/tibia ratio ranging from 1.333 to 1.500) and astout embolus emerging from the distal part of the tegulum(tegulum/embolus origin ratio ranging from 0.500 to 0.576).Conversely, the only H. cruciata male sampled combines ashort palpal tibia (1.529) with a slender, long embolus origi-nating from the basal half of the tegulum (0.278). The sametype of embolus is also observed in pubens males (tegulum/embolus origin ratio ranging from 0.281 to 0.450),although, in this case, they have a long palpal tibia (palpcymbium/tibia ratio ranging from 0.750 to 1.278). Alongwith face ornaments and male palp characters, othersomatic traits characterize the diVerent lineages. Specimensin the pubens clade have a dark brown carapace, with twolighter longitudinal bands on the dorsal part, and a creamcolored abdomen with a variable darker chevron patternand shiny scales on the dorsal side. Specimens in this cladeare stouter and more robustly built than species in otherclades. H. cruciata and the verecunda lineage have a blackcarapace, sometimes bearing a brush of white hairs on the

mid-frontal part, with two longitudinal spots of whitishhairs, and abdomen light cream colored with very variabledarker chevron patterns. Morphotype D stands in sharpcontrast to H. cruciata in terms of body coloration, beingalmost uniformly dark brown with grayish longitudinalpatterns on the abdomen.

Phylogenetic data indicate that island populations ineach lineage are monophyletic (considering Maui Nui as asingle island, see above), although none of the morphologi-cal characters considered seem to distinguish betweendiVerent island populations. Genetic divergences (Table 8)between islands in the same clade range from 0.022 to 0.082(within island values: 0.004–0.039) and provide further evi-dence that island populations are well characterized from amolecular standpoint. Lineages (pubens, verecunda) wereaccordingly further divided into island populations.

The morphological study of about 200 additional speci-mens identiWed additional island populations of the afore-mentioned lineages. The pubens lineage was sampled in themolecular analysis from West and East Maui but it is alsopresent in the former Maui Nui islands of Molokai and

Fig. 7. Maximum likelihood single tree of score ¡ln D resulting from analysis of ITS-2 using the TVM + I + ! model with following parameter values: Rmatrix D (2.5611, 3.3526, 3.8112, 0.6771, 3.3526); nucleotide frequencies: A D 0.1750, C D 0.2967, G D 0.3049, T D 0.2234; among-site rate variation withinvariant sites I D 0.5434 and gamma shape parameter # D 0.6870. Refer to Table 6 for clade support values.


Lanai. The same holds for the verecunda lineage, althoughmolecular data were also available for the Molokai popula-tion. Morphotype D was also found on Lanai, but not onMolokai. Males belonging to this morphotype were notsampled in the molecular analyses. Assignment of males tomorphotype D was made possible by the characteristicsomatic coloration observed in females and received furthersupport by the similarity of the bulb pattern of these speci-mens to H. cruciata, its putative sister taxa (Fig. 2). Finally,specimens collected in the Northwest Hawaiian Islands ofNecker and Nihoa did not Wt in any of the deWned morpho-types, although they more closely resemble verecunda andH. cruciata and C in terms of male genitalia and face orna-mentation.

3.5. Discriminant analysis of lineages

Table 9 summarizes size and male genitalia measure-ments obtained from the 240 Hawaiian Havaika specimensavailable for study, categorized into 10 groups correspond-ing to the lineages discussed above and their island popula-tions plus the Necker–Nihoa individuals. Both marginaland accumulative permutation tests (999 permutations)revealed signiWcant (p < 0.0167, after Bonferroni correction)discrimination power of the three variables (CL D carapacelength, Cy/Tb D palp cymbium/tibia ratio, Em/Te D embolous/tegulum ratio). However, discrimination

ability diVers among variables: Em/Te explains 0.720 fromthe 1.287 of the total explainable inertia, followed by Cy/Tb(0.568) and CL (0.438) last. The two Wrst canonical axes(eigenvalues 0.780 and 0.293, respectively; sum of all canon-ical eigenvalues 1.287) accounted for 12% of the cumulativepercentage variance of species data. Permutation tests ofthe Wrst canonical axis and all canonical axes combinedrevealed signiWcant diVerences among groups (lineages/islands) (Wrst axis: F-ratio D 8.449, p D 0.0010; trace: F-ratio D 4.951, p D 0.0010). The ability of the measurementsto discriminate among pubens, verecunda, and morphotypeD + H. cruciata lineages is clearly seen in a biplot with thelineages/island groups and the measured variables of thetwo Wrst discriminant axes (Fig. 8A). The plots in Fig. 8 alsoillustrate the similarity of genitalic characters of the indi-viduals from Necker–Nihoa to verecunda, although the Wrsttend to be of bigger size. Conversely, a scatter plot with allthe individuals analyzed and enveloped by lineage/island(Fig. 8B) shows that the diVerent island populations arehardly distinguishable based on the measurements consid-ered and that the same hold for the H. cruciata and mor-photype D. Interestingly, the very few cases whereindividuals from a particular lineage had canonical scoresthat overlapped with other lineages always corresponded todiVerent island populations. In the only example of coexis-tence of the three lineages, Maui Nui, the diVerent popula-tions were clearly very distinct.

3.6. Size variation

A plot of the mean and standard deviation of carapacelength in each island population is shown in Fig. 9. Resultsof the two-way analyses of variance are summarized inTable 10. Both the overall comparison and comparisonswithin islands for which male and female specimens wereavailable show that diVerence in size between sexes and theinteraction between lineage and sex were not signiWcant.However, diVerences between lineages were clearly signiW-cant in most of the comparisons performed. On Kauai,Molokai, and West Maui, where both sexes were not avail-able for one of the lineages, t tests showed signiWcant diVer-ences in all cases between the two lineages within islands(Kauai: t D 5.99, df D 67, p < 0.0001; Molokai: t D 3.96,df D 21, p D 0.0007; West Maui: t D 4.47, df D 9, p D 0.0016).The only situations where diVerences between lineages werenot signiWcant were those islands with three lineages. OnMaui Nui, pubens was signiWcantly larger than verecundaand morphotype D, but there was no signiWcant diVerencebetween verecunda and morphotype D. The same holdswhen the analysis was restricted to Lanai and East Maui,the only Maui Nui islands where the three lineages havebeen found. Interestingly, comparisons between island pop-ulations within lineages using one-way ANOVA test (Table11) reveal that individuals of both pubens and verecundafrom Oahu are signiWcantly smaller than their counterpartsfrom Maui Nui. None of the remaining comparisons weresigniWcant.

Table 6Clade support values of all the diVerent analyses performed on staticalignment of the ITS-2 data set

Values refer to (i) jackknife support in parsimony analyses; (ii) bootstrapproportions for maximum likelihood analysis; and (iii) posterior probabil-ities for Bayesian inference. Clade numbers as shown in Fig. 7. —: cladesupported below threshold value (50% for jackknife and bootstrap, 95%posterior probability); NS: clade not supported; 5th: parsimony analyseswith gaps as 5th state; A/P: parsimony analysis with gasp recoded asabsence/presence data; ?: parsimony analysis with gaps as missing data;ML: maximum likelihood analysis; BI: Bayesian inference analysis.

Clades Clade description 5th A/P ? ML BI

1 Hawaiian Havaika 100 100 100 100 1002 pubens 61 87 75 75 1003 pubens Big Island 82 85 60 NS —4 NS NS NS NS —5 NS NS NS — NS6 NS — — NS —7 NS — — NS —8 57 56 55 83 969 — — 53 75 97

10 NS — — 52 —11 — — — 68 —12 D + H. cruciata 100 100 96 97 10013 verecunda — — — — —14 verecunda Kauai — — — 71 —15 83 85 — 88 —16 — — 52 54 —17 — 57 — 78 —18 verecunda Maui Nui 68 58 62 53 —19 — — — 61 —

H. cruciata Big Island — — NS — NSverecunda + D + H. cruciata 69 58 — — —


4. Discussion

4.1. The Marquesan species

In the present study, the two Marquesan species werealways shown as sister taxa and consistently grouped withthe continental Habronattus, in some cases as sister taxa,and the representatives of Pellenes. These results support

the current taxonomic position of H. rufescens (recentlytransferred from Sandalodes Keyserling, 1883, to whichHavaika species were originally assigned, to Habronattus).However, H. rufescens lacks any of the synapomorphies ofHabronattus, including the presence of an elbowed tegularapophysis (TA), which makes its current taxonomic posi-tion questionable. Indeed, lack of a TA seems to be a diag-nostic character for Havaika and is used to distinguish the

Table 7Clade support values of analyses performed on the combined mitochondrial and nuclear data sets, and clade age estimates (see text for details)

Support values refer to jackknife in direct optimization (212, gap opening 2, gap extension 1, and tv/ts ratio 1) and static parsimony (5th: gaps as 5th state;A/P: gaps as absence/presence; ?: gaps as missing data;), bootstrap in maximum likelihood (ML) and posterior probabilities in Bayesian Inference (BI).Estimated age (age) and inferred conWdence intervals (min: minimum, max: maximum) obtained when Wxing alternatively Node 15 or Node 8 age to be 0.5my. In all cases, Marquesan species divergence was constrained to a minimum age of 2.01 my and a maximum age 4 my. Clade numbers as shown in Fig. 5.—: clade supported below threshold value (50% for jackknife and bootstrap, 95% posterior probability); NS: clade not supported.

Clades Clade description 212 5th A/P ? ML BI Node 15 D 0.5 my Node 8 D 0.5 my

Age Min Max Age Min Max

0 Marquesas — 2.01 4.00 — 2.01 4.001 Hawaiian Havaika 100 100 100 100 100 100 3.78 3.68 4.04 2.71 2.63 2.902 pubens lineage 100 100 99 100 — 100 2.11 2.07 2.24 1.28 1.24 1.353 0.20 0.20 0.22 0.13 0.13 0.144 — — — — — — 2.08 1.95 2.12 1.18 1.11 1.235 pubens Oahu 100 100 100 100 100 100 0.44 0.42 0.46 0.27 0.26 0.286 93 97 97 97 95 100 0.04 0.03 0.04 0.03 0.02 0.037 94 99 99 99 100 100 0.14 0.14 0.17 0.09 0.09 0.118 — — — — 57 — 1.61 1.57 1.71 0.50 — —9 pubens Big Island 100 100 100 100 100 100 0.13 0.12 0.15 0.06 0.06 0.07

10 — — — — 82 — 0.06 0.05 0.06 0.03 0.03 0.0311 pubens Maui Nui — — — — — 99 1.23 1.21 1.33 0.43 0.42 0.4412 93 92 89 91 99 100 0.96 0.94 1.03 0.36 0.36 0.3813 100 100 100 100 100 100 0.05 0.04 0.05 0.03 0.02 0.0214 verecunda + D + H. cruciata 95 99 93 92 — 100 2.48 2.42 2.63 1.79 1.74 1.9115 D + H. cruciata 98 100 100 100 100 100 0.50 — — 0.62 0.58 0.6416 H. cruciata 99 100 100 100 100 97 0.15 0.14 0.15 0.15 0.14 0.1617 verecunda lineage 98 99 98 99 — 100 1.46 1.31 1.42 1.04 0.93 1.0118 verecunda Maui Nui 99 100 99 100 100 100 0.61 0.57 0.64 0.43 0.40 0.4519 98 99 100 99 100 100 0.33 0.22 0.25 0.17 0.15 0.1720 — — — — — NS 1.35 1.31 1.42 0.96 0.93 1.0121 verecunda Kauai 100 100 100 100 95 100 0.42 0.40 0.45 0.29 0.28 0.3122 100 100 100 100 96 100 0.42 0.41 0.46 0.30 0.29 0.3223 verecunda Oahu 100 100 100 100 80 100 0.19 0.18 0.20 0.13 0.12 0.1424 69 71 54 71 76 — 0.26 0.26 0.30 0.18 0.18 0.2125 100 100 100 100 100 100 0.01 0.01 0.01 0.01 0.01 0.01

Table 8Average corrected (K81uf + I + G model) genetic distances between and within (diagonal) Marquesan representatives and island populations of eachHawaiian Havaika lineage

Mitochondrial and nuclear data sets combined, except *: only mitochondrial data available.

Marquesas pubens D Maui H. cruciata Big Island verecunda

Kauai Oahu W Oahu K Maui Big Island Kauai Oahu Maui N

Marquesas 0.068 0.19 0.196 0.196 0.194 0.199 0.16 0.203 0.185 0.197 0.190pubens Kauai 0.009* 0.082 0.079 0.075 0.069 0.11 0.128 0.154 0.152 0.145pubens Oahu W 0.009 0.024 0.073 0.074 0.1 0.124 0.146 0.121 0.126pubens Oahu K 0.006 0.07 0.067 0.095 0.118 0.148 0.125 0.124pubens Maui 0.039 0.056 0.101 0.119 0.142 0.128 0.123pubens Big Island 0.004 0.105 0.12 0.136 0.124 0.116D Maui — 0.022 0.081 0.072 0.071H. cruciata Big Island 0.017 0.095 0.094 0.080verecunda Kauai 0.021 0.067 0.059verecunda Oahu 0.012 0.048verecunda Maui N 0.014


genus from Habronattus (Prószyjski, 2002). The secondMarquesan species included in this analysis does have aTA, although it is not elbowed and is much thinner than incontinental Habronattus. A close comparison of the bulbsof the two Marquesan species analyzed (Fig. 2) shows themto be almost identical except for the presence of the TA andthe shape of the embolus tip. The absence of the TA in H.rufescens reveals that the embolus originates not on the lat-eral margin of the tegulum, as in the Hawaiian species, butcloser to its mid-part. A detailed examination of the onlyavailable drawing of the bulb of a Marquesan Havaika(Berland, 1933) shows clearly the embolus originating noton the lateral but towards the basal mid-part of the tegu-lum. If this structure reXects phylogenetic aYnity, then H.rufescens and the new species are likely to be the same evo-lutionary lineage as the Marquesan Havaika and can thusbe used as representatives for the phylogenetic position ofthis lineage. Genetic divergence between H. rufescens andthe undescribed species is about the same range reportedamong island populations of lineages in the HawaiianIslands (Table 8), which suggests that both Marquesan spe-cies originated in situ and are consequently endemic tothese islands.

The relationship between the Marquesan and theHawaiian species and, therefore, the monophyly of Havaikaas currently deWned deserves further consideration. Asstated above, all analyses agree in separating Marquesanfrom Hawaiian species. However, clades supporting thissplit are poorly supported by the data (e.g. Fig. 6, nodes 12and 15), and the trees obtained are not signiWcantly betterthan topologies resulting from constraining the monophylyof the Marquesan and Hawaiian representatives. However,

the observation that the TA has been independently lost inthe Marquesas suggests that a simple bulb is not a synapo-morphy of the endemics of both archipelagos. More datawill be required to reject completely the inclusion of theMarquesan species in the genus Havaika, although the dataat hand seem to point towards this direction.

4.2. Species delimitation in Hawaiian Havaika

Taxonomic delimitation of species within Havaika hasbeen hampered by the lack of clear-cut diagnostic charac-ters. Havaika specimens do not present many variable char-acters and the few available tend to be polymorphic. Thelability of some characters used in the original descriptionsof Havaika species was largely overlooked because of thereduced number of specimens available for examination(Simon, 1900). In this study, we circumvent some of thepast limitations by combining the use of molecular andmorphological information for a large sample of individu-als in a statistical framework. Our results clearly supportthe existence in Hawaiian Havaika of at least four geneti-cally distinct lineages that can be diagnosed on the basis ofcharacters used in traditional taxonomy: the pubens lineage,the verecunda lineage, H. cruciata, and morphotype D. Aputative Wfth lineage would include the specimens fromNecker and Nihoa, although molecular data are currentlyunavailable. Additionally, mitochondrial DNA sequencedata show that the island populations of each lineage formwell-supported, reciprocally monophyletic groups withoverall genetic divergences between islands well abovewithin island variation (pubens: between-islandmean D 0.070, S.D. D 0.008, within-island mean D 0.018,

Table 9Measurements of morphological structures (in mm)

Cy/Tb: cymbium/palp tibia ratio; Em/Te: embolous origin/tegulum ratio; m: males; f: females; N: number of specimens measured; SD: standard deviation;max: maximum values reported; min: minimum value reported.

Lineage Island Sex Carapace length Cy/Tb Em/Te

N Mean SD Max Min N Mean SD Mean SD

pubens Kauai m 10 3.091 0.485 4.335 2.704 6 1.040 0.099 0.420 0.033f 0 — — — —

Oahu m 20 2.922 0.461 3.927 2.244 14 1.071 0.223 0.351 0.034f 13 2.790 0.310 3.25 2.295

Maui Nui (Molokai, Lanai, Maui) m 9 3.273 0.463 4.160 2.601 10 0.994 0.105 0.338 0.040f 15 3.191 0.427 3.570 2.805

Big Island m 13 3.025 0.492 3.984 2.193 9 1.025 0.149 0.396 0.042f 15 3.143 0.413 3.927 2.601

D Maui Nui (Lanai, Maui) m 5 2.558 0.374 2.907 2.132 5 1.355 0.104 0.360 0.035f 3 2.832 0.411 2.938 2.703

H. cruciata Big Island m 10 2.372 0.303 2.652 1.785 7 1.434 0.222 0.395 0.087f 10 2.506 0.374 3.060 2.040

verecunda Kauai m 29 2.314 0.368 2.990 1.742 23 1.535 0.229 0.520 0.050f 30 2.468 0.229 2.964 2.091

Oahu m 13 2.363 0.298 2.856 1.836 11 1.336 0.148 0.485 0.040f 18 2.669 0.206 2.856 2.448

Maui Nui (Molokai, Lanai, Maui) m 10 2.611 0.276 3.162 2.193 7 1.477 0.168 0.564 0.052f 14 2.406 0.272 3.016 2.040

?? Necker, Nihoa m 3 2.960 0.479 3.237 2.407 3 1.180 0.143 0.537 0.032f 2 2.499 0.036 2.703 2.652


S.D. D 0.026; verecunda: between-islands meanD 0.058,S.D. D 0.010, within-island mean D 0.016, S.D. D 0.005).Genetic divergence between H. cruciata and morphotype Dis the lowest among groups, in the range of within-islandvariation divergences observed in other species groups(between-H. cruciata and morphotypes D mean D 0.024,S.D. D 0.001). Still, H. cruciata haplotypes form a well-sup-

ported clade and both lineages display diagnostic morpho-logical characters.

The pubens and verecunda lineages include several nomi-nal species each. However, our data suggest that withinthese lineages geographical distribution is a better indicatorof genealogical relationships than the morphological char-acters used in the original descriptions of the species, whichare either variable within island populations (color of theocular setae, Simon, 1900) or overlap across ranges (posi-tion of the embolus relative to the bulb, Prószyjski, 2002).In any case, formal description of new species and taxo-nomic amendments are beyond the scope of this study andwill be published elsewhere.

4.3. Origin and biogeography of Hawaiian Havaika

Hawaiian Havaika shared their most recent commonancestor between 3.86 my (3.68–4.04 my) and 2.7 my (2.63–2.90 my) ago depending on the time constraint imple-mented. These values represent the minimum ages of colo-

Fig. 8. Canonical correspondence analysis. (A) Plot of the Wrst two dis-criminant axes with discriminating characteristics (CL: carapace length,Cy/Tb: ratio between the length of the cymbium and the palpal tibia, E/Te: ratio between the distance from the base of the tegulum to the base ofthe embolus and the total length of the tegulum embolous and tegulumlength ratio) and centroids of classes (lineages and island populations)shown. (B) Plot of canonical scores of each specimen on Wrst two discrimi-nating axes. Points corresponding to diVerent lineages/islands are shownwith diVerent symbols and an envelope encloses all points belonging to aparticular class.

-6 6

-34

-4 6

-46

CL

Cy/Tb

Em/Te

pubens Kauaipubens Oahupubens Maui Nuipubens HawaiiH. cruciata Hawaii

morph Necker-Nihoa

verecunda Kauaiverecunda Oahuverecunda Maui Nui

morph D Maui Nui

B

A

Fig. 9. Plot of carapace length mean and standard deviation bars for eachlineage and island population.

pubensverecundaH. cruciatamorph DNecker-Nihoa

Necker-Nihoa

Car

apac

e le

ngth

(mm

's)

HawaiiMaui NuiOahuKauai0

0.5

1

1.5

2

2.5

3

3.5

4

Table 10Two-way ANOVA for lineage and sex eVects on size (carapace length)across the Hawaiian Archipelago (All) and in each island

In column ‘Comparisons’ the lineages that did not diVer signiWcantly(SNK test) are underlined.

Island d.f. MS F p Comparisons

AllLineage 3 7.5924 53.828 <0.0001 ACDBSex 1 0.2075 1.471 0.2264Interaction 3 0.0714 0.506 0.6783Error 234 0.1410

OahuLineage 1 8931.6 45.340 <0.0001 ACSex 1 175.6 0.891 0.3489Interaction 1 150.0 0.761 0.3864Error 60 364.4

WaianaesLineage 1 0.05658 22.222 0.0002 ACSex 1 0.00244 0.958 0.3400Interaction 1 0.00530 2.083 0.1652Error 19 0.00255

KoalasLineage 1 2.9824 19.261 <0.0001 ACSex 1 0.1464 0.945 0.3369Interaction 1 0.0539 0.348 0.5584Error 39 0.1548

Maui NuiLineage 2 3.088657 30.50239 <0.0001 ACDSex 1 0.000139 0.00138 0.9705Interaction 2 0.162440 1.60420 0.2112Error 50 0.101260

Lanai + EMauiLineage 2 1.47726 14.1543 0.0002 ACDSex 1 0.00869 0.0832 0.7763Interaction 2 0.06202 0.5943 0.5624Error 18 0.10437

Big IslandLineage 1 4.838069 28.88641 <0.0001 ABSex 1 0.183742 1.09706 0.3006Interaction 1 0.000749 0.00447 0.9470Error 44 0.187486


nization of the archipelago by Havaika. At that time, Kauaiand perhaps also Oahu were the only existing high islands(Price and Clague, 2002). Although the hotspot that formedthe current Hawaiian Archipelago has been active for 80my and islands with similar features have formed in thesame location during this period (Carson and Clague,1995), recent studies have shown that only during certainperiods of time were the extant islands high and closeenough to each other for existing islands to contribute tothe biota of new islands that emerged (Price and Clague,2002). SpeciWcally, there have been two peak periods duringwhich more than one island of elevation above 1000 mexisted, the Wrst, between 18 and 8 Mya, and the second 5Mya corresponding to the current high islands. These twotime periods were also associated with periods of maximumcloseness (even contact) between islands. Conversely, therewere two time periods (27–28 and 5–7 my) when volcanoeswere at their most distantly spaced and relatively low. Thefact that one of these periods was immediately prior to theformation of the current high islands might explain whymost of the current biodiversity in the Hawaiian Islandsseems to have been originated from outside the archipelagosince the formation of Kauai (Price and Elliott-Fisk, 2004).

The time and pattern of formation of the islands is alsoconsistent with age estimates of the origin of HawaiianHavaika. However, specimens collected in the 1950s on theislets of Necker (11.0 my) and Nihoa (7.3 my) have beenclearly identiWed during the course of the present study asHavaika. These two dry volcanic rocky islands with maxi-mum areas of 0.18 and 0.57 km2, and elevations of 84 and273 m, respectively, are the remnants of islands that reachedelevations above 1000 m immediately after their formation.Now the islets are mostly barren and vegetation isrestricted to a few species of grasses and bushes that surviveon their southern slopes. The species of Havaika on theseislands have adapted to a harsher and drier environmentthan their counterparts in the present high-islands, whichmostly occur in high elevation mesic and wet forests.Unfortunately, no recent specimens of Havaika have beencollected from these islets so they could not be included inany molecular analysis. Additional data will be required todetermine whether these species, because of the older age ofNecker and Nihoa, gave rise to those on the current high

islands. If they did, this might suggest a basal position forthe Necker and Nihoa morphotype, which is most similarto verecunda lineage. However, given that the moleculardata indicate that verecunda is more distal, the morphologywould better support a more recent colonization of theislets from the younger current high islands. The tree topol-ogy as it currently stands does not allow establishing thepolarity of these characters and thus they cannot be used toascertain the phylogenetic position of the Necker andNihoa morphotype. Therefore, resolution of the correctplacement of the Necker and Nihoa species will require newmaterial from these islands for DNA analyses.

The most general biogeographic pattern observed inHawaiian taxa is what has been referred to as the “progres-sion rule,” which results from dispersal events (often associ-ated with speciation) strongly skewed in the direction fromolder to younger islands (Funk and Wagner, 1995). Thispattern is found in many groups of Hawaiian arthropods(Roderick and Gillespie, 1998). In the case of the HawaiianHavaika, progression can neither be accepted nor refuted:H. cruciata and morphotype D are found on two adjacentisland and do not provide information on the progressionpattern. Similarly, although pubens and verecunda cladesinclude taxa from four and three islands, respectively, theexistence (or lack thereof) of a progression is not obvious.However, when one considers the ages of the island popula-tions in each clade (Table 7), then the most plausible expla-nation for lack of obvious progression becomes apparent.In the pubens lineage the time of split of the Kauai popula-tion is 2.11 my (2.07–2.24 my) or 1.28 my (1.24-1.35 my),depending on the time constraint, while the separation ofOahu from the Maui Nui + Big Island clade is set to 2.08my (1.95–2.12 my) or 1.18 my (1.11–1.23 my). In the vere-cunda lineage, the Maui Nui split is estimated to be approx.1.46 my (1.31–1.42 my) or 1.04 my (0.93–1.01 my) and thedivergence of the Kauai and Oahu population about 1.35my (1.31–1.42 my) or 0.96 my (0.93–1.01 my). Age estimatesfor each time constraint clearly show that time of diver-gence of the Kauai, Oahu, and Maui Nui populations of thepubens and verecunda lineages largely overlap, which is con-sistent with a rapid dispersal through the islands by eachlineage right after their origination. This is in agreementwith the geological timeframe given that Kauai, Oahu, andMaui Nui had already emerged at the estimated time ofdiversiWcation of these clades. Indeed, among plants wheretimes of colonization of the Hawaiian Islands have beenestimated to fall within a similar time frame, the same kindof rapid colonization of multiple islands (in particular,Kauai + Oahu, sometimes also Maui) has been found, forexample, in Viola (Violaceae) (Ballard and Sytsma, 2000),Kokia (Malvaceae), and Hesperomannia (Asteraceae)(Funk and Wagner, 1995).

An additional proof for the rapid dispersion across thearchipelago of the pubens and verecunda lineages can be foundin the incongruence detected between the mitochondrial andthe ITS-2 data sets. Mitochondrial data supported mono-phyly of the island populations of each clade (Figs. 3 and 6),

Table 11One-way ANOVA for island eVect on size (carapace length) of each line-age

Morph d.f. MS F p SigniWcant comparisons

pubensBetween islands 3 2594.9 3.71 0.0143 Oahu vs. Maui NuiError 91 699.0

verecundaBetween islands 3 0.5304 6.25 0.0006 Oahu vs. Maui NuiError 115 0.0849

DBetween islands 1 0.0941 1.16 0.3227 NoneError 6 0.0811


but there were some instances of non-monophyly in ITS-2sequence types from the same specimens (Figs. 4 and 7). ITS-2is part of the tandemly repeated nuclear ribosomal DNA clus-ter. Concerted evolution tends to homogenize variants withinspecies and populations, while divergences among them areaccentuated (Hillis et al., 1991). In spite of this, intragenomicvariation has been documented in several metazoan taxa(Duran et al., 2004; Hugall et al., 1999; Van Oppen et al.,2002; Vogler and Desalle, 1994). Several factors can beresponsible for the presence of multiple copies of ITS in a sin-gle genome, including the existence of rDNA loci at diVerentgenomic locations or hybridization events. Additionally, rapidspeciation can also generate intragenomic variability if con-certed evolution is then slower than the speciation rate (Harrisand Crandall, 2000). In the present study, ITS sequences weregenerated directly from PCR products and hence intrage-nomic variation may have been overlooked. However, duringmanual editing of the sequences, we observed some casescompatible with the presence of more than one ITS copy,including double peaks at single positions or unreadablesequence fragments because of overlapping peaks. In fact,extensive intragenomic variation in the same marker has beenfound in other Hawaiian spiders of the genera Theridion andArgyrodes (Arnedo and Gillespie, in preparation). The mis-matches observed in the mitochondrial and ITS-2 topologiesof pubens and verecunda lineages could then be explained byinvoking Wxation or dominance in consensus sequence ofalternative sequence types from ancestral polymorphisms as aresult of the rapid colonization of islands. An alternativeexplanation to the apparent polyphyly of some islandsequence types would be the occurrence of hybridizationevents among islands. However, the fact that conXictingsequence types (EM130, OW111) are related to sequencetypes from islands that are not geographically contiguous ren-ders this possibility quite unlikely.

4.4. Inferred evolutionary processes

The overall pattern of evolution of Havaika corresponds toan early split into diVerent morphologies directly upon coloni-zation of the Hawaiian Islands and subsequent dispersal ofeach lineage along the island chain. The pattern of early mor-phological diversiWcation matches what has been reported forother organisms. For example, a shift to cursorial behavior inthe web-building spider genus Tetragnatha seems to haveoccurred very early in the evolutionary history of the group inthe Hawaiian Islands (Gillespie et al., 1994). Planthoppers ofthe genus Nesosydne spread from an ancestral monocotyle-don host onto a larger array of diVerent dicotyledons shortlyafter colonization of the Hawaiian Islands, and maintainedtheir host-association as they progressed down the islandchain (Roderick, 1997). Additional cases would include theOrthotylus mirids (Polhemus, 2002) and Blackburnia carabidbeetles (Liebherr and Zimmerman, 1998).

Most of the examples listed above involved ecologicalshifts associated with the exploitation of new resources. Atpresent, there is no direct Weld or experimental evidence that

the diVerent lineages of Havaika exploit diVerent resources.However, size diVerentiation has frequently been linked toecological segregation (Schoener, 1970). In all cases wheretwo or more Havaika lineages were found to coexist in a sin-gle island, one of them was always signiWcantly larger thanthe others (Fig. 9 and Table 10). The lineages diVer by a ratioof 1.25–1.30, which is about the average of the valuesreported for other sympatric species (Schluter, 2000b). Thisrelationship was always in the same direction: pubens was thelargest lineage. On the other hand, in Necker and Nihoa, theonly islands inhabited by a single lineage specimens wereobserved to be intermediate in size relative to the size of lin-eages on other islands, except Oahu where the bigger lineageis similar in size to Necker and Nihoa. Three lineages coexistin Lanai and East Maui: pubens is signiWcantly larger that theother lineages, but there was no signiWcant size diVerencebetween verecunda and morphotype D. The Lanai and EastMaui populations are the largest island populations ofpubens while mean, maximum, and minimum sizes of mor-photype D are larger (but not signiWcantly so) than thosereported for its sister group, H. cruciata, which inhabits atwo-species island. However, sizes of the verecunda lineage donot seem to be smaller than they are on other islands. In fact,verecunda individuals on Oahu are signiWcantly smaller thanthose on Maui Nui. This pattern could suggest that the sys-tem is not in equilibrium, i.e. competition is still in the processof shaping this community, as suggested by the fact that vere-cunda colonized these island recently and much later than theother lineages (about 1.42–1.31 my—these are maximum esti-mates, see Table 7 for more recent alternatives). This generalpattern of accentuation of size diVerences when species co-occur resembles that reported for other island organisms forwhich it has been demonstrated to be the result of ecologicalcharacter displacement as, for example, Cnemidophorus liz-ards from the Sea of Cortez islands (Radtkey et al., 1997),Anolis lizards in the Lesser Antilles (Losos, 1990), mustelidsin Britain and Ireland (Dayan and SimberloV, 1994), mon-gooses in the Indian and PaciWc islands, and Darwin’s Wnchesin the Galapagos (Grant, 1972). Unfortunately, the lack ofphylogenetic information on the only solitary morphotype(Necker and Nihoa) and the weak support recovered for theinternal branching patterns of each lineage of Havaika pre-cluded testing some of the predictions of the character dis-placement hypothesis, namely that (1) size diVerencesbetween species should be greater in sympatry, (2) species onone-species islands should be intermediate in size, and (3) theancestors of sympatric species occurring on one-speciesislands should be of intermediate size (Schluter, 2000a).

Our data strongly support a single origin of size diVeren-tiation early after the arrival of Havaika in the HawaiianArchipelago. This result suggests that competitive exclusionprobably explains better the current distribution of lineagesthat size assortment (Losos, 1990). Under this scenario,only those species already suYciently diVerent in body sizecould invade and coexist on an island.

The Wnding that individuals of both pubens and vere-cunda on the island of Oahu are signiWcantly smaller than


those on Maui Nui presents a conundrum. As mentionedabove, the presence of a third morphotype may explain thelarger size of pubens in Maui Nui, but not the larger size ofverecunda in the same islands. Roughgarden (1992) andRoughgarden and Pacala (1989) developed the Taxon-Cycle hypothesis as an alternative to character displace-ment to explain the pattern of body size observed in theAnolis lizards in the Lesser Antilles: On an island alreadyinhabited by a species of intermediate size, subsequent colo-nization by a larger species might lead both species toevolve toward smaller sizes, resulting in the larger speciesapproaching the optimal intermediate solitary size, with theintermediate species being pushed, Wrst to the exploitationof marginal resources, and subsequently to extinction.However, reevaluation of the Taxon-Cycle in Anolis usingphylogenetic contrasts has failed to support this trend(Miles and Dunham, 1996). A simpler explanation for suchpatterns is that they are simply artifacts resulting fromsmall sample sizes.

In any case, to test the validity of the hypothesis inHavaika, more molecular information is needed, both toidentify the phylogenetic position of the Necker and Nihoamorphotype and to resolve the internal branching patternof the pubens and verecunda lineages.

In conclusion, our study shows that the Hawaiian Hava-ika arrived relatively late in the formation of the current setof high Hawaiian Islands. At the time they colonized,Kauai, Oahu, and perhaps also some part of Maui Nui, hadalready formed. More importantly, large numbers of spi-ders had already diVerentiated and radiated within thearchipelago template, in particular Tetragnatha, togetherwith thomisid (and likely also philodromid) crab spiders,would already have dominated the aerial environment(Garb, 1999; Gillespie, 2004). The abundant ecologicalspace that would have been available to these early coloniz-ers would have been more limited for Havaika. However, itdoes appear that diVerent islands within the Hawaiianarchipelago are inhabited by Havaika species that diVer sig-niWcantly in size. Moreover, our analyses support a singleorigin of the two-size classes that is in turn linked to amarked shift in sexual secondary characters. Therefore, wemay envision a scenario whereby Havaika, upon coloniza-tion of the islands, diversiWed into two lineages character-ized by secondary sexual and male genital characters. Sizesegregation may have occurred early in the diversiWcationof the group perhaps as a result of character displacementin sympatric lineages. Subsequent colonization of theremaining islands (both older and younger) appears to havebeen achieved by each lineage independently, with sizediVerences being accentuated on each island, perhapsthrough competitive exclusion.

The interesting evolutionary patterns displayed byHavaika spiders render the genus an excellent model forstudying the role of competition in shaping ecological com-munities. This research paves the way for more detailedstudies on the evolutionary biology of these previouslyoverlooked spiders.

Acknowledgments

This study would have not been possible without theextensive work on PaciWc Salticidae and the pioneeringstudies on Hawaiian Havaika of James Berry, JosephBeatty, and Jerzy Prószyjski. Wayne Maddison providedinsightful comments for the discussion, and together withMarshall Hedin supplied unpublished (at that time) DNAsequences. Xavier Turón and Eduardo Mateos contributedvaluable expertise on statistical methods. The Nature Con-servancy of Hawaii (M. White, P. Bily), the State Depart-ment of Land and Natural Resources, the Hawaii NaturalAreas Reserve System (B. Gagne), and West Maui Landand Pineapple (R. Bartlett) provided permits and logisticsupport for collecting in their lands. Many specimens usedin the study were collected in collaboration with or directlyprovided by the following colleagues: Ingi Agnarsson,Todd Blackledge, Jessica Garb, Mandy Heddle, GustavoHormiga, and Nikolaj ScharV. Additional specimens whereobtained as a loan from the following institutions and indi-viduals: Jane Beccaloni (BNHM), Frank Howarth(BPBM), Norman Platnick (AMNH), Christine Rollard(MNHN), and James Berry. Salvador Carranza providedsuggestions on the Wrst drafts of the paper, and GustavoHormiga contributed valuable comments to the Wnal ver-sion of the paper. Funding for this research was providedby a postdoctoral grant of the Spanish Ministry of Educa-tion (M.A.), the National Science Foundation and theSchlinger Foundation (R.G.).

References

Arnedo, M.A., Oromí, P., Ribera, C., 2001. Radiation of the spider genusDysdera (Araneae, Dysderidae) in the Canary Islands: cladistic assess-ment based on multiple data sets. Cladistics 17, 313–353.

Arnedo, M.A., Coddington, J., Agnarsson, I., Gillespie, R.G., 2004. From acomb to a tree: phylogenetic relationships of the comb-footed spiders(Araneae, Theridiidae) inferred from nuclear and mitochondrial genes.Mol. Phylogenet. Evol. 31, 225–245.

Baldwin, B.G., Sanderson, M.J., 1998. Age and rate of diversiWcation of theHawaiian Silversword Alliance (Compositae). Proc. Natl. Acad. Sci.USA 95, 9402–9406.

Ballard Jr., H.E., Sytsma, K.J., 2000. Evolution and biogeography of thewoody Hawaiian violets (Viola, Violaceae): Arctic origins,herbaceous ancestry and bird dispersal. Evol. Int. J. Org. Evol. 54,1521–1532.

Barker, F.K., Lutzoni, F.M., 2002. The utility of the incongruence lengthdiVerence test. Syst. Biol. 51, 625–637.

Berland, L., 1933. Araignees Des Iles Marquises. Bull. Bernice P BishopMus. 114, 39–70.

Berland, L., 1934. Araignées De Polynesie. Ann. Soc. Entomol. Fr. 103,321–336.

Berland, L., 1942. Polynesian spiders. Occasional Pap. Bernice P BishopMus. 17, 1–24.

Berry, J.W., Beatty, J.A., Prószyjski, J., 1998. Salticidae of the PaciWcIslands. III. Distribution of seven genera with descriptions of nineteennew species and two new genera. J. Arachnol. 26, 149–189.

Blackledge, T.A., Gillespie, R.G., 2004. Convergent evolution of behaviorin an adaptive radiation of Hawaiian web-building spiders. Proc. Natl.Acad. Sci. USA 101, 16228–16233.

Brocchieri, L., 2001. Phylogenetic inferences from molecular sequences:review and critique. Theor. Popul. Biol. 59, 27–40.


Brower, A.V.Z., 1994. Rapid morphological radiation and convergenceamong races of the butterXy Heliconius erato inferred from patterns ofmitochondrial DNA evolution. Proc. Natl. Acad. Sci. USA 91, 6491–6495.

Brown, F.B.H., 1921. Origin of the Hawaiian Xora. Special Publications.Bernice P. Bishop Mus. 7, 131–142.

Buckley, T.R., Arensburger, P., Simon, C., Chambers, G.K., 2002. Com-bined data, Bayesian phylogenetics, and the origin of the New Zealandcicada genera. Syst. Biol. 51, 4–18.

Caccone, A., Milinkovitch, M.C., Sbordoni, V., Powell, J.R., 1997. Mito-chondrial DNA rates and biogeography in European newts (genusEuproctus). Syst. Biol. 46, 126–144.

Campbell, D.H., 1933. The Xora of the Hawaiian Islands. Quart. Rev. Biol.8, 164–184.

Carson, H.L., 1989. Natural hybridization between the sympatric Hawai-ian species Drosophila silvestris and Drosophila heteroneura. Evolution43, 190–203.

Carson, H.L., Clague, D.A., 1995. Geology and biogeography of theHawaiian Islands. In: Wagner, W.L., Funk, V.A. (Eds.), Hawaiian Bio-geography. Smithsonian Institution, Washington, pp. 14–29.

Carson, H.L., Sato, J.E., 1969. Micro evolution within 3 species of Hawai-ian Drosophila. Evolution 23, 493–501.

Craig, D.A., Currie, D.C., Joy, D.A., 2001. Geographical history of thecentral-western PaciWc black Xy subgenus Inseliellum (Diptera: Simu-liidae: Simulium) based on a reconstructed phylogeny of the species,hot-spot archipelagoes and hydrological considerations. J. Biogeogr.28, 1101–1127.

Cunningham, C.W., 1997. Can three incongruence tests predict when datashould be combined? Mol. Biol. Evol. 14, 733–740.

Dayan, T., SimberloV, D., 1994. Character displacement, sexual dimor-phism, and morphological variation among British and Irish mustelids.Ecology 75, 1063–1073.

Duran, S., Giribet, G., Turon, X., 2004. Phylogeographical history of thesponge crambe crambe (Porifera, Poecilosclerida): range expansionand recent invasion of the Macaronesian Islands from the Mediterra-nean Sea. Mol. Ecol. 13, 109–122.

Emerson, B.C., Kolm, N., 2005. Species diversity can drive speciation.Nature 434, 1015–1016.

Emerson, B.C., Oromi, P., Hewitt, G.M., 2000. Tracking colonization anddiversiWcation of insect lineages on islands: mitochondrial DNA phylo-geography of Tarphius canariensis (Coleoptera: Colydiidae) on theCanary Islands. Proc. R. Soc. Biol. Sci. Ser. B 267, 2199–2205.

Farris, J.S., Kallersjo, M., Kluge, A.G., Bult, C., 1994. Testing signiWcanceof incongruence. Cladistics 10, 315–320.

Fleischer, R.C., Mcintosh, C.E., Tarr, C.L., 1998. Evolution on a volcanicconveyor belt: using phylogeographic reconstructions and K–Ar-basedages of the Hawaiian Islands to estimate molecular evolutionary rates.Mol. Ecol. 7, 533–545.

Funk, V.A., Wagner, W.L., 1995. Biogeographic patterns in the HawaiianIslands. In: Wagner, W.L., Funk, V.A. (Eds.), Hawaiian Biogeography:Evolution on a Hot-Spot Archipelago. Smithsonian Institution Press,Washington, pp. 379–419.

Ganders, F.R., Berbee, M., Pirseyedi, M., 2000. Its base sequence phylog-eny in Bidens (Asteraceae): evidence for the continental relatives ofHawaiian and Marquesan Bidens. Syst. Bot. 25, 122–133.

Garb, J.E., 1999. An adaptive radiation of Hawaiian Thomisidae: biogeo-graphic and genetic evidence. J. Arachnol. 27, 71–78.

Garb, J.E., Gillespie, R.G., 2006. Island hopping across the Central PaciWc:mitochondrial DNA detects sequential colonization of the AustralIslands by crab spiders (Araneae: Thomisidae). J. Biogeogr. 33, 201–220.

Gillespie, R.G., 1999. Naivete and novel perturbations: conservation ofnative spiders on an Oceanic Island system. J. Insect Conserv. 3, 263–272.

Gillespie, R.G., 2002. Biogeography of spiders on remote Oceanic Islands ofthe PaciWc: archipelagoes as stepping stones? J. Biogeogr. 29, 655–662.

Gillespie, G., 2004. Community assembly through adaptive radiation inHawaiian spiders. Science 303, 356–359.

Gillespie, R.G., 2005. Geographical context of speciation in a radiation ofHawaiian Tetragnatha spiders (Araneae, Tetragnathidae). J. Arachnol.33, 313–322.

Gillespie, R.G., Croom, H.B., Palumbi, S.R., 1994. Multiple origins of aspider radiation in Hawaii. Proc. Natl. Acad. Sci. USA 91, 2290–2294.

Giribet, G., 2001. Exploring the behavior of Poy, a program for directoptimization of molecular data. Cladistics 17, S60–S70.

GoloboV, P.A., 1993. Nona. <www.cladistics.com/>.GoloboV, P.A., Farris, J.S., Nixon, K.C., 2003. Tnt: tree analysis using new

technology, version 1.0. Available at: <http://www.zmuc.dk/public/phylogeny/TNT/>.

Grant, P.R., 1972. Convergent and divergent character displacement. Biol.J. Linnean Soc. 4, 39–68.

Gregory, H.E., 1928. Types of PaciWc Islands. Proc. Third Pan PaciWc Sci.Congr. 2, 1662.

Gregory, J.W., 1930. The geological history of the PaciWc Ocean. Proc.Geol. Soc. 86, 72–136.

Guillaumin, A., 1928. Les Regions Floristiques Du PaciWque D’apres LeurEndemisme Et La Repartition De Quelques Plantes Phanerogames.Proc. Third Pan-PaciWc Sci. Congr. 1, 920–928.

Guindon, S., Gascuel, O., 2003. A simple, fast, and accurate algorithmto estimate large phylogenies by maximum likelihood. Syst. Biol. 52,696–704.

Harris, D.J., Crandall, K.A., 2000. Intragenomic variation within Its1 andIts2 of freshwater crayWshes (Decapoda: Cambaridae): implications forphylogenetic and microsatellite studies. Mol. Biol. Evol. 17, 284–291.

Hedin, M.C., 1997. Molecular phylogenetics at the population/speciesinterface in cave spiders of the southern appalachians (Araneae: Nes-ticidae: Nesticus). Mol. Biol. Evol. 14, 309–324.

Hedin, M.C., Maddison, W.P., 2001. A combined molecular approach tophylogeny of the jumping spider subfamily Dendryphantinae (Ara-neae: Salticidae). Mol. Phylogenet. Evol. 18, 386–403.

Hillis, D.M., Moritz, C., Porter, C.A., Baker, R.J., 1991. Evidence for biasedgene conversion in concerted evolution of ribosomal DNA. Science251, 308–310.

Hormiga, G., Arnedo, M., Gillespie, R.G., 2003. Speciation on a conveyorbelt: sequential colonization of the Hawaiian Islands by orsonwellesspiders (Araneae, Linyphiidae). Syst. Biol. 52, 70–88.

Huelsenbeck, John, P., Larget, B., SwoVord, D., 2000. A compound Pois-son process for relaxing the molecular clock. Genetics 154, 1879–1892.

Hugall, A., Stanton, J., Moritz, C., 1999. Reticulate evolution and the ori-gins of ribosomal internal transcribed spacer diversity in apomicticmeloidogyne. Mol. Biol. Evol. 16, 157–164.

Jordan, S., Simon, C., Polhemus, D., 2003. Molecular systematics andadaptive radiation of Hawaii’s endemic damselXy genus megalagrion(Odonata: Coenagrionidae). Syst. Biol. 52, 89–109.

Liebherr, J.K., Zimmerman, E.C., 1998. Cladistic analysis, phylogeny andbiogeography of the Hawaiian Platynin (Coleoptera: Carabidae). Syst.Entomol. 23, 137–172.

Losos, J.B., 1990. A phylogenetic analysis of character displacement inCaribbean Anolis Lizards. Evolution 44, 558–569.

Losos, J.B., 2000. Ecological character displacement and the study ofadaptation. Proc. Natl. Acad. Sci. USA 97, 5693–5695.

Maddison, W., Hedin, M., 2003a. Phylogeny Of Habronattus jumping spi-ders (Araneae: Salticidae), with consideration of genital and courtshipevolution. Syst. Entomol. 28, 1–22.

Maddison, W.P., Hedin, M.C., 2003b. Jumping spider phylogeny (Araneae:Salticidae). Invertebr. Syst. 17, 529–549.

Masta, S.E., Maddison, W.P., 2002. Sexual selection driving diversiWcationin jumping spiders. Proc. Natl. Acad. Sci. USA 99, 4442–4447.

Meyrick, E., 1935a. Pyrales and microlepidoptera of the MarquesasIslands. Bernice P. Bishop Mus. Bull. 114, 333–355.

Meyrick, E., 1935b. Pyrales and microlepidoptera of the Society Islands.Bernice P. Bishop Mus. Bull. 113, 109–110.

Mickevich, M.F., Farris, J.S., 1981. The implications of incongruence inMenidia. Syst. Zool. 30, 351–370.

Miles, D.B., Dunham, A.E., 1996. The paradox of the phylogeny: characterdisplacement of analyses of body size in Island Anolis. Evolution 50,594–603.

Mindell, D.P., 1991. Similarity and congruence as criteria for molecularhomology. Mol. Biol. Evol. 8, 897–900.

http://www.cladistics.com


http://www.zmuc.dk/public/phylogeny/TNT/




Nixon, K.C., 2002. Winclada, version 1.00.08. Available at: <http://www.cladistics.com>.

Nunn, P.D., 1994. Oceanic Islands. Blackwell, Oxford.Page, R.D.M., 1993. Component: Tree Comparison Software For Micro-

soft Windows. The Natural History Museum, London.Pilsbry, H.A., 1900. The genesis of mid-paciWc faunas. Proc. Acad. Nat. Sci.

Philadelphia 52, 568–581.Polhemus, D.A., 2002. An initial review of Orthotylus in the Hawaiian

Islands, with descriptions of twenty-two new species (Heteroptera:Miridae). J. NY Entomol. Soc. 110, 270–340.

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

Potvin, C., RoV, D.A., 1993. Distribution-free and robust statistical meth-ods: viable alternatives to parametric statistics? Ecology 74, 1617.

Prager, E.M., Wilson, A.C., 1988. Ancient origin of lactalbumin from lyso-zyme: analysis of DNA and amino acid sequences. J. Mol. Evol. 27,326–335.

Price, J.P., Clague, D.A., 2002. How old is the Hawaiian biota? Geologyand phylogeny suggest recent divergence. Proc. R. Soc. Biol. Sci. Ser. B269, 2429–2435.

Price, J.P., Elliott-Fisk, D., 2004. Topographic history of the Maui NuiComplex, Hawai’i, and its implications for biogeography. PaciWc Sci.58, 27–45.

Prószyjski, J., 2003. Salticidae (Araneae) of the world. <http://saltic-idaeorg/salticid/mainhtm/>.

Prószyjski, J., 2002. Remarks on Salticidae (Aranei) from Hawaii, withdescription of Havaika. Gen. Nov. Arthropoda Selecta. 10, 225–241.

Radtkey, R.R., Fallon, S.M., Case, T.J., 1997. Character displacement insome cnemidophorus lizards revisited: a phylogenetic analysis. Proc.Natl. Acad. Sci. USA 94, 9740–9745.

Roderick, G.K., 1997. Herbivorous insects and the Hawaiian silverswordalliance: co-evolution or co-speciation? PaciWc Sci. 51, 440–449.

Roderick, G.K., Gillespie, R.G., 1998. Speciation and phylogeography ofHawaiian terrestrial arthropods. Mol. Ecol. 7, 519–531. Issn: 0962-1083.

Ronquist, F., Huelsenbeck, J.P., 2003. Mrbayes 3: Bayesian phylogeneticinference under mixed models. Bioinformatics 19, 1572–1574.

Roughgarden, J., 1992. Comments on the paper by Losos: character dis-placement versus taxon loop. Copeia, 1992, 288–295.

Roughgarden, J., Pacala, S., 1989. Taxon cycle among Anolis lizard popu-lations: review of the evidence. In: Otte, D., Endler, J.A. (Eds.), Specia-tion and Its Consequences. Sinauer, Sunderland, MA, pp. 403–442.

Sanderson, M.J., 1997. A nonparametric approach to estimating diver-gence times in the absence of rate constancy. Mol. Biol. Evol. 14,1218–1231.

Sanderson, M.J., 2002. Estimating absolute rates of molecular evolutionand divergence times: a penalized likelihood approach. Mol. Biol. Evol.19, 101–109.

Sanderson, M.J., 2003. R8s: inferring absolute rates of molecular evolutionand divergence times in the absence of a molecular clock. Bioinformat-ics 19, 301–302.

Schluter, D., 2000a. Ecological character displacement in adaptive radia-tion. Am. Nat. 156, S4–S16.

Schluter, D., 2000b. The Ecology of Adaptive Radiation. Oxford Univer-sity Press, Oxford.

Schoener, T.W., 1970. Size patterns of West Indian anolis lizards. II. Corre-lations with the sizes of particular sympatric species-displacement andconvergence. Am. Nat. 104, 155–174.

Shaw, K.L., 1996. Sequential radiations and patterns of speciation in theHawaiian cricket genus Laupala inferred from DNA sequences. Evolu-tion 50, 237–255.

Shimodaira, H., Hasegawa, M., 1999. Multiple comparisons of log-likeli-hoods with applications to phylogenetic inference. Mol. Biol. Evol. 16,1114–1116.

Simmons, M.P., 2004. Independence of alignment and tree search. Mol.Phylogenet. Evol. 31, 874–879.

Simmons, M.P., Ochoterena, H., 2000. Gaps as characters in sequence-based phylogenetic analyses. Syst. Biol. 49, 369–381.

Simon, C., 1987. Hawaiian evolutionary biology: an introduction. TrendsEcol. Evol. 2, 175–178.

Simon, E., 1900. Arachnida. In: Fauna Hawaiiensis, or the Zoology of theSandwich Isles. Royal Society of London Promoting Natural Knowl-edge and the British Association for the Advancement of Science, Lon-don, pp. 443–519.

Simon, C., Frati, F., Beckenbach, A., Crespi, B., Liu, H., Flook, P., 1994.Evolution, weighting, and phylogenetic utility of mitochondrial genesequences and a compilation of conserved polymerase chain reactionprimers. Ann. Entomol. Soc. Am. 87, 651–701.

SwoVord, D.L., 2001. Paup*: Phylogenetic Analysis Using Parsimony (andOther Methods). Sinauer Associates, Sunderland, MA.

Templeton, A.R., 1983. Phylogenetic inference from restriction endonucle-ase cleavage site maps with particular reference to the humans andapes. Evolution 37, 221–244.

Thompson, J.D., Gibson, T.J., Plewniak, F., Jeanmougin, F., Higgins, D.G.,1997. The Clustal_X windows interface: Xexible strategies for multiplesequence alignment aided by quality analysis tools. Nucl. Acids Res.25, 4876–4882.

Thorne, J.L., Kishino, H., 2002. Divergence time and evolutionary rateestimation with multilocus data. Syst. Biol. 51, 689–702.

Thorne, J.L., Kishino, H., Felsenstein, J., 1992. Inching toward reality: animproved likelihood model of sequence evolution. J. Mol. Evol. 34, 3–16.

Van Oppen, M.J.H., Willis, B.L., Van Rheede, T., Miller, D.J., 2002.Spawning times, reproductive compatibilities and genetic structur-ing in the Acropora Aspera group: evidence for natural hybridiza-tion and semi-permeable species boundaries in corals. Mol. Ecol.11, 1363–1376.

Vandergast, A.G., Gillespie, R.G., Roderick, G.K., 2004. InXuence of vol-canic activity on the population genetic structure of Hawaiian tetrag-natha spiders: fragmentation, rapid population growth and thepotential for accelerated evolution. Mol. Ecol. 13, 1729–1743.

Vingron, M., Von Haeseler, A., 1997. Towards integration of multiple align-ment and phylogenetic tree construction. J. Comput. Biol. 4, 23–34.

Vitousek, P.M., Loope, L.L., Stone, C.P., 1998. Introduced species inHawaii: biological eVects and opportunities for ecological research.Trends Ecol. Evol. 2, 224–227.

Vogler, A.P., Desalle, R., 1994. Evolution and phylogenetic informationcontent of the Its-1 region in the tiger beetle Cicindela dorsalis. Mol.Biol. Evol. 11, 393–405.

Wheeler, W.C., 1995. Sequence alignment, parameter sensitivity, and thephylogenetic analysis of molecular data. Syst. Biol. 44, 321–331.

Wheeler, W.C., 1996. Optimization alignment: the end of multiplesequence alignment in phylogenetics? Cladistics 12, 1–9.

Wheeler, W.C., Hayashi, C.Y., 1998. The phylogeny of the extant chelicer-ate orders. Cladistics 14, 173–192.

Wheeler, W.C., Gatesy, J., Desalle, R., 1995. Elision: a method for accom-modating multiple molecular sequence alignments with alignment-ambiguous sites. Mol. Phylogenet. Evol. 4, 1–9.

Wheeler, W.C., Gladstein, D.S., De Laet, J.E., 1996–2003. Poy. Version 3.0.ftp.amnh.org/pub/molecular/poy (Current Version 3.0.11). Documen-tation by Daniel Janies and Ward Wheeler. Commandline documenta-tion by J. De Laet and W.C. Wheeler. Amnh, New York.

White, T.J., Bruns, T., Lee, S., Taylor, J., 1990. AmpliWcation and directsequencing of fungal ribosomal RNA genes for phylogenetics. In: Innis,M., Gelfand, D., Swinsky, J., White, T.J. (Eds.), Pcr Protocols: A Guide toMethods and Applications. Academic Press, San Diego, CA, pp. 315–322.

Wiens, J.J., 1998. Combining data sets with diVerent phylogenetic histories.Syst. Biol. 47, 568–581.

Wilson, J.T., 1963. A possible origin of the Hawaiian Islands. Can. J. Phys.41, 135–138.

Yoder, A.D., Yang, Z., 2000. Estimation of primate speciation dates usinglocal molecular clocks. Mol. Biol. Evol. 17, 1081–1090.

Young, N.D., Healy, J., 2002. Gapcoder. <http://www.trinity.edu/nyoung/gapcoder/download.html/>.

Zar, J.H., 1984. Biostatistical Analysis, 2nd ed. Prentice-Hall, EnglewoodCliVs, NJ.




http://salticidaeorg/salticid/mainhtm



ftp://ftp.amnh.org/pub/molecular/poy

ftp://ftp.amnh.org/pub/molecular/poy

http://www.trinity.edu/nyoung/gapcoder/download.html



s p ecies d iversiw ca tion pa ttern s in the p o lyn esian ju mpi n g...

Documents