author's personal copy - ez lab zrc...

This article appeared in a journal published by Elsevier. The attachedcopy is furnished to the author for internal non-commercial researchand education use, including for instruction at the authors institution

and sharing with colleagues.

Other uses, including reproduction and distribution, or selling orlicensing copies, or posting to personal, institutional or third party

websites are prohibited.

In most cases authors are permitted to post their version of thearticle (e.g. in Word or Tex form) to their personal website orinstitutional repository. Authors requiring further information

regarding Elsevier’s archiving and manuscript policies areencouraged to visit:

http://www.elsevier.com/copyright

http://www.elsevier.com/copyright

Author's personal copy

Biogeography and diversification of hermit spiders on Indian Ocean islands(Nephilidae: Nephilengys)

Matjaz Kuntner a,b,⇑, Ingi Agnarsson a,b,c

a Institute of Biology, Scientific Research Centre, Slovenian Academy of Sciences and Arts, Novi trg. 2, P.O. Box 306, SI-1001 Ljubljana, Sloveniab Department of Entomology, National Museum of Natural History, Smithsonian Institution, NHB-105, P.O. Box 37012, Washington, DC 20013-7012, USAc Department of Biology, University of Puerto Rico–Rio Piedras (UPR-RP), San Juan, PR 00931, Puerto Rico

a r t i c l e i n f o

Article history:Received 8 November 2010Revised 19 January 2011Accepted 2 February 2011Available online 21 February 2011

Keywords:MadagascarMascarenesComorosCenozoic dispersalVicariancePhylogeographyNephila

a b s t r a c t

The origin of the terrestrial biota of Madagascar and, especially, the smaller island chains of the westernIndian Ocean is relatively poorly understood. Madagascar represents a mixture of Gondwanan vicariantlineages and more recent colonizers arriving via Cenozoic dispersal, mostly from Africa. Dispersal mustexplain the biota of the smaller islands such as the Comoros and the chain of Mascarene islands, but rel-atively few studies have pinpointed the source of colonizers, which may include mainland Africa, Asia,Australasia, and Madagascar. The pantropical hermit spiders (genus Nephilengys) seem to have colonizedthe Indian Ocean island arc stretching from Comoros through Madagascar and onto Mascarenes, and thusoffer one opportunity to reveal biogeographical patterns in the Indian Ocean. We test alternative hypoth-eses on the colonization route of Nephilengys spiders in the Indian Ocean and simultaneously test the cur-rent taxonomical hypothesis using genetic and morphological data. We used mitochondrial (COI) andnuclear (ITS2) markers to examine Nephilengys phylogenetic structure with samples from Africa, south-east Asia, and the Indian Ocean islands of Madagascar, Mayotte, Réunion and Mauritius. We used Bayes-ian and parsimony methods to reconstruct phylogenies and haplotype networks, and calculated geneticdistances and fixation indices. Our results suggest an African origin of Madagascar Nephilengys via Ceno-zoic dispersal, and subsequent colonization of the Mascarene islands from Madagascar. We find strongevidence of gene flow across Madagascar and through the neighboring islands north of it, while phyloge-netic trees, haplotype networks, and fixation indices all reveal genetically isolated and divergent lineageson Mauritius and Réunion, consistent with female color morphs. These results, and the discovery of thefirst males from Réunion and Mauritius, in turn falsify the existing taxonomic hypothesis of a single wide-spread species, Nephilengys borbonica, throughout the archipelago. Instead, we diagnose three Nephilengysspecies: Nephilengys livida (Vinson, 1863) from Madagascar and Comoros, N. borbonica (Vinson, 1863)from Réunion, and Nephilengys dodo new species from Mauritius. Nephilengys followed a colonizationroute to Madagascar from Africa, and on through to the Mascarenes, where it speciated on isolatedislands. The related golden orb-weaving spiders, genus Nephila, have followed the same colonizationroute, but Nephila shows shallower divergencies, implying recent colonization, or a moderate level ofgene flow across the archipelago preventing speciation. Unlike their synanthropic congeners, N. borbonicaand N. dodo are confined to pristine island forests and their discovery calls for evaluation of their conser-vation status.

� 2011 Elsevier Inc. All rights reserved.

1. Introduction

The native terrestrial and freshwater biotas of Madagascar rep-resent a mixture of geologically old lineages with vicariant origindating back to Gondwana over 100 million years ago (mya) (Briggs,2003), along with more recent arrivals originating via Cenozoic dis-

persal from Africa, Asia, or Australasia (Yoder and Nowak, 2006;Kohler and Glaubrecht, 2010). Examples of ancient Gondwananradiations on Madagascar include boid snakes, podocnemid turtles,and iguanid lizards (Noonan and Chippindale, 2006), typhlopidblindsnakes (Vidal et al., 2010), the extinct elephant birds (Cooperet al., 2001), and cichlid and rainbow fishes (Chakrabarty, 2004;Sparks and Smith, 2004; but, see Ali and Aitchison, 2008). How-ever, existing calibrated phylogenies for vertebrates, invertebrates,and plants are only rarely compatible with the vicariant model(Yoder and Nowak, 2006). Thus, the Cenozoic model seems to ap-ply to more lineages of Madagascar fauna and flora (Yoder and

1055-7903/$ - see front matter � 2011 Elsevier Inc. All rights reserved.doi:10.1016/j.ympev.2011.02.002

⇑ Corresponding author at: Institute of Biology, Scientific Research Centre,Slovenian Academy of Sciences and Arts, Novi trg. 2, P.O. Box 306, SI-1001Ljubljana, Slovenia.

E-mail address: [email protected] (M. Kuntner).

Molecular Phylogenetics and Evolution 59 (2011) 477–488

Contents lists available at ScienceDirect

Molecular Phylogenetics and Evolution

journal homepage: www.elsevier .com/ locate /ympev


Nowak, 2006), in particular to vertebrates (Vences et al., 2001;Raxworthy et al., 2002; Hume, 2007) and especially mammals (Tat-tersall, 2006; Masters et al., 2006, 2007; Russell et al., 2008).

Cenozoic dispersal directly from Africa or via Madagascar, inturn, may explain the origin of the biota of the smaller and morerecent islands of the Indian Ocean, such as the Seychelles, Comorosand Aldabra to the north, and the chain of volcanic Mascareneislands (Réunion, Mauritius and Rodrigues) to the east (Fulleret al., 2005; Yoder and Nowak, 2006; Raxworthy et al., 2007;Le Pechon et al., 2010). Colonizers might also have come frommainland Eurasia (Dijkstra, 2007; Hume, 2007; Cumberlidge,2008), or from Australia via Indomalaya (Jonsson and Fjeldsa,2006). Finally, they may be recent arrivals brought to the islandsby humans (Vences et al., 2004). Studies on the origin of the terres-trial biotas of these islands, especially the smaller ones, are rela-tively few (e.g., Austin et al., 2004; Vences et al., 2004; Rochaet al., 2005, 2006; Hume, 2007; Raxworthy et al., 2007), yet theseislands offer a unique system to study oceanic speciation and therole of dispersal ability in the generation of biodiversity. In partic-ular, clades that are present on both the African and Asian main-land and have colonized a number of the islands in the IndianOcean offer us a chance to reveal common patterns of colonizationand processes of diversification. Although spiders have colonizedand diversified across archipelagos worldwide and have beenprominent in studies of island diversification and biogeographyin general (e.g., Arnedo et al., 2001, 2007; Gillespie and Roderick,2002; Garb and Gillespie, 2009), none of these studies have focusedon the Indian Ocean.

Here we provide a study of Indian Ocean biogeography focusingon spiders of the pantropical nephilid genus Nephilengys L. Koch,1872. Despite its ubiquity and synathropic habits, and consequentlypresumed high mobility, Nephilengys, as currently understood, is aspecies poor genus with only four, fully allopatric species worldwide(Kuntner, 2007). Within this low species diversity, however, lies anamazing variation in morphology and size (Kuntner, 2007; Kuntnerand Coddington, 2009) as well as in certain behaviors (Kuntner et al.,2009, 2010). For example, Nephilengys cruentata (Fabricius, 1775)inhabits most of tropical Africa where variation in size and colorsis tremendous, and the species has also spread into the Neotropics,where it has established permanent synanthropic colonies (Leviand von Eickstedt, 1989; Kuntner, 2007). Similarly, the Asian popu-lations all seem to belong to a single, albeit morphologically variablespecies, Nephilengys malabarensis (Walckenaer, 1841), which livessynathropically and naturally in forests (Kuntner, 2007; Kuntneret al., 2010). According to the current taxonomic hypothesis,Nephilengys populations inhabiting the islands of the western IndianOcean islands are thought to belong to only one species, Nephilengysborbonica (Vinson, 1863) (Kuntner, 2007), yet, these populationsshow a striking pattern of color variation (Fig. 1; Vinson, 1863; Dahl,1912; Kuntner, 2007). Females in Madagascar range from shades ofgray to creamy, whitish, blue, or purple, whereas those in Mauritiusare bright white and those in Réunion shiny red (Fig. 1; Kuntner,2007: Fig. 18A, D, E, F). In the absence of any available males fromthe Mascarene islands in world museum collections, Kuntner(2007) noted that the known (female) anatomical features may notbe enough to distinguish valid species that would correspond tothese color morphs (contra Vinson, 1863). This taxonomic hypothe-sis has implications both for biogeography and conservation. If thesepopulations are indeed panmictic, one is left to wonder what ecolog-ical factors spawn such geographically fixed intraspecific variation.However, failing to recognize real diversity among island popula-tions may be detrimental for the local faunas, and may lead to unde-tected extinctions. Thus, testing both biogeographical and diversitypatterns is of obvious importance.

The aim of our study was to elucidate biogeographical and diver-sification patterns of Nephilengys across the islands of the western

Indian Ocean, and simultaneously test current taxonomy, using nu-clear and mitochondrial DNA markers. We sampled the spiders inAfrica, Asia, Madagascar, Comoros, Réunion and Mauritius, and re-port on their apparent absence from Rodrigues, the most geograph-ically isolated Mascarene island. Based on phylogenetic andpopulation genetic analyses, we investigated the biogeographicand genetic patterns within this lineage and the patterns of diversi-fication across the archipelago, and discuss how this lineage com-pares with that of its sister genus, Nephila Leach, 1815, inhabitingthe same islands.

2. Material and methods

Specimens were collected in the field and fixed in 95% ethanol.Specimens of Nephilengys were collected from Mayotte, Réunion,and Mauritius and from four distant localities in Madagascar(Table 1). An expedition to Rodrigues failed to find Nephilengys spec-imens. To test the origin of Indian Ocean Nephilengys, samples of N.cruentata were obtained from southern and western Africa, in addi-tion to specimens of Nephilengys papuana (from Australia), andN. malabarensis (from southeast Asia) (Table 1). Voucher specimensare deposited at the National Museum of Natural History, Smithso-nian Institution. Two sequences were obtained from GenBank, onefor ‘‘Nephilengys sp. (EU003303), which is N. cruentata, and one forN. malabarensis (FJ607575).

DNA was isolated from each individual using one leg, with theQIAGEN DNeasy Tissue Kit (Qiagen, Inc., Valencia, CA). We ampli-fied fragments of two loci that have proven to be useful to separatespecies and populations in araneoid spiders: the mitochondrialcytochrome c oxidase subunit 1 (COI) (e.g., Agnarsson et al.,2007; Su et al., 2007; Vink et al., 2008) and the nuclear internaltranscribed spacer 2 (ITS2) (see Agnarsson, 2010 for review of itsutility in spiders). We used the LCO1490 (Folmer et al., 1994),and C1-N-2776 (Hedin and Maddison, 2001) primer pair to amplifyCOI, and the ITS-5.8S (FITS) and ITS-28S (RITS) primers (Whiteet al., 1990) for ITS2. We used standard polymerase chain reaction(PCR) and electrophoresis protocols with a PCR annealing temper-ature of 47 �C and 30 PCR cycles (e.g., Agnarsson et al., 2007, 2010;Agnarsson, 2010). The PCR products were sequenced by theSequencing and Genotyping facility of the University of PuertoRico. Sequences were submitted to GenBank (for accession num-bers, see Table 1).

Sequences were annotated using Phred and Phrap to read andassemble the chromatograms (Green, 1999; Green and Ewing,2002) using the Chromaseq package test version 0.984 (Maddisonand Maddison, 2010a) in the evolutionary analysis programMesquite 2.74 (Maddison and Maddison, 2010b). Phred was runwith default options; phrap was used with options – qual_show20 – vector_bound 0. Sequence ends were trimmed by Chromasequsing a moving window analysis: the Wrst window of 10 baseswithin which at least six were above quality score 20 was usedas the start or end of the sequence. If a site had secondary peaksat least 0.3 the height of the primary peak, it was treated as ambig-uous. Subsequently the sequences were proofread by comparingthem with the chromatograms by eye. Alignments were done usingClustalW (Thompson et al., 1994) through Mesquite, with gapopening and extension costs set at 24/6. For both COI and ITS2the alignments were unambiguous, the former with no gaps, andthe latter with only a few, unambiguously placed, and informativeindels. In the aligned matrix, an insertion at position 93 unites theMadagascar individuals, at positions 235–247 a gap is shared withall the ingroup, at 294 a gap unites all individuals from Réunion,and finally at 297 an insertion is shared among all individuals fromMauritius (both matrices available from the authors). Furtherexploration of alignment parameters was therefore not necessary.

478 M. Kuntner, I. Agnarsson / Molecular Phylogenetics and Evolution 59 (2011) 477–488


For Bayesian analyses, the appropriate substitution model wasselected with jModeltest 0.1.1 (Posada, 2008), using the AIC crite-rion (Posada and Buckley, 2004) to select among the 24 modelsimplemented in MrBayes. The best model for ITS2 was HKY + C(nst = 2, rates = gamma) and for COI it was HKY + C + I (nst = 2,rates = invgamma). Bayesian analysis of each locus separately,and the two combined, was performed using MrBayes V3.1.2(Huelsenbeck and Ronquist, 2001). Due to the low number of var-iable sites throughout each locus, the analyses were not parti-tioned by codon. The Markov chain Monte Carlo was run withfour chains for 10,000,000 generations, sampling the Markov chainevery 1000 generations. Stationarity was reached within the first1,000,000 generations, while the sample points of the first

5,000,000 generations were discarded as ‘‘burnin’’. Parsimony anal-yses were done in Mesquite with 100 heuristic search replications.

Haplotype networks were constructed in a statistical parsimonyframework using TCS (Clement et al., 2000). Population geneticstructure was calculated in Arlequin 3.5 (Excoffier et al., 2005),and uncorrected genetic distances were calculated in Mesquite(Maddison and Maddison, 2010b).

Node ages were estimated using BEAST 1.6.1. under a relaxedclock model (Drummond et al., 2006; Drummond and Rambaut,2007). We used the geological age of Réunion islands (approxi-mately 2.1 mya) as a calibration point, in effect setting the maxi-mal age of N. borbonica as a normally distributed prior withmean of 1 mya and extremes of standard deviation reaching 2.1.

Fig. 1. Diversity, sexual size dimorphism, and taxonomic status of Nephilengys species in the Indian Ocean, with study area (delimited white, corresponding to the range ofN. borbonica s.l.) and precise sampling localities in Mayotte, Madagascar, Réunion, Mauritius and Rodrigues (stars). Note several female color morphs from Madagascar, andone for each island of Réunion and Mauritius. Species names exist for most populations, except for the population from Mauritius. Note that sampling in Rodrigues failed tofind Nephilengys.

M. Kuntner, I. Agnarsson / Molecular Phylogenetics and Evolution 59 (2011) 477–488 479


Prior to BEAST analysis, taxa with over 30% missing data werepruned.

Morphological methods are described in detail in Kuntner(2007). Nomenclature of the male palpal organ follows Kuntner(2005, 2006, 2007). All measurements are reported in millimeters,and were made using an Infinity K2 long distance microscope. Pro-soma and abdomen lengths and heights were measured in lateralview, and widths in dorsal view were all measured at widestpoints. Illustrations were prepared using a Visionary Digital imag-ing system, the core components being a Canon 5D digital camerabody and a K2 Infinity microscope equipped with Olympus metal-lurgical objectives. Successive images were combined with HeliconFocus 4.0, and thereafter minimally processed with Photoshop CS3to adjust for both contrast and brightness and to remove back-

ground blemishes. For photography, anatomical preparations weretemporarily mounted in alcohol-based hand sanitizer jelly (62%ethanol), and the specimen then covered with 70% ethanol. Wedeposited type specimens at the National Museum of Natural His-tory, Smithsonian Institution, in Washington, DC and additionalvoucher specimens are lodged in the Zoological Museum, Univer-sity of Puerto Rico, Río Piedras.

3. Results

The results of the Bayesian phylogenetic analyses of bothgenes independently, as well as a combined analysis, suggest thatIndian Ocean Nephilengys are monophyletic, and that they further

Table 1Specimen data for terminals used in the phylogenetic analysis. Excluded are two terminals from GenBank (see Methods).

Taxon code Species(after this study)

Locality Source Accession nu. COI Accession nu. ITS2

K02_Nephilengys_cruentata_SAfrica Nephilengys cruentata South Africa, Ndumo Kuntner et al. (unpublished) JF459575 JF419660K108_Mayotte Nephilengys livida Mayotte Kuntner et al. (unpublished) JF459576 -K126_Nephilengys_cruentata_SierraLeone Nephilengys cruentata Sierra Leone, Freetown Kuntner et al. (unpublished) JF459577 -K50_Nephilengys_papuana Nephilengys papuana Australia, Prosperine Kuntner et al. (unpublished) JF459578 -K62_Nephilengys_cruentata_SAfrica Nephilengys cruentata South Africa, Hluhluwe Kuntner et al. (unpublished) JF459579 -K86_Nephilengys_malabarensis Nephilengys

malabarensisIndonesia, Java Kuntner et al. (unpublished) JF459580 -

K87_Nephilengys_malabarensis Nephilengysmalabarensis

Indonesia, Sulawesi Kuntner et al. (unpublished) JF459581 -

K96_Nephilengys_cruentata_Brazil Nephilengys cruentata Brazil, Rio de Janeiro Kuntner et al. (unpublished) JF459582 -NG01_Perinet Nephilengys livida Madagascar, Perinet this study JF459583 JF419661NG02_Perinet Nephilengys livida Madagascar, Perinet this study JF459584 JF419662NG03_Perinet Nephilengys livida Madagascar, Perinet this study JF459585 -NG04_Perinet Nephilengys livida Madagascar, Perinet this study JF459586 JF419663NG05_Perinet Nephilengys livida Madagascar, Perinet this study JF459587 JF419664NG06_Mauritius Nephilengys dodo Mauritius, Brise Fer this study JF459588 JF419665NG07_Mauritius Nephilengys dodo Mauritius, Brise Fer this study JF459589 JF419666NG08_Mauritius Nephilengys dodo Mauritius, Brise Fer this study JF459590 JF419667NG09_Mauritius Nephilengys dodo Mauritius, Brise Fer this study - JF419668NG10_Mauritius Nephilengys dodo Mauritius, Brise Fer this study JF459591 JF419669NG11_Montagne Nephilengys livida Madagascar, Montagne

d’Ambrethis study JF459592 JF419670

NG12_Montagne Nephilengys livida Madagascar, Montagned’Ambre

this study JF459593 JF419671







NG16_Reunion Nephilengys borbonica Reunion this study JF459597 JF419675NG17_Reunion Nephilengys borbonica Reunion this study JF459598 -NG18_Reunion Nephilengys borbonica Reunion this study JF459599 -NG19_Reunion Nephilengys borbonica Reunion this study JF459600 JF419676NG20_Reunion Nephilengys borbonica Reunion this study JF459601 JF419677NG52_Perinet Nephilengys livida Madagascar, Perinet this study JF459602 JF419678NG53_Perinet Nephilengys livida Madagascar, Perinet this study JF459603 JF419679NG54_Perinet Nephilengys livida Madagascar, Perinet this study JF459604 JF419680NG55_Perinet Nephilengys livida Madagascar, Perinet this study JF459605 -NG56_Perinet Nephilengys livida Madagascar, Perinet this study JF459606 -NG57_Perinet Nephilengys livida Madagascar, Perinet this study JF459607 -NG58_Perinet Nephilengys livida Madagascar, Perinet this study JF459608 JF419681NG59_Mayotte Nephilengys livida Mayotte this study JF459609 -NG60_Mayotte Nephilengys livida Mayotte this study JF459610 -NG61_Reunion Nephilengys borbonica Reunion this study JF459611 JF419682NG62_Reunion Nephilengys borbonica Reunion this study JF459612 JF419683NG63_Ambohitantely Nephilengys livida Madagascar, Ambohitantely this study JF459613 -NG64_Ambohitantely Nephilengys livida Madagascar, Ambohitantely this study JF459614 JF419684NG65_Reunion Nephilengys borbonica Reunion this study JF459615 JF419685NG66_Antananarivo Nephilengys livida Madagascar, Antananarivo this study JF459616 JF419686NG67_Antananarivo Nephilengys livida Madagascar, Antananarivo this study JF459617 JF419687NG68_Antananarivo Nephilengys livida Madagascar, Antananarivo this study JF459618 JF419688NG69_Antananarivo Nephilengys livida Madagascar, Antananarivo this study JF459619 JF419689NG70_Antananarivo Nephilengys livida Madagascar, Antananarivo this study JF459620 JF419690NG71_Reunion Nephilengys borbonica Reunion this study JF459621 JF419691



contain a monophyletic lineage uniting a clade from Réunion anda clade from Mauritius, while individuals from Mayotte and Mad-agascar are intermixed (Fig. 2 shows results from the combinedBayesian analysis, parsimony results are near-identical). TCSreconstructs haplotype networks from both CO1 and ITS2(Fig. 3) as three unconnected networks, representing a Madagas-car and Comoros network, a Réunion network, and a Mauritius‘network’ (with only a single COI haplotype in Mauritius). TheFST estimates among each of these islands (Madagascar includingMayotte) are also consistent with a complete lack of gene flow(FST = 1.0).

Maximal uncorrected genetic distance (mitochondrial CO1)between the populations in Madagascar and the Mascarene is-lands was 5.3% and between Réunion and Mauritius was 3–4%,while the maximal uncorrected genetic distance between Mada-gascar and the African N. cruentata was 7–9.5%. Results of theBEAST analysis suggest a separation of island populations byroughly 1.9 million years (95% confidence intervals with maxi-mum age of 7.4), and a separation of Madagascar and Africanlineages (the latter N. cruentata) by roughly 1–3 million years(95% confidence intervals with maximum age of 12.6).

Together, the phylogenies, haplotype networks, fixation indices,and genetic divergences are consistent with an African origin ofthese populations via Cenozoic dispersal, with a subsequent dis-persal from Madagascar to the Mascarene islands. Genetic isolationamong Madagascar, Mauritius, and Réunion populations subse-quently resulted in divergence of these populations forming threedistinct species. Given our experience and expert knowledge inencountering nephilid spiders in the field, our failure to find Neph-ilengys on Rodrigues suggests that the lineage has failed to colonizethat remote island. Nevertheless, a questionable historic record ex-ists on Rodrigues, and it cannot be ruled out that we simply over-looked them (see Appendix A for detail).

All phylogenetic and haplotype results agree that the Neph-ilengys populations inhabiting the islands of the western IndianOcean in fact represent genetically isolated evolutionary lineages.Therefore, the taxonomic implications are that the broad definitionof N. borbonica in fact contains three valid species (Appendix A).One species, Nephilengys livida (Vinson, 1863) inhabits Madagascar,Comoros and the Seychelles, N. borbonica is redefined to containonly the population on Réunion, while the third species, describedhere as Nephilengys dodo n.sp., inhabits Mauritius.

Fig. 2. Results of the Bayesian analysis based on CO1 and ITS2. Node values are posterior probabilities; these are omitted from within species where they were typically lessthan 50%. Scale shows expected substitutions per site. Annotated numbers refer to estimated ages from BEAST with maximum ages in parentheses (see Methods). Thephylogeny of the Indian Ocean Nephilengys species supports the hypothesis of African origin with a dispersal from Madagascar to more remote islands, and is consistent withthree, rather than one species inhabiting these islands (color coded). (For interpretation of the references to color in this figure legend, the reader is referred to the webversion of this article.)



4. Discussion

The terrestrial biota of the Indian Ocean contains elements bothancient (vicariant) and more recent, originating from various re-gions including Africa, Asia, and Australasia. The complex tectonichistory of the region (Ali and Aitchison, 2008) and many potentialsource landmasses mean that identifying the primary biogeo-graphical forces contributing to the biota is challenging, and thepatterns are often taxon specific. Yoder and Nowak (2006) re-viewed the origin, patterns and timing of speciation in lineagesinhabiting Madagascar. They found that data for most lineagesare inconsistent with Gondwanan vicariant origins, and ratherimply dispersal from Africa during the Cenozoic as a major bio-geographical force. We add another model clade to test the gener-ality of this pattern, and also find evidence for an African origin ofthe Madagascar fauna and a Madagascar origin of the faunas on thelesser islands of the western Indian Ocean. We were interested intesting alternative hypotheses of the biogeographic and diversifi-cation patterns of hermit spiders on the Indian Ocean islands. Pre-vious studies on the taxonomy of this genus were limited by lack ofspecimens. We obtained more specimens through focused fieldwork on the islands of Madagascar, Mayotte (as part of the Comoroisland chain), and the three islands of the Mascarene chain (Réu-nion, Mauritius and Rodrigues). We sampled Nephilengys popula-tions on all of these islands with the exception of Rodrigues,where we failed to find any specimens. Our phylogenetic and pop-ulation genetic analyses based on mitochondrial and nuclear mark-ers suggest that there is no gene flow among the Madagascar,Mauritius and Réunion populations and that the populations onMadagascar and Comoros are panmictic. Based on all of the avail-able evidence, we hereby revise the taxonomy of Nephilenygs to in-clude three species in the western Indian Ocean (see Appendix A).

The phylogenetic pattern in Nephilengys is thus consistent withthe hypothesis of an African origin of the Madagascar lineage, witha subsequent dispersal to smaller islands to the east (Réunion andMauritius) and, we predict, to the north (Comoros, Mayotte, Aldab-ra, Seychelles). The dispersal to the northern island chain, here rep-resented by specimens from Mayotte, was not followed byspeciation, as these populations show evidence of recent gene flowwith Madagascar. Further studies are needed to genetically testpopulations from the Seychelles and Aldabra, though these do showtypical Madagascar color patterns (Kuntner, pers. obs.). However,the dispersal to the eastern islands was followed by speciation oneach of the two islands. The common ancestor of Indian OceanNephilengys populations likely split from the African N. cruentataapproximately 3 mya (Fig. 2) and certainly no more than 13 mya,thus ruling out Gondwanan vicariance. After colonizing MadagacarNephilengys reached the Mascarene islands about 2 mya (and nomore than 8 mya), and following genetic isolation this lineage re-sulted in three species, one each on Madagascar + Comoros, Réu-nion and Mauritius. Even though a large range of ages is includedin the confidence intervals, this timing fits well with the age ofthe Mascarene islands, thought to be less than 8–9 mya, and thatof Réunion, the youngest in the Mascarene chain, which is believedto be of volcanic origin about 2 million years old (Herrmann et al.,2010).

The biogeographic and diversification pattern detected in Neph-ilengys represents a common theme among many disparateMadagascar taxa that suggest dispersal rather than vicariantevents, and predominantly dispersal from the African mainlandvia the Mozambique Channel. For example, Kohler and Glaubrecht(2010) inferred the origin of a Madagascar radiation of pachychilidriver snails at 3–5 mya, hinting at a Cenozoic dispersal to Madagas-car. Fuller et al. (2005) also precluded vicariance in the allodapine

Fig. 3. Haplotype networks (A, CO1; B, ITS2) within Nephilengys borbonica s.l. are consistent with one species inhabiting Madagascar and the Comoro islands (N. livida, whichalso occurs on the Seychelles), and one island endemic each on Réunion (N. borbonica) and Mauritius (N. dodo n.sp.).



bee genus Braunsapis on Madagascar, for which they inferred aerialor over water dispersal from Africa. Russell et al. (2008) recon-structed the evolutionary history of Triaenops bats, which diversi-fied in Madagascar with several independent colonizations fromAfrica. Further cases of colonization of Madagascar from Africahave been inferred in frogs (Vences et al., 2004), freshwater crabs(Cumberlidge, 2008), and other spiders (Agnarsson and Kuntner,2005; Agnarsson et al., 2010). Herrmann et al. (2010) establishedmultiple colonization events to Réunion by the nematode Pristion-chus pacificus and their beetle hosts. Le Pechon et al. (2010), andMicheneau et al. (2008) inferred multiple colonizations from Mad-agascar by the Mascarene Dombeyoideae (Malvaceae) and angrae-coid orchids, respectively. Colonization of Mascarene islands fromMadagascar is also known in chameleons, and day geckos, genusPhelsuma (Raxworthy et al., 2002, 2007; Austin et al., 2004). Phel-suma probably colonized the Mascarenes once, around 4–5 mya(Austin et al., 2004).

The inferred dispersal events that differ from the pattern we de-tected in Nephilengys fall into the categories of the origin of lin-eages in the Indian Ocean from the Eurasian mainland, or from

Australasia. For example, parrots reached the Mascarene islandsby island-hopping from India (Hume, 2007), while Jonsson andFjeldsa (2006) established that the African lineage of oscine passer-ine birds had its origin in Australia, and dispersed to Africa over theIndian Ocean. Similarly, a radiation of coastal lizards (Cryptobleph-arus) originated in Australasia and reached Madagascar by trans-oceanic dispersal, where it diversified, then colonized Africa,Comoros and Mauritius. Passive dispersal by wind from Asia westacross the Indian Ocean and onto Africa is also known in dragon-flies (Dijkstra, 2007). Finally, chameleons colonized Africa fromMadagascar via repeated oceanic dispersal (Raxworthy et al.,2002).

According to Stankiewicz et al. (2006), dispersal by wind andocean currents is much more likely from Madagascar towardsAfrica than the other way. Regardless, the Nephilengys radiationon the islands of the western Indian Ocean apparently originatedfrom Africa. Orb-weaving spiders disperse aerially by ballooningon threads of silk (Bell et al., 2005). From our results we concludethat Nephilengys spiders are moderately good dispersers able tocross the Mozambique Channel, which has continuously separated

Fig. 4. Male taxonomically diagnostic features: (A) N. livida from Madagascar, male left palp, ectal view; (B) same, mesal view; (C) N. borbonica from Réunion, male left palp,ectal view; (D) same, mesal view; (E) N. borbonica male from Réunion, habitus, lateral view; (F) N. dodo n. sp. from Mauritius, male left palp, ectal view; (G) same, mesal view.Note images except E are to same scale.



Madagascar from Africa for over 100 million years (Yoder andNowak, 2006), but that the channel is large enough of a barrierto prevent gene flow between N. livida and N. cruentata. Similarly,the bodies of ocean between Madagascar and Réunion and Mauri-tius were crossed by the common ancestor of N. borbonica and N.dodo n.sp., yet precluded subsequent episodes of gene flow. Thisis not the case with the modest patch of ocean separating Mada-gascar and Comoros, which harbor a panmictic population. Inter-estingly, the sister lineage of Nephilengys, the golden orb-weavingspiders (Nephila) have followed the same colonization route fromAfrica to Madagascar and onto the Mascarene islands, with a pop-ulation also inhabiting the most remote island Rodrigues (Kuntnerand Agnarsson, unpublished). Yet Nephila is a better disperser, asevidenced by its global distribution and presence on numerousislands worldwide (Harvey et al., 2007), and has apparently re-tained sufficient gene flow across the entire archipelago to preventspeciation (Kuntner and Agnarsson, unpublished).

Finally, there might be assisted dispersal events in these spi-ders, which is also known in other animal groups, e.g., frogs(Vences et al., 2004). At least two of Nephilengys species are knownto be synanthropic, and potentially invasive. The African species N.cruentata has thus reached the Neotropics with several well estab-lished colonies in Brazil (Kuntner, 2007), such colonization almostcertainly assisted by human traffic between Africa and SouthAmerica. Another highly synanthropic species is the Asian N. mal-abarensis (Kuntner, 2007). We found the Madagascar populationsof N. livida to also be synanthropic, although they also inhabit na-tive forests. Judging from our field data, however, the Mascarenesisland species are not synanthropic, as we only found them in pris-tine forests.

In sum, Nephilengys spiders follow a common theme in the re-gion, Cenozoic dispersal to Madagascar from Africa, and onto theMascarenes. Island colonization was followed by genetic isolationand speciation on the most geographically isolated islands

Fig. 5. Female taxonomically diagnostic features: (A) N. livida from Madagascar, epigynum, ventral view; (B) same, dorsal view; (C) N. borbonica female from Réunion,epigynum, ventral view; (D) same, dorsal view; (E) N. dodo n.sp. female from Mauritius, epigynum, ventral view; (F) same, dorsal view. Note that images are to same scale.



(Madagscar, Réunion, Mauritius). The discovery of genetically dis-tinct species occupying small fragments of remaining forest on theMascarenes has implications for conservation. It is our hope thatrecognizing two distinct and charismatic species on Réunion andMauritius (Fig. 1) can contribute to the public awareness of thethreatened native habitats on these islands. Let us not allow N. bor-bonica and N. dodo to go the way of the dodo.

Acknowledgments

This is contribution number 6 resulting from the 2008 IndianOcean expedition, funded by the Slovenian Research Agency (GrantZ1-9799-0618-07 to I. Agnarsson) and the National Science Foun-dation (Grant DEB-0516038 to T. Blackledge). Additional fundingcame from the European Community 6th Framework Programme(a Marie Curie International Reintegration Grant MIRG-CT-2005036536 to M. Kuntner), and the National Geographic Society (Grant8655-09 to the authors). We thank Tjaša Lokovšek, Yadira Ortiz-Ruiz, Jason Rauscher, Peter Trontelj and Miquel Arnedo for lab helpor advice, Heine C. Kiesbüy and Dennis Hansen for field assistance,and Dominique Strasberg, Sonia Ribes-Beaudemoulin, BenoitLequettefor, and MICET for logistical help. We specially thankChristine Griffith for collecting the unknown N. dodo males. MikeRix and an anonymous reviewer provided valuable feedback.

Appendix A

Revised taxonomy of Nephilengys from the Indian Ocean islands.For works cited in synonymies, see Kuntner (2007).

A.1. Nephilidae Simon

Remarks: For classification of Nephilidae see Kuntner (2006)and Kuntner et al. (2008).

A.2. Nephilengys L. Koch, 1872

Remarks: For diagnosis, circumscription and taxonomic history,see Kuntner (2007).

Composition: With this contribution, the genus contains six spe-cies: N. cruentata (Fabricius, 1775) from Africa and South America,N. livida (Vinson, 1863) from Madagascar, Comoros, Aldabra andSeychelles, N. borbonica (Vinson, 1863) from Réunion, N. dodonew species from Mauritius, the type species N. malabarensis(Walckenaer, 1841) from south and southeast Asia, and N. papuanaThorell, 1881 from New Guinea and Australia.

Monophyly: Recent phylogenetic hypotheses based on morpho-logical and behavioral data support the monophyly of Nephilengys

(Kuntner, 2005, 2006, 2007; Kuntner et al., 2008; Kuntner andCoddington, 2009). However, preliminary molecular phylogeniessuggest that the Australasian and African plus Indian Ocean speciesmay represent two independent lineages (Kuntner et al., unpub-lished data). If so, the species treated below will have to be as-signed to a different genus because the Asian N. malabarensis isthe type species.

A.3. Nephilengys livida (Vinson, 1863)

Epeira livida Vinson, 1863: 175, 310, pl. 14, f. 1, description offemale (from Madagascar); type(s) not found, presumed lostNephilengys cruentata livida: Dahl, 1912: 48; Roewer, 1942: 933Nephilengys borbonica livida: Benoit, 1963: 368; Benoit, 1964:312; Schmidt & Jocqué, 1986: 209, f. 7–9, description of maleand femaleNephilengys cruentata: Saaristo, 1978: 120, f. 211–223, descrip-tion of male and female, misidentification; Roberts, 1983: 284,285, f. 222–224, misidentificationNephilengys borbonica (in part): Kuntner, 2007; Kuntner et al.,2008.

Remarks: Kuntner (2007) treated this species as synonymouswith N. borbonica. We here remove it from this synonymy.

Taxonomic history. See Kuntner (2007).Diagnosis: Females with a creamy, brown, blue, or purple abdo-

men (Fig. 1). Female ventral epigynal area is larger (2.5 mm wide)than in N. borbonica and N. dodo (Fig. 5A, C, E). Compared with N.borbonica and N. dodo, the male palp is larger, the embolic conduc-tor exhibits a wider gap between the distal twist and the subdistalarch, and the tip of the embolic conductor, in ectal view, is at 6o’clock (Fig. 4A and B, compare with 4B and C, F and G; see alsoKuntner, 2007: Fig. 14A and B). The species can also be diagnosedbased on a number of unique combinations of COI and ITS2 nucle-otide substitutions (see Fig. 6).

DNA barcode: Matrices of both genes, including DNA barcodeswill be submitted to Dryad (http://datadryad.org/).

Description and variation: Female (ng93 from Ranomafana, Mad-agascar; Kuntner, 2007: Figs. 13, 15–16, 18C and D): Total length23.6. Prosoma 9.3 long, 6.3 wide, 5.6 high at head region; red-brown. Chelicerae dark red-brown. Sternum light red, with darkbrown edge. Labium, maxillae dark brown, with distal white edge.Ventral, frontal femora 1 and 2 with numerous short stout spines.Opisthosoma 16.7 long, 11.7 wide, 9.9 high. Dorsum creamy brown(but, see variation), venter dark brown. Epigynum (Kuntner, 2007:Figs. 13, 16C and D) with prominent lateral sclerotized chambersand copulatory openings opening anteriorly. Tubular copulatoryducts connect to spermathecae in their posterior part. Spermathe-

Fig. 6. Genetically diagnostic features, in part: typical CO1 haplotypes (above) and ITS2 haplotypes (below). Both genes show highly consistent patterns of DNA substitutionsacross islands.



cae juxtaposed. Prosoma length ranges from 7.4 to 10.2; totallength from 15.5 to 23.6 (n = 10). Sternum color varies from darkred (Kuntner, 2007: Fig. 18B) to orange and white, or can be cen-trally dark (Kuntner, 2007: Fig. 18C). Abdomen color in live ani-mals varies from dark gray to brown (Kuntner, 2007: Fig. 18Cand D) to purple (Kuntner, 2007: Fig. 18A and B).

Male (ng93/m1 from Ranomafana, Madagascar; Kuntner, 2007:Figs. 14, 17): Total length 3.5. Prosoma 1.9 long, 1.4 wide, 1.0 high;yellow-brown. Sternum yellow-brown, darker laterally. Opisthoso-ma 2.1 long, 1.4 wide, 1.0 high. Dorsum gray, with white pigmentdots and dark brown central area, venter dark gray with two pairsof white dots, laterally with black longitudinal stripes on lateralintegument folds. Pedipalp as diagnosed (Kuntner, 2007: Fig. 14).Prosoma length ranges from 1.9 to 2.5; total length from 3.1 to4.9 (n = 10).

Distribution: Madagascar, Comoros, Seychelle Islands, AldabraAtoll.

A.4. Nephilengys borbonica (Vinson, 1863)

Epeira borbonica Vinson, 1863: 170, 309, pl. 4, f. 1, description offemale (from Réunion); type(s) not found, presumed lost.

Nephilengys cruentata borbonica: Dahl, 1912: 48; Roewer, 1942:933Nephila borbonica: Bonnet, 1958: 3067Nephilengys borbonica: Benoit, 1963: 369; Kuntner, 2007; Kunt-ner, Agnarsson & Gregoric, 2009: 266, Fig. 2A and B.Nephilengys borbonica borbonica: Benoit, 1964: 312Taxonomic history: See Kuntner (2007).

Diagnosis: Females with a red abdomen ranging from strikingbright red (typical, Fig. 1) to whitish-red (e.g., Kuntner et al.,2009: Fig. 1). Female ventral epigynal area is substantially smaller(1.6 mm wide) than in N. livida and N. dodo (Fig. 5A, C, E). Com-pared with N. livida and N. dodo, the male palp is much smaller,the embolic conductor exhibits a narrower gap between the distaltwist and the subdistal arch, and the tip of the embolic conductor,in ectal view, is at 5 o’clock (Fig. 4C and D compare with 4A and B, Fand G, and with Kuntner, 2007: Fig. 14A and B). The species canalso be diagnosed based on a number of unique combinations ofCOI and ITS2 nucleotide substitutions (see Fig. 6).


Description and variation: Female (ng759 from Colorado, Réu-nion): Total length 21.8. Prosoma 9.2 long, 6.3 wide, 6.4 high athead region; red-brown. Chelicerae dark red-brown. Sternum darkbrown, with a narrow median light band. Opisthosoma 13.2 long,10.7 wide, 10.2 high, entirely bright red (but, see variation). Epigy-num as diagnosed. Prosoma length ranges from 5.6 to 9.2; totallength from 14.1 to 21.8 (n = 4). Abdomen color in smaller femaleswhite and red, but bright red in larger ones (these colors disappearin old museum animals).

Male (ng768 from Réunion): Total length 3.8. Prosoma 2.0 long,1.4 wide, 1.2 high; yellow-brown. Opisthosoma 2.0 long, 1.5 wide,1.2 high. Dorsum gray, with white pigment dots and dark browncentral pair of dots, venter dark gray. Pedipalp as diagnosed. Pro-soma length ranges from 2.0 to 3.0; total length from 3.8 to 6.1(n = 2).

Distribution: La Réunion (France).

A.5. Nephilengys dodo new species

Nephilengys borbonica (in part): Kuntner, 2007; Kuntner et al.,2008.

Types: Female holotype and paratype in USNM (National Mu-seum of Natural History, Smithsonian Institution) from MAURI-TIUS: Brise Fer Station at S20.37628 E57.44340. 20-APR-08Kuntner, Agnarsson, et al.

Remarks: Kuntner (2007) treated this species as synonymouswith N. borbonica. We here remove it from this synonymy.

Etymology: Named after the vernacular of the extinct flightlessbird from Mauritius, the dodo (Raphus cucullatus). These two spe-cies once shared their habitat, the increasingly rare native forestsin Mauritius. The specific name, a noun in apposition, is meant toincrease awareness of the need for urgent conservation of the Mau-ritius biota.

Diagnosis: Females with a strikingly white abdomen (Fig. 1). Fe-male ventral epigynal area is of intermediate size between N. lividaand N. borbonica (Fig. 5A, C, E). Compared with N. livida and N. bor-bonica, the male palp is of intermediate size, the embolic conductorexhibits a moderate gap between the distal twist and the subdistalarch, and the tip of the embolic conductor, in ectal view, is at 7o’clock (Fig. 4F and G compare with 4A and B, C and D, and withKuntner, 2007: Fig. 14A and B). The species can also be diagnosedbased on a number of unique combinations of COI and ITS2 nucle-otide substitutions (see Fig. 6).


Description and variation: Female (ng758 from Brise Fer, Mauri-tius): Total length 22.6. Prosoma 8.8 long, 6.7 wide, 5.6 high at headregion. Prosoma, chelicerae and sternum dark red. Opisthosoma 14.9long, 11.1 wide, 9.5 high, white. Epigynum as diagnosed. Prosomalength ranges from 8.5 to 8.8; total length from 22.6 to 23.4 (n = 2).

Male (from MAURITIUS: Black River Gorges National Park, campat S20.37962 E57.44750, coll. by C. Griffith): Total length 6.6. Pro-soma 3.6 long, 2.4 wide, 1.8 high; yellow-brown. Opisthosoma 4.1long, 2.4 wide, 1.7 high. Dorsum gray, with white pigment dotsand dark brown central dots, venter gray. Pedipalp as diagnosed.Prosoma length ranges from 2.5 to 3.6; total length from 4.6 to6.6 (n = 3 from the same locality).

Distribution: Mauritius.

A.6. Nephilengys instigans (Butler, 1876), nomen dubium

Nephila instigans Butler, 1876: 442, description of female (fromRodrigues); female and juvenile syntype (ng269) in BMNH, labeled‘‘76.13, Nephila cruentata Fabr., Rodriguez, G. Guliver (c). Transit ofVenus Exped. 1874–1875. [Nephila instigans Butler Cotypes]. Hirstrevised 1876’’, examined.

Nephila instigans: Butler, 1878: 504, pl. 52, f. 10, redescription offemale.

Remarks: Kuntner (2007) treated this species as synonymouswith N. borbonica. As we here redefine the taxonomy of N. borbo-nica sensu Kuntner (2007), the status of this name is in question.We cannot, based the single female type from the 19th century,diagnose and revalidate the species. Further, we failed to findany Nephilengys in Rodrigues. It is possible that we simply over-looked them and a population persists in remaining forest frag-ments, similar to N. dodo in Mauritius. Should a population inRodrigues eventually be discovered, the name N. instigans is avail-able. However, since Butler’s types were collected during a 19thcentury expedition to various islands (see above), and given thisis the only putative record of Nephilengys from Rodrigues, we findit plausible that the specimens were mislabelled and actually camefrom another island.

References

Agnarsson, I., 2010. The utility of ITS2 in spider phylogenetics: notes on prior workand an example from Anelosimus. J. Arachnol. 38, 377–382.



Agnarsson, I., Avilés, L., Maddison, W.P., 2010. Complete separation along matrilinesin a social spider metapopulation inferred from hypervariable mitochondrialDNA region. Mol. Ecol. 19, 3052–3063.

Agnarsson, I., Kuntner, M., 2005. Madagascar: an unexpected hotspot of socialAnelosimus spider diversity (Araneae: Theridiidae). Syst. Entomol. 30, 575–592.

Agnarsson, I., Kuntner, M., Coddington, J.A., Blackledge, T.A., 2010. Shiftingcontinents, not behaviors: independent colonization of solitary and subsocialAnelosimus spider lineages on Madagascar (Araneae, Theridiidae). Zool. Scr. 39,75–87.

Agnarsson, I., Maddison, W.P., Avilés, L., 2007. The phylogeny of the socialAnelosimus spiders (Araneae: Theridiidae) inferred from six molecular loci andmorphology. Mol. Phylogenet. Evol. 43, 833–851.

Ali, J.R., Aitchison, J.C., 2008. Gondwana to Asia: Plate tectonics, paleogeography andthe biological connectivity of the Indian sub-continent from the Middle Jurassicthrough latest Eocene (166-35 Ma). Earth-Sci. Rev. 88, 145–166.

Arnedo, M.A., Oromi, P., Ribera, C., 2001. Radiation of the spider genus Dysdera(Araneae, Dysderidae) in the Canary Islands: cladistic assessment based onmultiple data sets. Cladistics 17, 313–353.

Arnedo, M.A., Agnarsson, I., Gillespie, R.G., 2007. Molecular insights into thephylogenetic structure of the spider genus Theridion (Araneae, Theridiidae) andthe origin of the Hawaiian Theridion -like fauna. Zool. Scr. 36, 337–352.

Austin, J.J., Arnold, E.N., Jones, C.G., 2004. Reconstructing an island radiation usingancient and recent DNA: the extinct and living day geckos (Phelsuma) of theMascarene islands. Mol. Phylogenet. Evol. 31, 109–122.

Bell, J.R., Bohan, D.A., Shaw, E.M., Weyman, G.S., 2005. Ballooning dispersal using silk:world fauna, phylogenies, genetics and models. Bull. Entomol. Res. 95, 69–114.

Briggs, J.C., 2003. The biogeographic and tectonic history of India. J. Biogeogr. 30,381–388.

Chakrabarty, P., 2004. Cichlid biogeography: comment and review. Fish Fish. 5, 97–119.

Clement, M., Posada, D., Crandall, K.A., 2000. TCS: a computer program to estimategene genealogies. Mol. Ecol. 9, 1657–1659.

Cooper, A., Lalueza-Fox, C., Anderson, S., Rambaut, A., Austin, J., Ward, R., 2001.Complete mitochondrial genome sequences of two extinct moas clarify ratiteevolution. Nature 409, 704–707.

Cumberlidge, N., 2008. Insular species of Afrotropical freshwater crabs (Crustacea:Decapoda: Brachyura: Potamonautidae and Potamidae) with special referenceto Madagascar and the Seychelles. Contrib. Zool. 77, 71–81.

Dahl, F., 1912. Seidenspinne und Spinnenseide. Mitt. Mus. Naturkunde Berl. 6, 1–90.Dijkstra, K.D.B., 2007. Gone with the wind: westward dispersal across the Indian

Ocean and island speciation in Hemicordulia dragonflies (Odonata: Corduliidae).Zootaxa 1438, 27–48.

Drummond, A.J., Ho, S.Y.W., Phillips, M.J., Rambaut, A., 2006. Relaxed phylogeneticsand dating with confidence. PLoS Biology 4, e88.

Drummond, A.J., Rambaut, A., 2007. BEAST: Bayesian evolutionary analysis bysampling trees. BMC Evol. Biol. 7, 214.

Excoffier, L., Laval, G., Schneider, S., 2005. Arlequin ver. 3.0: an integrated softwarepackage for population genetics data analysis. Evol. Bioinform. Online 1, 47–50.

Folmer, O., Black, M., Hoeh, W., Lutz, R., Vrijenhoek, R., 1994. DNA primers for theamplification of mitochondrial cytochrome c oxidase subunit I from diversemetazoan invertebrates. Mol. Mar. Biol. Biotechnol. 3, 294–299.

Fuller, S., Schwarz, M., Tierney, S., 2005. Phylogenetics of the allodapine bee genusBraunsapis: historical biogeography and long-range dispersal over water. J.Biogeogr. 32, 2135–2144.

Garb, J.E., Gillespie, R.G., 2009. Diversity despite dispersal: colonization history andphylogeography of Hawaiian crab spiders inferred from multilocus genetic data.Mol. Ecol. 18, 1746–1764.

Gillespie, R.G., Roderick, G.K., 2002. Arthropods on islands: colonization, speciation,and conservation. Ann. Rev. Ent. 47, 595–632.

Green, P., 1999. Phrap. <http://phrap.org/>.Green, P., Ewing, B., 2002. Phred. <http://phrap.org/>.Harvey, M.S., Austin, A.D., Adams, M., 2007. The systematics and biology of the

spider genus Nephila (Araneae: Nephilidae) in the Australasian region. Invert.Syst. 21, 407–451.

Hedin, M.C., Maddison, W.P., 2001. A combined molecular approach to phylogeny ofthe jumping spider subfamily Dendryphantinae (Araneae: Salticidae). Mol.Phylogenet. Evol. 18, 386–403.

Herrmann, M., Kienle, S., Rochat, J., Mayer, W.E., Sommer, R.J., 2010. Haplotypediversity of the nematode Pristionchus pacificus on Réunion in the Indian Oceansuggests multiple independent invasions. Biol. J. Linn. Soc. 100, 170–179.

Huelsenbeck, J.P., Ronquist, F., 2001. MrBayes: Bayesian inference of phylogenetictrees. Bioinfomatics 17, 754–755.

Hume, J.P., 2007. Reappraisal of the parrots (Aves: Psittacidae) from the MascareneIslands, with comments on their ecology, morphology, and affinities. Zootaxa1513, 3–76.

Jonsson, K.A., Fjeldsa, J., 2006. Determining biogeographical patterns of dispersaland diversification in oscine passerine birds in Australia, Southeast Asia andAfrica. J. Biogeogr. 33, 1155–1165.

Kohler, F., Glaubrecht, M., 2010. Uncovering an overlooked radiation: molecularphylogeny and biogeography of Madagascar’s endemic river snails(Caenogastropoda: Pachychilidae: Madagasikara gen. nov.). Biol. J. Linn. Soc.99, 867–894.

Kuntner, M., 2005. A revision of Herennia (Araneae, Nephilidae, Nephilinae), theAustralasian ‘coin spiders’. Invert. Syst. 19, 391–436.

Kuntner, M., 2006. Phylogenetic systematics of the Gondwanan nephilid spiderlineage Clitaetrinae (Araneae, Nephilidae). Zool. Scr. 35, 19–62.

Kuntner, M., 2007. A monograph of Nephilengys, the pantropical ‘hermit spiders’(Araneae, Nephilidae, Nephilinae). Syst. Entomol. 32, 95–135.

Kuntner, M., Agnarsson, I., Gregoric, M., 2009. Nephilid spider eunuchphenomenon induced by female or rival male aggressiveness. J. Arachnol.37, 266–271.

Kuntner, M., Coddington, J.A., 2009. Discovery of the largest orbweaving spiderspecies: the evolution of gigantism in Nephila. PLoS ONE 4, e7516.

Kuntner, M., Coddington, J.A., Hormiga, G., 2008. Phylogeny of extant nephilid orb-weaving spiders (Araneae, Nephilidae): testing morphological and ethologicalhomologies. Cladistics 24, 147–217.

Kuntner, M., Kralj-Fišer, S., Gregoric, M., 2010. Ladder webs in orb-web spiders:ontogenetic and evolutionary patterns in Nephilidae. Biol. J. Linn. Soc. 99, 849–866.

Le Pechon, T., Dubuisson, J.Y., Haevermans, T., Cruaud, C., Couloux, A., Gigord,L.D.B., 2010. Multiple colonizations from Madagascar and convergedacquisition of dioecy in the Mascarene Dombeyoideae (Malvaceae) asinferred from chloroplast and nuclear DNA sequence analyses. Ann. Bot.106, 343–357.

Levi, H.W., von Eickstedt, V.R.D., 1989. The Nephilinae spiders of the Neotropics(Araneae: Tetragnathidae). Mem. Inst. Butantan 51, 43–56.

Maddison, D.R., Maddison, W.P., 2010a. Chromaseq: a Mesquite Module forAnalyzing Sequence Chromatograms. Version 0.984. <http://mesquiteproject.org/packages/chromaseq>.

Maddison, W.P., Maddison, D.R., 2010b. Mesquite: a Modular System forEvolutionary Analysis. Version 2.74. <http://mesquiteproject.org>.

Masters, J.C., de Wit, M.J., Asher, R.J., 2006. Reconciling the origins of Africa, Indiaand Madagascar with vertebrate dispersal scenarios. Folia Primatol. 77, 399–418.

Masters, J.C., Lovegrove, B.G., de Wit, M.J., 2007. Eyes wide shut: canhypometabolism really explain the primate colonization of Madagascar? J.Biogeogr. 34, 21–37.

Micheneau, C., Carlsward, B.S., Fay, M.F., Bytebier, B., Pailler, T., Chase, M.W., 2008.Phylogenetics and biogeography of Mascarene angraecoid orchids (Vandeae,Orchidaceae). Mol. Phylogenet. Evol. 46, 908–922.

Noonan, B.P., Chippindale, P.T., 2006. Vicariant origin of Malagasy reptiles supportslate cretaceous antarctic land bridge. Am. Naturalist 168, 730–741.

Posada, D., 2008. jModelTest: Phylogenetic Model Averaging. Mol. Biol. Evol. 25,1253–1256.

Raxworthy, C.J., Forstner, M.R.J., Nussbaum, R.A., 2002. Chameleon radiation byoceanic dispersal. Nature 415, 784–787.

Raxworthy, C.J., Ingram, C.M., Rabibisoa, N., Pearson, R.G., 2007. Applications ofecological niche modeling for species delimitation: a review and empiricalevaluation using day geckos (Phelsuma) from Madagascar. Syst. Biol. 56, 907–923.

Rocha, S., Carretero, M.A., Harris, D.J., 2005. Mitochondrial DNA sequence datasuggests two independent colonizations of the Comoros archipelago byChameleons of the genus Furcifer. Belg. J. Zool. 135, 39–42.

Rocha, S., Carretero, M.A., Vences, M., Glaw, F., Harris, D.J., 2006. Decipheringpatterns of transoceanic dispersal: the evolutionary origin and biogeography ofcoastal lizards (Cryptoblepharus) in the Western Indian Ocean region. J.Biogeogr. 33, 13–22.

Russell, A.L., Goodman, S.M., Cox, M.P., 2008. Coalescent analyses support multiplemainland-to-island dispersals in the evolution of Malagasy Triaenops bats(Chiroptera: Hipposideridae). J. Biogeogr. 35, 995–1003.

Sparks, J.S., Smith, W.L., 2004. Phylogeny and biogeography of the Malagasyand Australasian rainbowfishes (Teleostei: Melanotaenioidei): Gondwananvicariance and evolution in freshwater. Mol. Phylogenet. Evol. 33, 719–734.

Stankiewicz, J., Thiart, C., Masters, J.C., de Wit, M.J., 2006. Did lemurs havesweepstake tickets? An exploration of Simpson’s model for the colonization ofMadagascar by mammals. J. Biogeogr. 33, 221–235.

Su, Y.-C., Chang, Y.-H., Lee, S.-C., Tso, I.-M., 2007. Phylogeography of the giant woodspider (Nephila pilipes, Araneae) from Asian–Australian regions. J. Biogeogr. 34,177–191.

Tattersall, I., 2006. Historical biogeography of the strepsirhine primates ofMadagascar. Folia Primatol. 77, 477–487.

Thompson, J.D., Higgins, D.G., Gibson, T.J., 1994. CLUSTAL W: improving thesensitivity of progressive multiple sequence alignments through sequenceweighting, position specific gap penalties and weight matrix choice. NucleicAcids Res. 22, 4673–4680.

Vences, M., Freyhof, J., Sonnenberg, R., Kosuch, J., Veith, M., 2001. Reconciling fossilsand molecules: cenozoic divergence of cichlid fishes and the biogeography ofMadagascar. J. Biogeogr. 28, 1091–1099.

Vences, M., Kosuch, J., Rodel, M.O., Lotters, S., Channing, A., Glaw, F., Bohme, W.,2004. Phylogeography of Ptychadena mascareniensis suggests transoceanicdispersal in a widespread African-Malagasy frog lineage. J. Biogeogr. 31, 593–601.

Vidal, N., Marin, J., Morini, M., Donnellan, S., Branch, W.R., Thomas, R., Vences, M.,Wynn, A., Cruaud, C., Hedges, S.B., 2010. Blindsnake evolutionary tree revealslong history on Gondwana. Biol. Lett. 6, 558–561.

Vink, C.J., Sirvid, P.J., Malumbres-Olarte, J., Griffiths, J.W., Paquin, P., Paterson,A.M., 2008. Species status and conservation issues of New Zealand’sendemic Latrodectus spider species (Araneae: Theridiidae). Inv. Syst. 6,589–604.

Vinson, A., 1863. Aranéides des îles de la Réunion, Maurice et Madagascar. Roret,Paris.



White, T.J., Bruns, T.D., Lee, S.B., Taylor, J.W., 1990. Amplification and directsequencing of fungal ribosomal RNA genes for phylogenetics. In: Innis, M.A.,Gefland, D.H., Sninsky, J.J., White, T.J. (Eds.), PCR Protocols: A Guide to Methodsand Applications. Academic Press Inc., San Diego, pp. 315–322.

Yoder, A.D., Nowak, M.D., 2006. Has vicariance or dispersal been the predominantbiogeographic force in Madagascar? Only time will tell. Annu. Rev. Ecol. Evol.Syst. 37, 405–431.


author's personal copy - ez lab zrc...

Documents