The dark side of an island radiation: systematics and evolution of troglobitic spiders of the genus...

Introduction

The discovery of a rich and well adapted fauna in tropical cavesin the late 1960s and 1970s (Jarrige 1957; Glover et al. 1964;Leleup and Leleup 1968; Howarth 1972; Peck 1975; Reddell1977) challenged the traditional paradigm on the origin and evo-lution of troglobitic species. Before these findings, obligatecave animals were considered relict lineages restricted to tem-perate limestone caves. These views were articulated in the‘Pleistocene effect’ or ‘climatic relict hypothesis’, which inter-preted adaptation to the underground environment as a passiveprocess chiefly driven by external factors (Barr 1967; Sbordoni1982; Barr and Holsinger 1985; Holsinger 1988; Sbordoni et al.2000). It suggested that major climatic changes, namelyPleistocene glaciations, caused extirpation of epigean sourcepopulations of cave-dwellers, promoting allopatric speciationand reinforcing cave adaptation, mainly through characterregression. The discovery of tropical troglobites prompted theproposal of the ‘adaptive shift hypothesis’ or ‘local habitat shifthypothesis’, which considered cave adaptation as an activeprocess involving an ecological shift to a completely new arrayof resources. Extinction of the epigean population was not con-sidered to be a necessary condition for speciation, given thatresource segregation and ecological requirements would have

prevented gene flow (Chapman 1982; Howarth 1987, 1991;Peck and Finston 1993). The applicability and viability of thetwo competing hypotheses is still a matter of debate (Trajano1995; Desutter-Grandcolas and Grandcolas 1996; Rivera et al.2002; Leys et al. 2003). It has been argued, however, that the cri-teria followed to distinguish between the two hypotheses arehighly questionable (Juberthie 1984) and that neither of themprovides an explanation for the colonisation of the undergroundenvironment, arguably the single most important step towardsthe evolution of hypogean life (Desutter-Grandcolas 1997b). Onthe other hand, there seems to exist a clear consensus regardingthe importance of the so-called ‘preadaptations’ or exaptations(Gould and Vrba 1982) for cave colonisation, that is, the pres-ence in a nonhypogean ancestor of character-states that promoteor facilitate exploitation of new resources in a harsh environ-ment (Desutter-Grandcolas 1997b).

Traditionally, cave-dwelling organisms have been classifiedfrom an ecological standpoint into trogloxenes (accidental cave-dwellers that cannot reproduce in caves) troglophiles (faculta-tive cave-dwellers that can either complete their life cycle insidecaves or outside), and troglobites (strict cave-dwellers thatcannot reproduce outside caves) (Schiner 1854; Racovitza

Invertebrate Systematics, 2007, 21, 623–660

10.1071/IS07015 1445-5226/07/060623© CSIRO 2007

Miquel A. ArnedoA,D, Pedro Oromí B, Cesc MúrriaC, Nuria Macías-HernándezA,B

and Carles RiberaA

ADepartament de Biologia Animal, Universitat de Barcelona, Avinguda Diagonal 645, 08028,Barcelona, Spain.

BDepartamento de Biología Animal, Universidad de La Laguna, Tenerife, Islas Canarias, Spain.CDepartament d’Ecologia, Universitat de Barcelona, Avinguda Diagonal 645, 08028, Barcelona, Spain.DCorresponding author. Email: [email protected]

Abstract. The spider genus Dysdera Latreille is an excellent model for the study of the evolution of cave life: ten speciesare known to exist exclusively in the subterranean environment of the Canary Islands, where the genus has undergone localdiversification. In the present paper, two new troglobitic species (Dysdera madai, sp. nov. and D. sibyllina, sp. nov.) andthe previously unknown sex of five additional species are described and illustrated: the males of D. gollumi Ribera &Arnedo, 1994, D. hernandezi Arnedo & Ribera, 1999 and D. labradaensis Wunderlich, 1991; and the females ofD. andamanae Arnedo & Ribera, 1997 and D. gibbifera Wunderlich, 1991. The first direct evidence of troglobitic membersof Dysdera in micro- and mesocaverns are reported. The evolution of cave life as hypothesised following a combinedmorphological and molecular phylogeny is investigated. Troglobitic Canarian Dysdera species have colonised the under-ground on eight independent occasions. The Dysderidae groundplan represents a preadaptation to cave life and has facil-itated the colonisation of caves. Canarian members of Dysdera have a predominantly parapatric mode of speciation,although postspeciation changes in distribution may have obscured allopatric processes. Eye regression and, to a lesserextent, larger body size and appendage elongation characterise troglobitic species. The different levels of troglobiomor-phism are interpreted as local adaptations to heterogeneous subterranean conditions. The high levels of sympatry amongtroglobites are explained by trophic segregation and changes in prey capture strategy were involved in the single identi-fied case of subterranean speciation in the group.

The dark side of an island radiation: systematics andevolution of troglobitic spiders of the genus DysderaLatreille (Araneae:Dysderidae) in the Canary Islands

www.publish.csiro.au/journals/is

CSIRO PUBLISHING

M. A. Arnedo et al.624 Invertebrate Systematics

1907). In many cases, however, the paucity of data precludes theassignment of taxa to these categories. Alternative definitionshave been proposed based on the morphological interpretationof the relationship between species and their environment. Theterm troglobiomorphism, originally coined as ‘troglomorphism’by Christiansen (1962) and subsequently amended by Juberthie(1984), refers to phenotypes resulting from the adaptation of anorganism to caves, which in turn characterises them.Troglobiomorphic organisms are characterised by (Christiansen1992): specialisation of the sense organs; appendage elonga-tion; reduction of the eyes, pigments, cuticle and wings;increase in egg size; overdevelopment of the tarsal claws, andadditional physiological and behavioural traits related to decel-eration of the metabolic rate and the loss of circadian and sea-sonal rhythms. The evolutionary significance of regressivecharacters has been the subject of debate between selectionists,who claim that regressive evolution is adaptive because it allowsa reduction of energy investment in useless structures, and neu-tralists, who interpret regressive characters as the byproduct ofthe absence of purifying selection (Kane and Culver 1992).

The incorporation of phylogenetic information in ecologicaland evolutionary studies firstly allows the testability of hypothe-ses regarding processes; secondly, the establishment of the polar-ity (i.e. evolutionary sequence) of the character statetransformation; and, thirdly, the distinction between phylogeneticconstraint and ecological adaptation (Miller and Wenzel 1995;Grandcolas 1997). The need to address the evolution of cave lifein a phylogenetic context has been stressed by several authors(Deeleman-Reinhold and Deeleman 1988; Peck 1990; Ashmole1993; Peck and Finston 1993; Desutter-Grandcolas 1997b), anda growing number of studies on cave-dwelling organisms incor-porate explicit phylogenetic information (Desutter-Grandcolas1993; Desutter-Grandcolas 1994; Hedin 1997a; Caccone andSbordoni 2001; Rivera et al. 2002; Leys et al. 2003; Allegrucciet al. 2005; Miller 2005; Moulds et al. 2007).

The spider family Dysderidae includes nocturnal huntersthat spend the daytime in a silk nest under stones, dead logs orbark; they are usually found in slightly damp but warm groundhabitats. Specimens of this family are frequently collected incaves and related underground environments, and several exam-ples of troglobitic species have been reported in the family(Ribera and Juberthie 1994). The largest genus of the family,Dysdera Latreille, has a circum-Mediterranean distribution andincludes some 250 species (Platnick 2006). The Macaronesianarchipelagos (Azores, Madeira, Selvagens, Canaries and CapeVerde) represent the westernmost limit of its range. The genushas undergone a process of local diversification in the CanaryIslands, a volcanic archipelago 100 km off the north-easternAfrican cost. At present, 43 endemic species are known (Arnedoet al. 1996; Arnedo and Ribera 1997; Arnedo et al. 2000;Arnedo et al. 2001), most of which evolved from a singlecolonist (Arnedo et al. 2001). Several Canarian endemic speciesof Dysdera have been collected exclusively in hypogean habi-tats, which include lava tubes, volcanic pits, networks of voidsand cracks, and the mesovoid shallow substratum (MSS, orterrestrial interstitial habitat) (Culver 2001).

The existence of closely related surface and cave-dwellingspecies combined with the experimental-like conditions ofoceanic islands for the study of evolution, render Canarian

members of Dysdera as ideal models for the study of the originand evolution of troglobites. The aim of the present study is,first, to complete the taxonomic knowledge of troglobiticCanarian members of Dysdera based on an exhaustive samplingof hypogean habitats of the Canaries, and second, to use phylo-genetic information to investigate geographical and morpho-logical patterns of colonisation and adaptation to thesubterranean environment.

Material and methodsThe use of a priori evolutionary hypotheses in the definition ofcave organisms has been heavily criticised on the basis of theindependence principle (Desutter-Grandcolas 1997a). Instead,we have chosen to use simple habitat definitions to avoid circu-lar reasoning, and hereafter we will refer to taxa exclusively col-lected in caves or the MSS as troglobitic taxa, and those collectedindistinctly in hypogean and epigean habitats as nontroglobitic orepigean taxa. Under the former definition, the followingCanarian species were considered troglobitic taxa: D. ambulo-tenta Ribera, Ferrández & Blasco, 1985; D. chioensisWunderlich, 1991; D. esquiveli Ribera & Blasco, 1986;D. gollumi Ribera & Arnedo, 1994; D. hernandezi Arnedo &Ribera, 1999; D. labradaensis Wunderlich, 1991; D. unguimma-nis Ribera, Ferrández & Blasco, 1985, and D. ratonensisWunderlich, 1991, along with two new species described below.In recent years, extensive fieldwork conducted in the under-ground environment of Tenerife, La Palma and El Hierro, hasprovided a large collection of new cave-dwelling specimens.Along with the two new species, the collection includes speci-mens of previously unknown sexes of the following species:D. andamanae Arnedo & Ribera, 1997; D. gibbifera Wunderlich,1991 (females previously unknown); and D. gollumi; D. hernan-dezi, and D. labradaensis (males previously unknown). Thedescriptions of these species are completed in the present study.

TaxonomyMorphological methods are described in detail in Arnedo andRibera (1999) and Arnedo et al. (2000). Taxonomic descriptionsfollow the format of Arnedo at al. (1999). Specimens wereexamined using a Leica MZ16A stereoscopic microscopeequipped with a Nikon DXM1200 digital camera. Digital micro-scope images were edited using the Auto-Montage softwarepackage. Digital illustrations were generated following guide-lines provided in Coleman (2003) with the assistance of aWACOM digitiser board and Adobe Illustrator 10 and Photoshop8.0.1 software. Male bulbi, tarsi and spinnerets were removed,cleaned by ultrasound and examined using either a HITACHI S-2300 Scanning Electron Microscope (SEM) (SCT, Universitat deBarcelona, Barcelona, Spain) or LEO 1340 VP SEM (GeorgeWashington University, Washington, DC, USA) operated at10–15 Kv. Left male palps were illustrated and right palps wereused for SEM images, unless otherwise stated. Measurementswere taken using an ocular measuring graticule mounted on aLeica MZ16A or a Zeiss 475022 dissection microscope. Allmorphological measurements are given in millimetres. Eye dia-meters were taken from the spans of the lens. Carapace andabdomen measurements were taken in dorsal view, abdomenhairs were measured in lateral view, and cheliceral basal segmentlength was measured in lateral view. The prolateral groove of the

Invertebrate Systematics 625

chelicera was measured in dorsal view, by positioning the che-licera parallel to the background, and measuring from the distalend of the margin to the beginning of the cheliceral lamina. Thefang was measured in ventral view, from the basal segmentcondyl to the fang distal tip. The largest leg article lengths weremeasured in lateral view without detaching the legs from thespecimen, by placing the article being measured in a perpendic-ular position. The female vulva was removed and muscle tissueswere digested using a KOH (35%) solution before observation.Leg spination was recorded using the codification method fullydescribed in Arnedo and Ribera (1997).

Abbreviations used in the text and figures are as follows:Repositories

CRBA Centre de Recerca de Biodiversitat Animal,Universitat de Barcelona, Spain

RG The personal collection of Mr Rafael GarciaSMF Senckenberg Museum, Frankfurt, Germany

UB Universitat de Barcelona, SpainULL Universidad de La Laguna, Spain

EyesAME anterior medial eyesPME posterior medial eyesPLE posterior lateral eyes

Male copulatory bulbusT tegulum

DD distal divisionIS internal sclerite

ES external scleriteDH distal haematodoca

F flagellumC crest

AC additional crestAR arch-like ridgeLF lateral fold over lateral sheet between internal and

external scleritesL lateral sheet

LA lateral sheet anterior apophysisAL additional lateral sheet at the internal border

P posterior apophysis

Female genitaliaDA dorsal archDF dorsal arch foldVS ventral sclerotisation

S spermathecaTB transversal bar

SpinneretsALS anterior lateral spinneretsPMS posterior medial spinneretsPLS posterior lateral spinneretsMS major ampulate gland spigotPS polar piriform gland spigot

Phylogeny

Taxonomic sampling

Taxa analysed in the present study are listed in Appendix 1and cave localities are mapped in Fig. 1. Taxonomic sampling

was based on a previous cladistic analysis of Canarian membersof Dysdera (Arnedo et al. 2001). Canarian species, D. hirguan(Arnedo, Oromí and Ribera, 1997) and D. minutissimaWunderlich, 1991, the continental D. caeca Ribera, 1993 andD. drescoi Ribera, 1983, as well as Dysdera, sp. nov. from theAzores, and D. vermicularis Berland, 1936 from Cape Verdewere not included in the present analyses as there was no mole-cular information available on these species. The continentalspecies, D. vivesi Ribera & Ferrández, 1986, was replaced by themorphologically similar D. edumifera Ferrández,1983 forsimilar reasons. The new species described in the present studywere incorporated into the analyses. The final matrix was com-posed of 74 terminals, including all but the two aforementionedCanarian species, 18 continental Dysdera and one endemictaxon from Madeira, along with representatives of four addi-tional Dysderidae genera. All analyses were rooted on thebranch connecting Harpactea hombergi (Scopoli, 1763) (sub-family Harpacteinae) to the remaining taxa.

Morphological charactersThe new species and information on previously unknown

sexes were coded into the phylogenetic data matrix of Arnedoet al. (2001), with the modifications in taxonomic samplingstated above. An additional state was added to character 29(state 4: IS and ES sclerites of the DD male bulb similarly devel-oped), which happened to be an autopomorphy of D. sibyllina,sp. nov. The final morphological data matrix consisted of 66parsimony informative discrete characters, with multistate char-acters treated as unordered.

Molecular charactersPublished sequences of the cytochrome c oxidase subunit I

(cox1) and 16S rRNA (rrnL) mitochondrial genes of the speciesscored in the morphological matrix were retrieved fromGenBank. New sequences were obtained for the speciesdescribed in this study, along with some newly collectedCanarian populations and four additional species for whichmolecular information was not available in Arnedo et al. (2001),using the following molecular protocols. Partial fragments ofcox1 and rrnL were amplified using the primer pairs C1-J-1490and C1-N-2198 (Folmer et al. 1994); C1-J-1718 and C1-N-2191(Simon et al. 1994) (cox1); and LR-N-13398 (Simon et al.1994), and LR-J-12864 (Arnedo et al. 2004) (rrnL). PCR con-ditions were as follows: 2 min at 94°C followed by 35 cycles ofdenaturalisation at 94°C for 30 s, annealing at 42–45°C for 35 s,and extension at 72°C for 45 s, with a final single extension stepat 72°C for 5 min. In some cases, a successful amplification wasachieved starting with five cycles at 38°C and raising thetemperature to 42°C for 30 more cycles. The PCR reaction mixcontained a final concentration of 0.2 µM of each primer,0.2 mM of each dNTPs, 1,25 U µL–1 BIOTAQ DNA polymerase(Bioline, www.bioline.com), with the supplied buffer, and1.5–2.5 mM Mg Cl2 in a final volume of 50 µL. PCR productswere checked in a 1.5% agarose gel, and subsequently purifiedusing MultiScreen 96- well filter plates from Millipore(www.millipore.com). The forward and reverse strands werecycle-sequenced using dye terminators (Sanger et al. 1977) andthe ABI PRISM Big Dye Terminator Cycle Sequencing ReadyReaction (Applied Biosystems, www.appliedbiosystems.com)

Systematics and evolution of Canarian troglobitic Dysdera


with AmpliTaq DNA Polymerase FS kit, and sequenced in anABI 3700 automated sequencer at the facilities of the Scientificand Technical Services of the University of Barcelona, Spain(www.sct.ub.es).

Phylogenetic analysesRaw sequences were analysed and edited using the Pregap andGap4 programs included in the Staden software package(http://staden.sourceforge.net/). Sequences were manipulatedand preliminary manual alignments constructed using Bioedit(Hall 1999). Alignment of cox1 was trivial since no evidence ofinsertions/deletions was observed. Ribosomal gene sequences,however, showed fragment length polymorphism. Automaticmultiple sequence alignments of the ribosomal sequences wereconstructed with Clustal X (Thompson et al. 1997). The effectof alternative gap opening (GOP) and extension (GEP) costschemes was investigated by constructing different automaticalignments under the following combinations (GOP/GEP): 8/2,8/4, 20/2, 24/4 and 24/6 (in all cases transition weight was fixedto 0.5). Congruence among data partitions as measured by therescaled incongruence length difference (RILD) (Wheeler and

Hayashi 1998) was chosen as the optimality criterion to selectthe best ribosomal alignment, which was subsequently used inall the phylogenetic analyses performed. In all analyses, gapswere treated as separate presence/absence characters, accordingto a set of rules based on gap overlapping and sharing of the 5′and/or the 3′ termini (Simmons and Ochoterena 2000). Thiscoding scheme allows the incorporation of indel informationinto phylogenetic reconstruction using not only parsimony butalso Bayesian inference methods, while minimising the effect ofthe increase in weight of overlapping multiple non-homologousgaps as a result of scoring gaps as an additional state (Pons andVogler 2006). The GapCoder computer program (Young andHealy 2003) was used to facilitate the automatic recoding of thealignments based on the simple method proposed by Simmonset al. (2001). The WINCLADA ver. 1.00.08 (Nixon 2002)program was used to obtain a combined data matrix by concate-nating the preferred automatic rrnL alignment with the cox1sequences and morphology. Parsimony analysis was conductedwith the TNT ver. 1.0 program (Goloboff et al. 2003). Heuristicsearches consisted of 1000 iterations of Wagner trees con-structed with random addition taxa and subsequent TBR branch

12

34

5

7

6

89

Fuerteventura22 My

Lanzarote15.5 My

Gran Canaria14.5 My

Tenerife12 My

La Gomera11 My

La Palma2 My

El Hierro1.2 My

Fig. 1. Maps showing the location of the Macaronesian Archipelagoes (Cape Verde not included), a close-up of theCanary Islands and a larger-scale map of Tenerife, showing sampling locations listed in Table 3. Ages of the subaerialstage based on radiometric dating in millions of years before the present date (Carracedo and Day 2002).


swapping, holding five trees per iteration and up to a totalmaximum of 10000 trees. Clade support was assessed via stan-dard bootstrap resampling (Felsenstein 1985) as implemented inTNT, based on 1000 replicates with individual heuristicsearches consisting of 15 iterations of Wagner tree constructionusing random addition of taxa, holding five trees per iterationand an overall maximum of 10000. The relative contributions ofeach gene fragment to clade support in the simultaneousanalysis was visualised by means of the partitioned BremerSupports (Baker and DeSalle 1997; Baker et al. 1998) calcu-lated in PAUP* (Swofford 2001) supported by the TreeRot ver.2c (Sorenson 1999) software.

The program MODELTEST ver. 3.06 (Posada and Crandall1998) was used to select the substitution model that best fittedthe data with the fewer parameters as indicated by the Akaikeinformation criterion (AIC) (Akaike 1973), which allows thecomparison of multiple nested models and accounts for modelselection uncertainty (Posada and Buckley 2004).

Bayesian inference is the only available model-based methodthat allows incorporation of morphological data into the phylo-genetic analysis. Bayesian analyses were performed withMRBAYES ver. 3.1.1 (Ronquist and Huelsenbeck 2003). Thebest model of evolution selected by MODELTEST was specifiedfor each gene fragment and independent standard discretemodels were implemented for the morphological characters andgaps scored as absence/presence data. The substitution estimateswere allowed to vary independently between each partition. Twoindependent runs were performed to assess convergence of theresults. Four simultaneous MCMC (Markov Chain Monte Carlo)chains (one cold and three heated) beginning with random start-ing trees were run to 4 million generations, sampling the Markovchain every 1000 generations. The standard deviation of the splitfrequencies of the two (< 0.01) and the TRACER ver. 1.3 soft-ware (Rambaut and Drummond 2003) were used to ensure thateach MCMC run had reached stationarity, as well as to ensureconvergence of the two runs and to determine the number of gen-erations to be discard as a burn-in.

The concatenated matrix of morphological characters andaligned sequences is available for download from the Inverte-brate Systematics website as ‘Accessory Material’ to this paper.

Uncorrected genetic distances between troglobitic and theirnontroglobitic sister-taxa were estimated with MEGA ver. 3.0(Kumar et al. 2004) and used as a proxy of the time of coloni-sation of the underground environment.

Statistical analyses of morphological variablesWe analysed patterns of morphological variation betweensurface and troglobitic sister-species, as inferred from ourphylogenetic analyses. Variation in quantitative morphologicalcharacters and their association with underground life, lineageand/or sex was investigated by means of standard statisticalanalyses of the following measurements: maximum carapacelength (P1), minimum (P2min), and maximum carapace width(P2max); length of the basal segment of the chelicera in lateralview (Q1); maximum width of the basal segment in lateral view(Q2); cheliceral fang length (F); length of the prolateral marginof the basal segment (Esc); length of the femur of leg 1 (fe1),and length of the metatarsus of leg 4 (me4). Pearson correlationswere conducted between all morphological measured variables

and individuals to assess allometric differences among individ-uals as a consequence of adaptation and/or speciation. A simi-larity matrix was estimated for each individual and the ninemorphological characters using the Bray–Curtis distance.Individuals with similar morphological measures were groupedusing a hierarchical agglomerative cluster based on the similar-ity matrix. These exploratory analyses were performed with thesoftware packages SPSS (www.spss.com) and Primer ver. 5.2.2(www.primer-e.com). Several methods were used to identifydifferences between troglobitic and epigean species, as well aswithin troglobites. Interespecific regressions were conductedfor each variable against P1 to remove the effect of body size,and the residual values were used in subsequent analyses. Wefirst performed a nonparametric one-way Kruskal–Wallis test(since morphological measurements did not meet the assump-tions of normality and homogeneity) in order to elucidate theexistence of sexual dimorphism in targeted species. A principalcomponents analysis (PCA) was conducted for all individualson the residual variables and P1 values to reduce variability of‘morphological measurements’ to fewer dimensions, and toobtain the variance explained by these few independent axes(Legendre and Legendre 1998). A one-way analysis of variancewith factor habitat (subterranean versus surface), and dependentvariable ‘morphological measurement’ was performed using anonparametric Kruskal–Wallis test for each pair of troglobiteand epigean sister-species. A second PCA was performed toinvestigate patterns of species exclusion and/or association insubterranean habitats, using all individual values of each vari-able for each species. Kruskal–Wallis tests and PCAs were runwith STATISTICA (www.statsoft.com).

ResultsPhylogenetic analysesAutomatic alignment of the rrnL under the 5-parameter costschemes assayed resulted in alignments with the number of gapcharacters ranging from 83 (24/6) to 138 (8/2) (Table 1). TherrnL alignment with gap opening 24 and gap extension 4, max-imised congruence among data partitions as measured by theRILD (Table 1). Parsimony analyses of the combined datamatrix with the preferred alignment yielded five trees of length5764, the strict consensus of which is shown in Fig. 2. Standardbootstrap support and partition Bremer supports for each cladeare listed in Appendix 2. The AIC as implemented in MODEL-TEST selected the GTR+I+Γ as the preferred evolutionarymodel for both the cox1 and rrnL genes. The two Bayesian runsconverged after 1.947 × 106 generations. Posterior probabilityvalues compatible with the clades recovered in the parsimonyanalyses are listed in Appendix 2.

Parsimony and Bayesian analyses are congruent in splittingCanarian members of Dysdera into two groups: the speciesD. lancerotensis, whose relationships to the Moroccan speciesare well supported in both analyses, and a large clade with highposterior probability that includes all remaining Canarianspecies sampled, along with the Madeiran endemic D. coiffaitiDenis, 1962. All endemic species from the western and centralCanaries form a monophyletic group with high posterior proba-bility, except D. andamanae and D. sibyllina, sp. nov., which areshown as more closely related to endemic species from the


http://www.publish.csiro.au/?act=view_file&file_id=IS07015_AC.nex


eastern Canary Islands – albeit with low bootstrap and posteriorprobability. Canarian troglobitic species are spread along themain Canarian clade and are clearly polyphyletic, except forD. hernandezi and D. esquiveli, which are sister-species.Bayesian analysis also suggests a sister-species relationshipbetween D. ambulotenta and D. labradaensis, albeit with lowposterior probability (0.73). Full descriptions of the actualphylogenetic relationships of each troglobitic species aredetailed in the Taxonomy section (see below). Average uncor-rected genetic distances between cave-dwelling species andtheir identified sister-groups, as well as within species, are listedin Table 2.

Statistical analysesMeasurements were taken for 91 specimens of the targetedspecies and are summarised in Tables 4 and 5. No significantdifferences were found between sexes within the Dysderaspecies investigated. Consequently, all individuals of eachspecies were pooled together for subsequent analyses. Pearsoncorrelations showed a strong positive linear relationship amongthe eight analysed variables (n = 90, all P values < 0.05, data notshown), which suggests a linear function of allometries amongthe studied species. Targeted species were divided into threegroups based on 73% of Bray–Curtis similarity among individ-uals. These groups are related to body size and include the largespecies (D. ambulotenta, D. gibbifera, D. labradaensis, andD. ratonensis), the medium size species (D. brevisetaeWunderlich, 1991, D. chioensis, D. guayota Arnedo & Ribera,1999, D. liostethus Simon, 1907, D. unguimmanis, and D. igua-nensis Wunderlich, 1987), and the small species (D. anda-manae, D. esquiveli, D. gollumi, D. hernandezi, D. levipesWunderlich, 1987, D. madai, sp. nov., and D. sibyllina, sp. nov.).The Kruskal–Wallis test found significant differences betweenhabitats (subterranean versus epigean) in at least one morpho-logical variable in seven out of eight lineages analysed (Table 6).In general, troglobites are larger than their epigean sister-species, although significant differences were only found in lin-eages 2, 6 and 9. In lineage 3, appendage length was the onlymeasurement significantly longer in troglobites. In all remain-ing lineages, troglobites showed a significantly longer fe1, butonly in three lineages was the me4 also significantly longer.Conversely, there was no evidence for longer appendages introglobites of lineages 1, 2 and 7. No significant differences

among epigean and subterranean individuals were found inlineage 7, which included the only intraspecific comparison.

The first axis of the PCA of the morphological variablesbetween species (data not shown) suggested a negative relation-ship between appendage length (fe1, me4) and minimum widthof the prosome (P2min) (accounting for 30.78% of the vari-ance). The second axis accounted for 25.43% of the variationand included associations related to cheliceral differences (Q1,Esc, F). The distribution of species using their scores on the twofirst axes (Fig. 3A) illustrates substitution of epigean, mostly onthe negative area of the first axis by troglobites towards the pos-itive area of the first axis, under the influence of cheliceral vari-ation. These differences are less conspicuous in themedium-sized group. The plot of the first PCA two axes isshown in Fig. 3B. The first axis displays a negative relationshipbetween appendage length and carapace minimum length thatexplains 31.99% of the variation. The second axis accountingfor 28.74% of the variation, is mostly explained by cheliceraldifferences (Q1 and F).

Taxonomy

Family DYSDERIDAE

Genus Dysdera Latreille

Dysdera andamanae Arnedo & Ribera

(Fig. 4A–B; Table 7)

Dysdera andamanae Arnedo & Ribera, 1997: 206–208, fig. 2–4.

Material examinedHolotype. 1�, Canary Is, Gran Canaria, Santa María de Guía, Brezal del

Palmital; 9.ii.1996, coll. M. A. Arnedo, B. Emerson and P. Oromí (UB2976).

Additional material examined. Canary Is: 1�, Gran Canaria, SantaMaría de Guía, Barranco Oscuro, (CRBA 694). 1�, Brezal del Palmital(CRBA 692).

Diagnosis

Dysdera andamanae is distinguished from other endemics inGran Canaria by its smaller size (± 2 mm carapace length). It

Table 1. Summary of the parsimony searches conducted separately on the morphological characters (morphology), the protein coding gene cox1and the ribosomal gene rrnL alignments constructed under different parameter combinations, and conducted on the combined data matrices of

morphology and cox1 with each ribosomal alignment# trees, number of most parsimonious trees (MPTs); length, number of steps of the MPTs; gaps, number of gap absence/presence characters; sum, sum of thelength of the partial analyses; info, number of parsimony informative sites in the combined data matrices; max, maximum number of steps of the combineddata matrices; RILD, rescaled incongruence length difference among morphology, cox1 and each ribosomal alignment; GOP, gap opening penalty; GEP, gap

extension penalty. Best parameter combination shown in bold

Morphology cox1 rrnL Combined RILD# trees length # trees length GOP/GEP gaps # trees length sum info # trees length max

536 417 12 2776 8/2 138 4 2238 5431 554 4 5706 8874 0.0798722536 417 12 2776 8/4 111 6 2241 5434 542 14 5696 8849 0.07672035536 417 12 2776 20/2 90 15 2280 5473 535 14 5738 8943 0.07636888536 417 12 2776 24/4 84 152 2305 5498 529 4 5764 9033 0.07524752536 417 12 2776 24/6 83 20 2302 5495 527 2 5762 9026 0.07561597

Invertebrate Systematics 629Systematics and evolution of Canarian troglobitic Dysdera

Harpactea hombergiStalita stygiaHarpactocrates raduliferDysderocrates silvestrisDysdera adriatica

Dysdera atlantica

Dysdera coiffaiti

Dysdera crocota

Dysdera erythrina

Dysdera fuscipes

Dysdera lucidipes melillensis

Dysdera mucronata

Dysdera ninniDysdera scabricula

Dysdera cf. seclusa

Dysdera sp. MA

Dysdera sp. MB

Dysdera sp. MC

Dysdera sp. MD

Dysdera sp. MF

Dysdera sp. MHDysdera edumifera

Dysdera alegranzaensis

Dysdera lancerotensis

Dysdera longa Dysdera nesiotes

Dysdera sanborondonDysdera spinidorsum

Dysdera sibyllina.sp.

Dysdera arabisenen

Dysdera bandamae

Dysdera iguanensis ADysdera iguanensis T

Dysdera insulana

Dysdera paucispinosaDysdera rugichelis

Dysdera tilosensis

Dysdera liostethusDysdera liostethus

Dysdera yguanirae

Dysdera ambulotenta

Dysdera brevisetae

Dysdera brevispina

Dysdera chioensis EDysdera chioensis W

Dysdera cribellata

Dysdera curvisetae

Dysdera esquiveli

Dysdera gibbifera

Dysdera levipes T

Dysdera guayota

Dysdera hernandezi

Dysdera labradaensis

Dysdera macra

Dysdera montanetensis

Dysdera verneaui

Dysdera unguimmanis

Dysdera volcania

Dysdera calderensis GDysdera calderensis P

Dysdera gomerensis GDysdera gomerensis H

Dysdera enghoffi

Dysdera levipes G

Dysdera gollumi

Dysdera orahan Dysdera ramblae

Dysdera ratonensis

Dysdera silvatica GDysdera silvatica P

Dysdera silvatica H

Dysdera madai n.sp.Dysdera andamanae

Dysdera mauritanica

EA

ST

ER

N C

AN

AR

IES

WE

ST

ER

N-C

EN

TR

AL C

AN

AR

IES

1

6564

63

6261

60

59

58

5756

55

5453

52

5150

4948

4746

45

4443

42

41

40

39

3837

36

3534

33

32

31

3029

28

2726

25

24

2322

2120

19

18

1716

1514

13

1211

109

87

6

5

4

3

2

Fig. 2. Strict consensus tree of the five most parsimonious trees for the genus Dysdera (5764 steps; CI = 0.19, RI = 0.41). Clade numberscorrespond to support values listed in Appendix 2. Thick branches denote clades recovered under all alignment parameter cost combinations.Black dots identify troglobites, white dots refer to endemic Canarian species exclusively reported from epigean localities and grey dots showepigean species occasionally collected in subterranean habitats (lava tubes or the mesovoid shallow substratum).


differs from closely related D. minutissima from Tenerife by thepresence of LF in the male bulb and a more complete scleroti-sation of the vulva VA (Fig. 4A); and from D. sibyllina, sp. nov.by the presence of well-developed eyes and the absence ofspines on the femora.

DescriptionFemale CRBA 694Fig. 4A–B. All characters as in male except: carapace 1.84

mm long; maximum width 1.38 mm; minimum width 0.87 mm.AME diameter 0.09 mm; PLE 0.08 mm; PME 0.08 mm; AMEseparated from one another by ~1 diameter or more. Labiumslightly notched at anterior tip. Sternum brownish-orange, uni-formly distributed.

Chelicerae 0.69 mm long, fang short, 0.41 mm; basalsegment dorsal side completely covered with piligerous granu-lations, ventral side smooth. Chelicera tooth D trapezoid,located roughly at centre of groove. Leg lengths describedabove: fe1 1.40 mm (all measurements in mm); pa1 0.82; ti11.12; me1 1.12; ta1 0.38; total 4.85; fe2 1.33; pa2 0.77; ti2 1.1;me2 1.07; ta2 0.41; total 4.67; fe3 1.02; pa3 0.56; ti3 0.77; me3

0.92; ta3 0.36; total 3.62; fe4 1.33; pa4 0.66; ti4 1.05; me4 1.35;ta4 0.43; total 4.82; fe Pdp 0.77; pa Pdp 0.36; ti Pdp 0.38; ta Pdp0.51; total 2.01; relative lengths 1>4>2>3. Spination: plap, leg1,leg2 spineless. Tb3d spines arranged in two bands: proximal0–1.0.0, distal 1.0.0, tb3v spineless, with one terminal spine onforward margin. Tb4d spines arranged in two bands: proximal0.0.1, distal 0.0.1, tb4v spines arranged in one band: proximal0.0–1.0, with one terminal spine at one leg and two on other one.Dorsal side of anterior legs smooth; ventral side of pedipalpsmooth.

Abdomen 2.80 mm long; cream-coloured; cylindrical.Abdominal dorsal hairs 0.05 mm long; thin, curved, not com-pressed, pointed; uniformly, thickly distributed.

Vulva DA clearly distinguishable from VA (Fig. 4A); DAslightly wider than long; DF wide in dorsal view. MF marginsfused, sheet-like, well developed, completely sclerotised. VArectangular; anterior region completely sclerotised; posteriorregion sclerotised except for innermost area; AVD clearly recog-nisable (Fig. 4B). S attachment projected under VA; armsgreatly reduced almost absent; neck as wide as arms. TB usualshape.

Table 2. Uncorrected genetic distances for the cox1 between troglobitic and epigean sister-species and within species

# specs, number of specimens included; standard deviation shown in brackets. nc = not calculated

# specs D. ambulotenta D. labradaensis D. gibbifera

D. ambulotenta 2 0.069 (0.014)D. labradaensis 2 0.146 (0.016) 0.002 (0.002)D. gibbifera 3 0.113 (0.013) 0.137 (0.015) 0.007 (0.003)

D. chioensis D. guayotaD. chioensis 3 0.078 (0.01)D. guayota 1 0.132 (0.014) n/c

D. esquiveli D. hernandezi D. brevisetaeD. esquiveli 1 n/cD. hernandezi 1 0.1105 (n/c) n/cD. brevisetae 1 0.113 (n/c) 0.116 (n/c) n/c

D. unguimmanisD. unguimmanis 2 0.089 (0.013)

D. gollumi D. levipesD. gollumi 2 0.011 (0.05)D. levipes 3 0.091 (0.01) 0.129 (0.012)

D. madai, sp. nov. D. iguanensisD. madai, sp. nov. 1 n/cD. iguanensis 2 0.125 (0.015) 0.004 (0.003)

D. silvaticaD. silvatica 3 0.076 (0.01)

D. ratonensis D. liostethusD. ratonensis 9 0.016 (0.003)D. liostethus 2 0.08 (0.01) 0.102 (0.013)

D. sibyllina, sp. nov. D. andamanaeD. sybyllina, sp. nov. 1 n/cD. andamanae 2 0.169 (0.018) 0.007 (0.004)


Intraspecific variationThe carapace ranges in length from 1.84–2.27 mm; the femaleis smaller in size. Spination variability is shown in Table 7.

DistributionThis species is only known from two close localities in northernGran Canaria that contain the last remnants of laurel forest onthis island.

RemarksDysdera andamanae has never been collected from undergroundhabitats. The description of its previously unknown female,however, has been included in the present revision because of itsclose similarity to a new troglobite described below.

DNA sequencesMitochondrial rrnL (GenBank accession number EU068056,EU068066) and cox1 (EU068027, EU068028) DNA sequencesof two individuals are reported here for the first time.

Phylogenetic relationshipsBoth parsimony and Bayesian inference analyses of themorphological data and the mitochondrial genes support the

sister-species relationship of D. andamanae to D. sibyllina, sp.nov. In the cladistic analyses of Arnedo et al. (2001), whichincluded species with no molecular information, D. andamanaewas shown as sister to D. minutissima, and hence these threespecies most likely belong to the same evolutionary lineage.

Dysdera chioensis Wunderlich

(Fig. 5H)

Dysdera chioensis Wunderlich, 1991: 291, figs 21–23. – Arnedo &Ribera, 1999: 618–620, figs 49–53, 73, 74.

Material examinedHolotype. 1�, Canary Is, Tenerife, Cueva Grande de Chío, Guía de

Isora, 29.iv.1985, coll. G. I. E. T. (ULL T-GC-5). Additional material examined. Canary Is: 1�, Tenerife, Güímar,

Cueva Honda de Güímar (ULL HG-1, C/AÑ085).

Distribution

This species was previously reported from Cueva de LosRoques and Cueva Grande de Chío in central and north-westTenerife and here reported for the first time from a cave ineastern Tenerife.


Table 3. List of the lava tubes and MSS localities where troglobitic Dysdera species have been collected in TenerifeCodes as shown in Fig. 1

Locality Code Species

MSS Barranco de los Cochinos 1 D. madai, sp. nov., D. esquiveliCueva de Felipe Reventón 2 D. ambulotenta, D. esquiveli, D. labradaensis, D. ungimmanisCueva del Viento-Sobrado 3 D. ambulotenta, D. esquiveli, D. labradaensis, D. ungimmanis, D. sibyllina, sp. nov.Cueva del Bucio 4 D. ambulotenta, D. labradaensis, D. ungimmanisCueva Labrada-Mechas 5 D. ambulotenta, D. esquiveli, D. hernandezi, D. labradaensis,Cueva de la Puerta 6 D. hernandeziCueva Honda de Güímar 7 D. chioensisCueva de los Roques 8 D. ambulotenta, D. chioensis, D. gollumiCueva Grande de Chío 9 D. chioensis

Table 4. Morphological measurement data for Dysdera spp.Data shown are mean values ± 1 s.e. for all individuals of the species investigated. See text for abbreviations of morphological variables

Lineage Species Habitat P1 P2max P2min Q1L Q2 Esc Fang Fe1 me4

1 D. ambulotenta subterranean 7.14 ± 0.85 5.63 ± 0.50 3.62 ± 0.31 3.98 ± 0.36 2.16 ± 0.26 1.67 ± 0.21 3.20 ± 0.32 6.68 ± 0.71 6.15 ± 0.60D. labradaensis subterranean 7.49 ± 1.45 5.79 ± 1.31 3.37 ± 0.67 3.50 ± 0.73 2.13 ± 0.38 1.45 ± 0.37 2.47 ± 0.55 7.65 ± 1.33 8.97 ± 1.53D. gibbifera epigean 6.28 ± 0.48 5.11 ± 0.29 3.24 ± 0.27 3.04 ± 0.12 1.89 ± 0.16 1.15 ± 0.16 2.17 ± 0.16 5.65 ± 0.39 5.86 ± 0.18

2 D. chioensis subterranean 4.13 ± 0.43 3.29 ± 0.31 2.17 ± 0.24 1.80 ± 0.27 1.29 ± 0.10 0.62 ± 0.08 1.29 ± 0.15 2.97 ± 0.33 2.95 ± 0.27D. guayota epigean 3.51 ± 0.21 2.96 ± 0.20 2.00 ± 0.11 1.56 ± 0.11 1.15 ± 0.09 0.53 ± 0.11 1.13 ± 0.10 2.73 ± 0.40 2.39 ± 0.32

3 D. esquiveli subterranean 2.21 ± 0.17 1.62 ± 0.16 1.01 ± 0.12 0.83 ± 0.13 0.48 ± 0.10 0.35 ± 0.08 0.63 ± 0.11 1.79 ± 0.20 1.47 ± 0.12D. hernandezi subterranean 2.18 ± 0.16 1.62 ± 0.10 1.07 ± 0.08 0.91 ± 0.02 0.52 ± 0.05 0.37 ± 0.07 0.73 ± 0.07 1.70 ± 0.06 1.38 ± 0.08D. brevisetae epigean 3.60 ± 0.35 2.70 ± 0.20 1.95 ± 0.20 1.67 ± 0.15 0.96 ± 0.08 0.67 ± 0.13 1.42 ± 0.14 2.65 ± 0.17 2.33 ± 0.20

4 D. unguimmanis subterranean 2.79 ± 0.13 2.07 ± 0.10 1.22 ± 0.07 1.26 ± 0.10 0.70 ± 0.06 0.59 ± 0.03 1.12 ± 0.07 3.94 ± 0.35 3.51 ± 0.355 D. gollumi subterranean 1.98 ± 0.11 1.45 ± 0.04 0.71 ± 0.13 0.62 ± 0.05 0.39 ± 0.05 0.27 ± 0.03 0.46 ± 0.03 2.02 ± 0.09 1.93 ± 0.09

D. levipes epigean 1.99 ± 0.47 1.55 ± 0.34 0.84 ± 0.20 0.61 ± 0.18 0.34 ± 0.11 0.19 ± 0.05 0.46 ± 0.11 1.60 ± 0.41 1.49 ± 0.356 D. madai, sp. nov. subterranean 2.50 ± 0.07 1.90 ± 0.06 1.17 ± 0.05 1.06 ± 0.04 0.85 ± 0.19 0.43 ± 0.09 0.72 ± 0.03 2.21 ± 0.08 2.27 ± 0.08

D.iguanensis epigean 3.12 ± 0.10 2.30 ± 0.11 1.44 ± 0.07 1.25 ± 0.06 0.79 ± 0.03 0.44 ± 0.03 0.85 ± 0.06 2.62 ± 0.23 2.84 ± 0.187 D. silvatica subter/epigean 6.12 ± 0.15 4.92 ± 0.20 3.05 ± 0.22 2.94 ± 0.26 1.71 ± 0.12 1.20 ± 0.40 1.98 ± 0.15 5.96 ± 0.56 5.28 ± 0.368 D. ratonensis subterranean 6.49 ± 0.26 5.03 ± 0.25 3.34 ± 0.14 3.08 ± 0.18 1.80 ± 0.08 1.10 ± 0.11 2.34 ± 0.06 6.07 ± 0.38 5.53 ± 0.23

D. liostethus epigean 4.14 ± 0.31 3.29 ± 0.26 2.20 ± 0.17 2.05 ± 0.10 1.14 ± 0.09 0.65 ± 0.13 1.60 ± 0.09 3.37 ± 0.21 2.92 ± 0.169 D. sibyllina, sp. nov. subterranean 2.26 ± 0.08 1.77 ± 0.06 1.14 ± 0.03 1.08 ± 0.06 0.70 ± 0.08 0.38 ± 0.04 0.78 ± 0.09 2.33 ± 0.06 2.24 ± 0.08

D. andamanae epigean 2.10 ± 0.23 1.59 ± 0.19 1.04 ± 0.15 0.80 ± 0.10 0.56 ± 0.05 0.25 ± 0.03 0.52 ± 0.10 1.54 ± 0.12 1.41 ± 0.06


Tab

le 5

.M

orph

olog

ical

dat

a fo

r D

ysde

rasp

p.tr

ansf

orm

atio

n of

mor

phol

ogic

al m

easu

rem

ents

pro

duci

ng b

ody-

size

fre

e sh

ape

vari

able

sD

ata

show

n ar

e m

ean

valu

e ±

1 s.

e. f

or a

ll in

divi

dual

s of

the

spec

ies

inve

stig

ated

. See

text

for

abb

revi

atio

n in

mor

phol

ogic

al v

aria

bles

Lin

eage

Spe

cies

Hab

itat

P2m

axP

2min

Q1L

Q2

Esc

Fang

Fe1

me4

1D

. am

bulo

tent

asu

bter

rane

an–0

.027

9 ±

0.21

–0.0

024

±0.

140.

4616

±0.

170.

0707

±0.

090.

3078

±0.

040.

5918

±0.

080.

0621

±0.

23–0

.378

1 ±

0.25

D. l

abra

daen

sis

subt

erra

nean

–0.1

524

±0.

18–0

.435

6 ±

0.13

–0.2

101

±0.

18–0

.062

2 ±

0.20

0.01

42 ±

0.07

–0.2

768

±0.

050.

6987

±0.

102.

0929

±0.

34D

. gib

bife

raep

igea

n0.

1461

±0.

160.

0647

±0.

14–0

.023

4 ±

0.21

0.06

54 ±

0.11

–0.0

371

±0.

08–0

.105

0 ±

0.06

–0.1

459

±0.

310.

1712

±0.

432

D. c

hioe

nsis

subt

erra

nean

0.07

31 ±

0.05

0.12

07 ±

0.07

–0.0

983

±0.

140.

1353

±0.

05–0

.109

3 ±

0.05

–0.1

261

±0.

06–0

.766

3 ±

0.14

–0.6

287

±0.

17D

. gua

yota

epig

ean

0.23

75 ±

0.06

0.27

45 ±

0.01

–0.0

024

±0.

220.

1893

±0.

04–0

.068

0 ±

0.11

–0.0

441

±0.

10–0

.427

2 ±

0.21

–0.5

865

±0.

133

D. e

squi

veli

subt

erra

nean

–0.0

444

±0.

03–0

.035

7 ±

0.04

–0.0

318

±0.

05–0

.077

8 ±

0.06

0.02

77 ±

0.05

–0.0

264

±0.

06–0

.116

0 ±

0.08

–0.2

240

±0.

11D

. her

nand

ezi

subt

erra

nean

–0.0

201

±0.

040.

0384

±0.

020.

0627

±0.

07–0

.022

5 ±

0.05

0.05

32 ±

0.06

0.08

45 ±

0.07

–0.1

716

±0.

09–0

.290

0 ±

0.18

D. b

revi

seta

eep

igea

n–0

.088

3 ±

0.13

0.17

81 ±

0.08

0.05

87 ±

0.09

–0.0

226

±0.

040.

0485

±0.

100.

2081

±0.

06–0

.586

2 ±

0.23

–0.7

357

±0.

184

D. u

ngui

mm

anis

subt

erra

nean

–0.0

629

±0.

06–0

.126

8 ±

0.07

0.07

98 ±

0.05

–0.0

364

±0.

060.

1401

±0.

030.

2295

±0.

041.

4791

±0.

241.

2480

±0.

225

D. g

ollu

mi

subt

erra

nean

–0.0

243

±0.

05–0

.213

2 ±

0.08

–0.1

261

±0.

04–0

.095

0 ±

0.08

–0.0

067

±0.

03–0

.109

5 ±

0.04

0.32

97 ±

0.07

0.45

25 ±

0.06

D. l

evip

esep

igea

n0.

0659

±0.

06–0

.088

3 ±

0.05

–0.1

327

±0.

08–0

.141

4 ±

0.04

–0.0

878

±0.

05–0

.113

6 ±

0.07

–0.0

958

±0.

040.

0098

±0.

106

D. m

adai

, sp.

nov

.su

bter

rane

an0.

0077

±0.

05–0

.027

2 ±

0.03

0.03

85 ±

0.02

0.20

17 ±

0.18

0.04

72 ±

0.09

–0.0

561

±0.

020.

0270

±0.

060.

2877

±0.

05D

.igua

nens

isep

igea

n–0

.091

9 ±

0.06

–0.0

768

±0.

04–0

.106

9 ±

0.05

–0.0

450

±0.

01–0

.078

2 ±

0.03

–0.1

681

±0.

03–0

.153

5 ±

0.18

0.25

33 ±

0.13

7D

. sil

vati

casu

bter

rane

an /

epig

ean

0.08

45 ±

0.11

–0.0

394

±0.

17–0

.032

3 ±

0.24

–0.0

550

±0.

140.

0537

±0.

41–0

.224

2 ±

0.12

0.31

16 ±

0.53

–0.2

534

±0.

308

D. r

aton

ensi

ssu

bter

rane

an–0

.097

3 ±

0.12

0.06

40 ±

0.04

–0.0

930

±0.

13–0

.080

6 ±

0.05

–0.1

284

±0.

10–0

.009

5 ±

0.10

0.07

92 ±

0.23

–0.3

552

±0.

10D

. lio

stet

hus

epig

ean

0.06

88 ±

0.05

0.14

94 ±

0.06

0.14

39 ±

0.08

–0.0

105

±0.

07–0

.085

4 ±

0.08

0.17

48 ±

0.04

–0.3

750

±0.

19–0

.674

0 ±

0.21

9D

. sib

ylli

na,s

p. n

ov.

subt

erra

nean

0.06

45 ±

0.00

0.06

72 ±

0.02

0.18

81 ±

0.02

0.12

69 ±

0.06

0.04

62 ±

0.04

0.10

18 ±

0.06

0.37

33 ±

0.03

0.48

70 ±

0.07

D. a

ndam

anae

epig

ean

0.02

17 ±

0.00

0.04

94 ±

0.03

–0.0

052

±0.

030.

0414

±0.

12–0

.047

3 ±

0.03

–0.0

929

±0.

04–0

.262

4 ±

0.10

–0.1

791

±0.

17


Tab

le 6

.R

esul

ts o

f th

e K

rusk

al–W

allis

tes

t fo

r th

e ef

fect

of

habi

tat

(sub

terr

anea

n ve

rsus

epi

gean

) on

mor

phol

ogic

al v

aria

bles

for

eac

h si

ster

-gro

up c

ompa

riso

nsS

eete

xtfo

rab

brev

iatio

nsof

mor

phol

ogic

alva

riab

les.

Gre

ysh

ade

indi

cate

sva

riab

les

whe

reth

eva

lues

for

trog

lobi

tesp

ecie

sar

egr

eate

rth

anfo

rep

igea

nsp

ecie

s;bo

ldte

xtin

dica

tes

sign

ifan

tval

ues

(P<

0.05

)

P1

P2m

axP

2min

Q1L

Q2

Esc

Fang

fe1

me4

HP

HP

HP

HP

HP

HP

HP

HP

HP

Gr

12.

573

0.10

874.

3333

0.03

743.

1025

0.07

820.

4102

0.52

180.

0256

0.87

284.

3485

0.03

70

13.

6923

0.05

470.

2307

0.63

1G

r 2

5.06

020.

0245

5.06

020.

0245

50.

0245

1.08

880.

2967

1.8

0.17

970.

0224

0.88

081.

102

0.29

385

0.02

540.

0224

0.88

08G

r 3

12.3

127

0.00

051.

7526

0.18

5612

.272

70.

0005

3.20

810.

0733

2.42

420.

1195

0.05

450.

8153

11.7

409

0.00

0612

.272

70.

0005

12.2

727

0.00

05G

r 5

0.74

110.

3893

3.84

0.05

14.

860.

0275

0.59

990.

8065

0.54

0.46

246

0.01

430.

0599

0.80

656

0.01

436

0.01

43G

r 6

8.39

570.

0038

7.43

620.

0064

3.69

230.

0547

8.36

610.

0038

7.43

620.

0064

8.33

680.

0039

8.33

680.

0039

5.02

560.

025

0.23

070.

631

Gr

70.

60.

4386

0.6

0.43

860.

60.

4386

2.4

0.12

132.

40.

1213

0.6

0.43

862.

40.

1213

2.4

0.12

132.

40.

1213

Gr

88.

3368

0.00

395.

0256

0.02

54.

3485

0.03

77.

141

0.00

653.

1025

0.07

820.

923

0.33

678.

3076

0.00

47.

4102

0.00

656.

5641

0.01

04G

r 9

1.19

040.

2752

3.85

710.

0495

1.19

040.

2752

3.85

710.

0495

0.42

850.

5127

3.85

710.

0495

3.85

710.

0495

3.85

710.

0495

3.85

710.

0495


DNA sequences

Mitochondrial rrnL (GenBank accession numbers AF244189,AF244188) and cox1 (AF244282, AF244281) DNA sequencesfrom Cueva de los Roques and Cueva Grande de Chío werereported in Arnedo et al. (2001). Here we report new sequencesfrom Cueva Honda de Güimar (rrnL EU068067, cox1EU068030).

Phylogenetic relationships

Parsimony analyses of the morphological data and the mito-chondrial genes support the sister-species relationship ofD. chioensis and D. guayota, as suggested in Arnedo et al.(2001), although with low support. Bayesian inference contra-dicts these results and suggests D. chioensis is sister toD. verneaui Simon, 1883 + D. brevispina Wunderlich, 1991.

Dysdera curvisetae WunderlichDysdera curvisetae Wunderlich, 1987: 291, figs 12–17. – Wunderlich,

1991: 284–287; Arnedo & Ribera, 1999: 624–626, figs 66–72.

Material examinedHolotype. 1�, Canary Is, Tenerife, Icod de los Vinos, from small cave

on north coast of San Marcos, coll. J. Wunderlich (SMF). Additional material examined. Canary Is: 1�, Tenerife, Santiago del

Teide, Playa del Barranco de Natero (CRBA 514).

Distribution

This species was formerly known from a single male specimenfrom Cueva de San Marcos, but has subsequently been collectedin epigean habitats.

DNA sequences

Mitochondrial rrnL (GenBank accession number EU068068)and cox1 (EU068031) DNA sequences are reported here for thefirst time.

Phylogenetic relationshipsBoth parsimony and Bayesian inference analyses of themorphological and the mitochondrial data support former sug-gestions based solely on morphology that D. curvisetae isclosely allied to the Gran Canarian species D. liostethus andD. paucispinosa Wunderlich, 1991, as well as to D. ratonensisfrom La Palma.

Dysdera esquiveli Ribera & Blasco(Fig. 5J)

Dysdera esquiveli Ribera & Blasco, 1986: 42–44, fig. 1A–F. –Wunderlich, 1991: 284–287; Arnedo & Ribera, 1999: 625–629, fig.75–86.

Material examinedHolotype. 1�, Canary Is, Tenerife, Icod de los Vinos, Cueva del

Viento-Sobrado, 23.iii.1983, coll. J. L. Martín (ULL T-CV-118). Paratype. Canary Is: 1�, data as for holotype (ULL T-CV-119).Additional material examined. Canary Is: 1�, Tenerife, Buenavista,

Cabecera Barranco de Cochinos, Teno (CRBA 584). 1�, Icod de los Vinos,Cueva de Felipe Reventón, Laberinto Maestro Pepe (CRBA 1091).

Intraspecific variationThe specimen from the MSS is slightly larger than the remain-ing material (2.55 compared with 1.96–2.28 for remainingmaterial). Eyes are absent except for the right AME and leftPME that are reduced to tiny, fuzzy, clear spots (Fig. 5J). Cave-dwelling specimens also show different levels of eye reduction(from eyeless to all six eyes visible, but spot-like).

DistributionThe species was formerly known from several lava tubes in theIcod de los Vinos region, in northern Tenerife (Fig. 1). Thespecies is reported here for the first time in the MSS. The onlyspecimen found was collected in a MSS pitfall trap in the regionof Teno, north-west Tenerife (Fig. 1).

madmadmad

PC1

-3 -2 -1 0 1 2 3 4

PC

2

-4

-3

-2

-1

0

1

2

3

amb

ambamb

lablablabgib

gib

gib gib

gib

gib

gua

gua

guachi

chichichi

chiesq

esqesqesqesq

esq

esq

esqesqesq

esq

her

herher

her

bre

bre

brebrebre

bre

ungungungung

ung ung

lev lev

lev

lev golgolgolgol

gol

madmadmad

iguiguiguiguigu

igu

sil

sil

sil

sil

lio liolio

lio

liolio

ratrat

ratrat

rat

rat

and

andand

sybsybsyb

PC1

-3 -2 -1 0 1 2 3 4

PC

2

-5

-4

-3

-2

-1

0

1

2

3

amb

lab

chi esq

her ung

gol

mad

rat

syb

A B

Fig. 3. Principal components analysis of morphological variation in Dysdera. A, plot of the distribution of Dysdera epigean and troglobite lineages basedon scores from the first two principal components axes, using all individuals and morphologic measurements (see Appendix 1 for species abbreviations);B, plot of the distribution of troglobites based on the mean principal component scores and 95% confidence limits for each morphological measure and foreach species.


DNA sequences

Mitochondrial rrnL (GenBank accession numbers AF244206,AF244205) and cox1 (AF244298) DNA sequences from Cuevade Felipe Reventón were reported in Arnedo et al. (2001).


Parsimony and Bayesian analyses of the morphological andmolecular data support D. exquiveli and D. hernandezi as sister-taxa and that both species form a clade with the epigean Tenerifespecies D. macra and D. brevisetae.

Dysdera gibbifera Wunderlich

(Figs 4C–E, 5E)

Dysdera gibbifera Wunderlich, 1991: 293–294, figs 35, 36, 38, 39. –Wunderlich, 1991: 284–287; Arnedo & Ribera, 1997; Arnedo &Ribera, 1999: 629–632, fig. 87–95.

Material examinedHolotype. 1� (non �, incorrect identification), Canary Is, Tenerife,

Los Silos, Monte del Agua, MSS-3, 10.vii.1988, coll. A. L. Medina (UL T-H3–124).

Additional material examined. Canary Is: 2 juv., Tenerife, Teno,Buenavista, Cabecera Barranco de Cochinos (CRBA 580). 1�, Los Silos,Monte del Agua (CRBA 1082); 1� (CRBA 1083); 1� (CRBA 1084); 1�

subadult., 1� subadult. (CRBA 1085); 1�, 1� (ULL 2284). 1� subadult,Anaga, Santa Cruz de Tenerife, El Bailadero (CRBA 1081).

DiagnosisDysdera gibbifera differs from the closely related speciesD. ambulotenta and D. labradaensis by the presence of welldeveloped eyes (Fig. 5E) and reduced number of spines on thelegs, and from D. montanetensis and D. volcania by its largersize (carapace length ≥6.00 mm).

DescriptionFemale CRBA 1083Fig. 4C–E. All characters as in male except: carapace

6.94 mm long; maximum width 5.56 mm; minimum width3.57 mm. Slightly foveate at borders, slightly wrinkled withsmall black granulation mainly anteriorly. Anterior borderroughly round. AME diameter 0.22 mm; PLE 0.22 mm; PME0.20 mm; AME slightly back from anterior border, separated


0.5

A

B

C

D

E

Fig. 4. Vulva: Dysdera andamanae, A, ventral view; B, dorsal view. D. gibbifera, C, ventral view; D, dorsal view; E, lateral view.

Table 7. Intraspecific spination variability in D. andamanae

Proximal Medio-proximal Medio-distal Distal

Tibia 3 dorsal 0–1.0.0 0 0 1.0.0–1Tibia 4 dorsal 0–1.0.1 0 0 0.0.1Tibia 3 ventral 0 0 0 0Tibia 4 ventral 0.0–1.0 0 0 0


from one another by ~1 diameter or more, close to PLE; PMEabout one quarter of diameter apart, ~1 PME diameter from PLE(Fig. 5E). Sternum very slightly wrinkled, mainly between legsand anterior border; uniformly covered in slender black hairs.

Chelicerae 3.21 mm long. Leg lengths of female describedabove: fe1 5.86 mm (all measurements in mm), pa1 3.98; ti1

5.1; me1 4.99; ta1 0.97; total 20.91; fe2 5.46; pa2 3.72; ti2 4.84;me2 4.74; ta2 97; total 19.74; fe3 4.59; pa3 2.55; ti3 3.57; me34.59; ta3 1.07; total 16.37; fe4 5.66; pa4 3.11; ti4 4.79; me46.12; ta4 1.22; total 20.91; fe Pdp 3.77; pa Pdp 1.68; ti Pdp 1.73;ta Pdp 2.04; total 9.23; relative length: 4=1>2>3. Spination:palp, leg1, leg2 spineless. Fe3d spineless; pa3 spineless; tb3d

A B

j

G H

E F

DC

I

LK

Fig. 5. Carapace eye region, dorsal view. A, Dysdera silvatica (UB 2907), male from Barranco Aramaqué, LaGomera; B, D. silvatica (ULL 3917), male from Cueva Longueras, El Hierro; C, D. ratonensis (CRBA 1099), femalefrom Cueva Honda de Gallegos, La Palma; D, D. ratonensis (CRBA 1098), female from Cueva de los Palmeros, LaPalma; E, D. gibbifera (CRBA 1083), female from MSS Monte del Agua, Tenerife; F, D. labradaensis (ULL 2705),male from Cueva de Felipe Reventón, Tenerife; G, D. madai, sp. nov. (CRBA 1112), male from MSS Barrancode Cochinos, Tenerife; H, D. chioensis (ULL 85), female from Cueva Honda de Güímar, Tenerife; I, D. hernandezi(CRBA 1320), male from Cueva de la Puerta, Tenerife; J, D. esquiveli (CRBA 584), female from MSS Barrancode Cochinos, Tenerife; K, D. gollumi (CRBA 1161), male from Cueva de los Roques, Tenerife; L, D. sibyllina,sp. nov. (ULL 3166), male from Galería de los Ingleses, Tenerife.


spines arranged in three bands: proximal 1.0.0; medio-proximal1.1.1; distal 1.0.0; tb3v spines arranged in two bands: proximal1.1.0; distal 1.0.0; with two terminal spines. Fe4d spines in onerow; 1–2; pa4 spineless; tb4d spines arranged in three bands:proximal 1.0.1; medio-proximal 1.1.1; distal 1.1.1; tb4v spinesarranged in three bands: proximal 1.1.1; medio-proximal1.1–2.1; distal 1.1.1; with two terminal spines. Claws with10–14 teeth; hardly larger than claw width.

Abdomen 9.43 mm long; cream-coloured; cylindrical.Abdominal dorsal hairs 0.25–0.40 mm long; thin, curved,slightly compressed, pointed; uniformly, thickly distributed.Vulva DA (Fig. 4D) clearly distinguishable from VA; DA longerthan wide; DF wide in dorsal view. MF margins fused, sheet-like, well developed, completely sclerotised (Fig. 4E). VA rect-angular (Fig. 4D); anterior region completely sclerotised;posterior region sclerotised except for most internal area; AVDcleraly recognisable. S attachment projected under VA; arms aslong as DA, straight; ends projected forwards; neck hardlyvisible. TB usual shape.

Intraspecific variationThe somatic traits of the newly found females fall well withinthe variability reported for the males. The single specimen col-lected in the Anaga region (north-east Tenerife) is a subadultfemale but shows all the diagnostic characters for the species,including spination. There is no variation in eye size.

DistributionThe species is mostly known from MSS localities. One speci-men was from a lava tube (Cueva de Felipe Reventón, Icod delos Vinos) and another was from a single epigean locality atMonte del Agua. The species was originally thought to berestricted to the Teno region (north-west Tenerife), but asubadult female specimen collected in the MSS from Anaga(north-east Tenerife) is reported here.

DNA sequencesMitochondrial rrnL (GenBank accession number AF244182)and cox1 (AF244277) DNA sequences from Cueva de FelipeReventón were reported in Arnedo et al. (2001). New sequencesof the same genes have been obtained from MSS CabeceraBarranco de Cochinos, Teno (rrnL EU068033, EU068070 cox1EU068034).

Phylogenetic relationshipsDysdera gibbifera is sister to D. ambulotenta in the parsimonycombined analyses, albeit with low bootstrap support. Bayesianinference, however, suggests (PP = 0.92) that it is the sister-species to a clade formed by D. ambulotenta and D. labradaensis.

Dysdera gollumi Ribera & Arnedo(Figs 5K, 6A, 7A–C)

Dysdera gollumi Ribera & Arnedo, 1994: 115–119, fig. 1–3. – Arnedo& Ribera, 1999: 632, figs 118–119.

Material examinedHolotype. 1�, Canary Is, Tenerife, La Orotava, Cueva de Los Roques,

27.x.1991, coll. C. Ribera (UB 2567).

Additional material examined. Canary Is: 1�, Tenerife, P.N delTeide, La Orotava, Cueva de los Roques (ULL 1161); 1� (CRBA 585). 1�,Cueva de los Roques, ramal abierto (CRBA 1094).

DiagnosisThe combination of eye reduction (Fig. 5K), spineless legs andsmall size (carapace ≤2.00 mm long) distinguishes D. gollumifrom any other Canarian species of Dysdera. Dysdera gollumidiffers from other small and troglobitic species like D. esquiveliand D. hernandezi in having a foveate carapace and slenderbody shape.

DescriptionMale ULL 1161Figs 5K, 6A, 7A–C. All characters as in female except:

Carapace 1.84 mm long; maximum width 1.40 mm; minimumwidth 0.50 mm. Brownish-orange, uniformly distributed;heavily wrinkled, foveate, covered with small black granules;hairy, uniformly covered with white hairs. Anterior borderroughly round; posterior margin narrow, straight. Eyeless (5K).Sternum brownish-orange, uniformly distributed; uniformlycovered in slender black hairs.

Chelicerae 0.60 mm long, basal segment proximal border ofdorsal side strongly covered with setiferous granulations.Chelicera inner groove short, ~1/3 cheliceral length; M close toB. Legs pale yellow. Lengths of male described above: fe1 2 mm(all measurements in mm); pa1 1; ti1 1.8; me1 1.76; ta1 0.48;total 7.04; fe2 1.6; pa2 0.96; ti2 1.56; me2 1.66; ta2 0.44; total6.22; fe3 1.32; pa3 0.6; ti3 1.06; me3 1.4; ta3 0.4; total 4.78; fe41.76; pa4 0.84; ti4 1.52; me4 1.82; ta4 0.44; total 6.38; fe Pdp0.6; pa Pdp 0.32; ti Pdp 0.36; ta Pdp 0.42; total 1.7; relativelength: 1>4> 2>3. Spineless. Ventral side of anterior legssmooth.

Abdomen 2.30 mm long; whitish; cylindrical. Abdominaldorsal hairs 0.03–0.04 mm long; medium-sized, roughlystraight, not compressed, blunt, tip not enlarged; uniformly,thickly distributed.

Male copulatory bulb (Fig. 6A) T slightly shorter than DD;external distal border sloped backwards; internal sloped back-wards. DD bent ~45° in lateral view; internal distal border notexpanded. ES wider, more sclerotised than IS; IS truncated atDD middle part. DD tip straight in lateral view. C present, short;distal end on DD internal tip; well developed; located close toDD distal tip (Figs 7A, C); proximal border sharply decreasing;distal border stepped, upper tip projected, pointed, external sidehollowed. AC present. LF absent. L well developed; externalborder not sclerotised, laterally slightly folded; distal borderdivergent, not continuous, upper sheet strongly folded atmiddle. LA absent. F absent. AL absent. P fused to T; perpen-dicular to T in lateral view; lateral length from 2/5 to 1/2 of Twidth; ridge present, perpendicular to T; not expanded, uppermargin slightly toothed, mainly on external side, on its distalpart; very few teeth (1–3); not distally projected; posteriormargin slightly folded towards internal side.

Intraspecific variationSomatic variability in males falls well within the ranges previ-ously reported for females. There are different levels of eyeregression in this species: some specimens are eyeless, whereas



others have remains of both AME (female 1094, male 585) andtraces of one (1094) or two PLE (585).

DistributionThis species is only known from the lava tubes of Los Roques,located above 2000 m altitude on the south slope of the Teidevolcano in central Tenerife.

DNA sequencesMitochondrial rrnL (GenBank accession number AF244204)and cox1 (AF244297) DNA sequences were reported in Arnedoet al. (2001).


All analyses support the close relationship between D. gollumiand Tenerife representatives of D. levipes, rendering the laterspecies as paraphyletic.

Dysdera hernandezi Arnedo & Ribera

(Figs 5I, 6B, 8E)

Dysdera hernandezi Arnedo & Ribera, 1999: 636–637, figs 108–112,120, 121.

Material examinedHolotype. 1�, Canary Is, Tenerife, Cueva Labrada, El Sauzal,

11.xii.1984, coll. J. J. Hernández (ULL. 3214).

0.5

0.5

A

B

C

Fig. 6. Left male palp, prolateral view. A, Dysdera gollumi; B, D. hernandezi; C, D. labradaensis.


Additional material examined. Canary Is: 1�, Tenerife, Candelaria:Barranco Hondo (Candelaria), Cueva de la Puerta (CRBA 1320). 1�,Tenerife, El Sauzal, Cueva Labrada (ULL 1072).

DiagnosisA flattened cheliceral fang distinguishes D. hernandezi fromclosely related D. esquiveli. This kind of chelicerae is alsoobserved in D. ramblae from La Gomera from which D. hernan-dezi differs by its much smaller size (average carapace lengths5.70 mm versus 2.18 mm respectively) and eye reduction (Fig. 5I).

DescriptionMale CRBA 1320Fig. 6B. All characters as in female except: carapace

2.24 mm long; maximum width 1.63 mm; minimum width1.07 mm. Reddish dark brown, darkened at borders; foveate atborders, slightly wrinkled at middle, covered with small blackgranules. Anterior lateral borders parallel. Eyes markedlyreduced but all present; AME diameter 0.03 mm; PLE 0.01 mm;PME 0.01 mm; AME separation 0.17 mm; AME–PLE separa-tion 0.03 mm; PLE–PME separation 0.32 mm; PME separation0.09 mm (Fig. 5I). Labium wider bassaly than distally.

Chelicerae 0.92 mm long, ~1/3 of carapace length in dorsalview; fang medium-sized, 0.66 mm; basal segment proximaldorsal side scantly covered with piligerous granulations.Chelicera teeth B>D=M. Legs bicoloured, darker on proximalborder, becoming lighter distally. Lengths of male describedabove: fe1 1.73 mm (all measurements in mm); pa1 1.07; ti1 1.48;me1 1.38; ta1 0.41; total 6.07; fe2 1.68; pa2 1.12; ti2 1.48; me21.38; ta2 0.41; total 6.07; fe3 1.17; pa3 0.61; ti3 0.82; me3 1.07;ta3 0.36; total 4.03; fe4 1.43; pa4 0.82; ti4 1.17; me4 1.43; ta40.41; total 5.25; fe Pdp 0.97; pa Pdp 0.56; ti Pdp 0.48; ta Pdp 0.51;total 2.25; relative length: 1=2>4>3. Spination: palp, leg1, leg2spineless. Tb3d spines arranged in two bands: proximal 1.0.0;distal 1.0.0; tb3v spines arranged in two bands: proximal 1.0.1;distal 1.0.0, with two terminal spines. Pa4 1 ventral medial; tb4dspines arranged in two bands: proximal 1.0.1; distal 1.0.1; tb4vspines arranged in four bands: proximal 1.1.0; medio-proximal1.1.0; medio-distal 1.1.0; distal 0.1.0, with two terminal spines.Dorsal side of anterior legs smooth; ventral side of pedipalpsmooth. Claws with 10–14 teeth; hardly larger than claw width.

Abdomen 2.29 mm long; whitish; cylindrical.Abdominal dor-sal hairs 0.02 mm long; medium-sized, roughly straight, not com-pressed, blunt, tip not enlarged; uniformly, scantly distributed.

Male copulatory bulbus T slightly shorter than DD (Fig. 6B);external distal border straight; internal sloped backwards. DDslightly bent in lateral view, clearly less than 45°; internal distalborder not expanded. IS, ES equally developed; IS truncated atDD middle part. DD tip straight in lateral view. C present, short;distal end on DD internal tip; well developed; located close toDD distal tip (Fig. 8A–E); proximal border sharply decreasing;distal border stepped, upper tip not projected, rounded, externalside hollowed. AC present. LF absent. L poorly developed;external border not sclerotised, laterally slightly folded; distalborder approximately parallel, not continuous, upper sheetslightly folded at middle. LA absent. F absent. AL present, verypoorly developed; proximal border in posterior view fused withDH. P fused to T; perpendicular to T in lateral view; laterallength from 2/3 to as long as T width; ridge present, perpendic-

ular to T; not expanded, upper margin slightly toothed, mainlyon external side, on its distal part; few teeth (4–6); not distallyprojected; posterior margin slightly folded towards internal side.

Intraspecific variation

Carapace length ranges from 2.00–2.40 mm. The male speci-men shows a more ornamented carapace with slight dif-ferences in shape (parallel anterior borders), but otherwisefalls well within the reported somatic variability of females.There is a certain polymorphism in eye reduction; the malespecimen has all eyes, whereas most females have lost one ortwo PME.

Distribution

This species was previously known exclusively from CuevaLabrada; we report here the new locality Cueva de la Puerta, atonly 50 m altitude on the eastern slope of the dorsal ridge inTenerife.

DNA sequences

Mitochondrial rrnL (GenBank accession number EU068059)and cox1 (EU068037) DNA sequences from Cueva Labrada(1072) are here reported for the first time.


All phylogenetic analyses supports the sister-species relation-ship between D. hernandezi and D. esquiveli, and include thesetwo species in a clade also containing the Tenerife epigeanspecies D. macra and D. brevisetae.

Dysdera labradaensis Wunderlich

(Figs 5F, 6C, 9A–D)

Dysdera labradaensis Wunderlich, 1991: 296, figs 47–49. –Wunderlich, 1991: 284–287; Arnedo & Ribera, 1999: 637–639, figs113–117, 122, 123.

Material examinedHolotype. 1�, Canary Is, Tenerife, Cueva Labrada, El Sauzal,

12.ix.1984, coll. G. I. E. T. (ULL T-CL-59). Additional material examined. Canary Is: 1�, Tenerife, Icod de los

Vinos, Cueva de Felipe Reventón (ULL 2705).

Diagnosis

Dysdera labradaensis differs from similar species, D. gibbiferaand D. ambulotenta, by the presence of six reduced eyes (5F),and spinated anterior femora. These three species share acommon genitalic pattern with D. montanetensis and D. volca-nia, also from Tenerife, from which they differ by their largersize (average carapace lengths 5.5 versus 2.7 respectively).

Description

Male ULL 2705.

Figs 5F, 6D. All characters as in female except: carapace5.81 mm long; maximum width 4.28 mm; minimum width2.60 mm. Brownish-orange, anteriorly darker, becoming lighterposteriorly; smooth with some small black granules mainly ante-riorly. Anterior border roughly round, from 1/2 to 3/5 carapace



length; anterior lateral borders convergent; rounded at maximumdorsal width, posterior lateral borders straight; posterior marginnarrow, straight. Eyes markedly reduced but all present(Fig. 5F): AME diameter 0.13 mm; PLE 0.07 mm; PME0.09 mm; AME separation 0.30 mm; AME–PLE separation0.04 mm; PLE–PME separation 0.18 mm; PME separation0.06 mm. Sternum light orange, uniformly distributed; smooth;uniformly covered in slender black hairs.

Chelicerae 2.68 mm long, ~1/3 of carapace length in dorsalview; fang medium-sized, 1.84 mm; basal segment dorsal,ventral side completely covered with setiferous granulations.Chelicera inner groove short, ~1/3 cheliceral length. Legsbicoloured, darker on proximal border, becoming lighter dis-tally. Lengths of male described above: fe1 6.12 mm (all mea-surements in mm); pa1 3.57; ti1 5.87; me1 5.61; ta1 1.17; total22.3; fe2 5.61; pa2 3.32; ti2 5.81; me2 5.61; ta2 1.17; total 21.5;fe3 4.85; pa3 2.35; ti3 4.23; me3 5; ta3 1.17; total 17.6; fe4 6.12;pa4 2.81; ti4 5.66; me4 7.24; ta4 1.33; total 23.2; fe Pdp 3.06;pa Pdp 1.16; ti Pdp 1.73; ta Pdp 1.5; total 7.88; relative length:4>1>2>3. Spination: all femora spinated; fe1 3 terminal spineson forward margin; fe2 4 terminal spines on forward margin.Fe3d spines in two rows; forward 5; backward 4–2; pa3 1 pos-terodorsal; tb3d spines arranged in four bands: proximal1–0.1–0.1; medio-proximal 1.1.1; medio-distal 1.1–0.0; distal1.0.1.; tb3v spines arranged in four bands: proximal 0–1.0.0;medio-proximal 1.2–1.1; medio-distal 1.1–2.0–1; distal1.1–0.1; with two terminal spines. Fe4d spines in two rows:forward 3–2; backward 8; pa4 1 posterodorsal; tb4d spinesarranged in four bands: proximal 1.2–1.1; medio-proximal1.2–1.1; medio-distal 1.1.1; distal 1.1–0.1; tb4v spines arrangedin four bands: proximal 1.3–1.1; medio-proximal 1.1.1; medio-distal 1.1.1; distal 1.1.1; with two terminal spines. Dorsal sideof anterior legs smooth; ventral side of pedipalp smooth. Clawswith more than 15 teeth; hardly larger than claw width.

Abdomen 5.35 mm long; cream-coloured; cylindrical.Abdominal dorsal hairs 0.04–0.06 mm long; medium-sized,curved, not compressed, blunt, tip enlarged; uniformly, thicklydistributed.

Male copulatory bulbus (Fig. 5D) T slightly longer than DD;external distal border sloped backwards; internal sloped back-wards. DD slightly bent in lateral view, clearly less than 45°;internal distal border not expanded. ES wider, more sclerotisedthan IS; IS continuous to tip. DD tip straight in lateral view.C present, short; distal end on DD internal tip; well developed;located far from DD distal tip (9A–D); proximal border contin-uously decreasing; distal border sloping in its base, upper tip notprojected, rounded, external side smooth. AC present. LFpresent; distally not projected. L well developed. LF well devel-oped. L external border sclerotised, laterally markedly foldedbackwards; distal border divergent, continuous. LA absent.F absent. AL absent. P fused to T; markedly sloped on its prox-imal part, perpendicular on distal; lateral length as long as orlonger than T width; ridge present, parallel to T, not sclerotised;distinctly expanded, rounded, upper margin smooth; not distallyprojected; posterior margin not folded.

Intraspecific variationMale specimens are smaller than the range reported for females(7.00–8.33 mm), with slight differences in carapace shape andcoloration and larger, yet reduced, eyes.

DistributionThe species is known from several lava tubes in Icod de losVinos area and La Orotava (north central Tenerife).

DNA sequencesMitochondrial rrnL (GenBank accession number AF244179)and cox1 (AF244275) DNA sequences from Cueva del Viento-

A CB

Fig. 7. Dysdera gollumi right male bulb. A, retrolateral view; B, anterior view; C, posterior view.


Sobrado were reported in Arnedo et al. (2001). New sequencesof the same genes from Cueva de Felipe Reventón are herereported (ULL2705 x114, rrnL EU068074, cox1 EU068040).

Phylogenetic relationshipsParsimony and Bayesian inference analyses of the morpho-logical and molecular data support inclusion of D. labradaensisin the same clade as D. montanetensis, D. volcania, D. ambulo-tenta and D. gibbifera, although they disagree in its actual posi-tion, the species occurring either as a sister species toD. ambulotenta (Bayesian) or as the first species branching offthe clade (parsimony).

Dysdera madai, sp. nov. Arnedo(Figs 5G, 10A–G, 11A–H; Table 8)

Material examinedHolotype. 1�, Canary Is, Tenerife, Teno, Monte del Agua, Los Silos,

Cabecera Barranco de Cochinos, MSS pitfall; 28.iv.2004; coll. Contreras &Arnedo (CRBA 1112).

Paratypes. Canary Is: 1� same locality as holotype, 14.xii.2003; coll.H. López (UB 1095). 2�, Tenerife, Los Silos, Monte del Agua, CabeceraBarranco de Cochinos, MSS pitfall, 21.i.2003 (CRBA 583) (spinnerets andright leg1 removed for SEM); 1�, 25.x.2005 (CRBA 1328); 1�, 1� coll.G. I. E. T. (ULL 1330); 1�, coll. H. Contreras, 1.x.2003 (CRBA578).

Tissue collection. Canary Is, same locality as holotype: 1� (ULL968); 2 juv. (freezer collection CRBA 579); 1� (freezer collection CRBA1086); 1�, 1� (freezer collection CRBA 1331).

DiagnosisDysdera madai, sp. nov. differs from sympatric and morpho-logically similar D. iguanensis in having reduced eyes (Fig. 5G),smaller size (average carapace length 2.50 mm versus3.10 mm), reduced spination, lack of male bulb LF (Fig. 11B)


A C

D

B

E

Fig. 8. Dysdera hernandezi right male bulb. A, anterior view; B, prolateral view; C, posterior view; D, retrolateralview; E, posterior apophysis, retrolateral view.

Table 8. Intraspecific spination variability in D. madai, sp. nov.


Tibia 3 dorsal 0–1.0.1 0 0 1.0.0–1Tibia 4 dorsal 0–1.0.0–1 1.0.1 0 1.1.0–1Tibia 3 ventral 0–1.1–2.0 0 0 1.0.0Tibia 4 ventral 0.0–1.0 1.2.0–1 0 1.0–1.1


and the triangle-like shape of the vulva VA (Fig. 10F). Smallersize and a smoother carapace distinguish this species from sym-patric D. gibbifera and D. brevisetae, respectively.

Description

Holotype male

Figs 5G, 10A–D, 11A–F. Carapace (Fig. 10A) 2.42 mm long;maximum width 1.91 mm; minimum width 1.17 mm. Darkbrown, darkened at borders; slightly foveate at borders, slightlywrinkled with small black granules mainly anteriorly. Anteriorborder roughly round, markedly smaller than 1/2 carapacelength; anterior lateral borders divergent; rounded at maximumdorsal width, posterior lateral borders straight; posterior marginwide, straight. Eyes markedly reduced but all present (Fig. 5G):AME diameter 0.05 mm; PLE 0.03–0.05 mm; PME0.03–0.05 mm; AME separation 0.20 mm; AME–PLE separa-tion 0.02 mm; PLE–PME separation 0.07 mm; PME separation0.09 mm. Labium trapezoid-shaped, base wider than distal part;longer than wide at base; semicircular groove at tip. Sternumbrownish-orange, anteriorly darker, becoming lighter posteri-orly; wrinkled; uniformly covered in slender black hairs.

Chelicerae (Fig. 10B) 1.02 mm long, ~2/5 of carapace lengthin dorsal view; fang medium-sized, 0.71 mm; basal segmentdorsal, ventral side completely covered with setiferous granula-tions. Chelicera inner groove short, ~1/3 cheliceral length;armed with three teeth and lamina at base; D>B>M; D trape-zoid, located roughly at centre of groove; B close to basallamina; M close to B. Anterior legs dark orange, posterior legsyellow. Lengths of male described above: fe1 2.19 mm (all mea-surements in mm); pa1 1.38; ti1 1.91; me1 1.84; ta1 0.51; total7.83; fe2 2.04; pa2 1.27; ti2 1.73; me2 1.73; ta2 0.53; total 7.32;fe3 1.66; pa3 0.87; ti3 1.12; me3 1.58; ta3 0.51; total 5.74; fe42.19; pa4 1.15; ti4 1.79; me4 2.17; ta4 0.56; total 7.85; fe Pdp1.27; pa Pdp 0.61; ti Pdp 0.71; ta Pdp 0.64; total 3.24; relativelength: 4>1>2>3. Spination: leg1, leg2 spineless; Fe3d spine-less; pa3 spineless; tb3d spines arranged in two bands: proximal1.0.1; distal 1.0.1.; tb3v spines arranged in two bands: proximal0.1–2.0; distal 1.0.0; with two terminal spines. Fe4d spineless;pa4 spineless; tb4d spines arranged in three bands: proximal0.0.1; distal 1.0.1; tb4v spines arranged in two bands: proximal1.2.1; distal 1.1.1; with two terminal spines. Dorsal side of ante-rior legs covered with small setiferous granules; ventral side ofpedipalp covered with small setiferous granules. Claws witheight teeth or less; hardly larger than claw width (Fig. 11F).

Abdomen 3.01 mm long; whitish; cylindrical. Abdominaldorsal hairs 0.03 mm long; medium-sized, roughly straight,not compressed, blunt, tip not enlarged; uniformly, thicklydistributed.

Male copulatory bulbus (Fig. 10C–D) T slightly longer thanDD; external distal border straight; internal border projected atmiddle. DD slightly bent in lateral view, clearly less than 45°;internal distal border not expanded. ES wider, more sclerotisedthan IS; IS truncated at DD middle part. DD tip (Fig. 11A–E)straight in lateral view; anterior (upper) sheet internal partmarkedly projected above posterior (lower) sheet. C present,long; distal end beside DD internal tip; well developed; locatedfar from DD distal tip; proximal border continuously decreas-ing; distal border sloping in its base, upper tip projected,

rounded, external side hollowed. AC absent. LF absent. L welldeveloped; external border not sclerotised, not folded; distalborder perpendicular, continuous. LA absent. F absent. ALpresent, very poorly developed; proximal border in posteriorview smooth, not fused with distal haematodoca. P fused to T;perpendicular to T in lateral view; lateral length from 1/2 to 2/3of T width; ridge present, perpendicular to T; not expanded,upper margin slightly toothed, mainly on external side, on itsdistal part; very few teeth (1–3); not distally projected; posteriormargin slightly folded towards internal side.

Paratype female

Figs 10E–G, 11G–H. All characters as in male except: cara-pace 2.50 mm long; maximum width 1.84 mm; minimum width1.12 mm. Eyes markedly reduced but all present: AME diameter0.04 mm; PLE 0.03 mm; PME 0.04 mm; AME separation0.20 mm; AME–PLE separation 0.01 mm; PLE–PME separa-tion 0.07 mm; PME separation 0.01 mm.

Chelicerae 1.07 mm long, fang medium-sized, 0.71 mm.Chelicera teeth D=B>M. Lengths of female described above:fe1 2.19 mm (all measurements in mm); pa1 1.43; ti1 1.94; me11.79; ta1 0.53; total 7.78; fe2 1.99; pa2 1.28; ti2 1.73; me2 1.78;ta2 0.51; total 7.29; fe3 1.68; pa3 0.87; ti3 1.27; me3 1.63; ta30.53; total 5.99; fe4 2.29; pa4 1.12; ti4 1.88; me4 2.27; ta4 0.56;total 8.13; fe Pdp 1.27; pa Pdp 0.61; ti Pdp 0.57; ta Pdp 0.76;total 3.24; relative length 4>1>2>3. Spination: leg1, leg2 spine-less. Fe3d spineless; pa3 spineless; tb3d spines arranged in twobands: proximal 1.0.1; distal 1.0.0; tb3v spines arranged in twobands: proximal 0.1–2.0; distal 1.0.0; with two terminal spines.Fe4d spineless; pa4 spineless; tb4d spines arranged in threebands: proximal 0.0.1–0; medio-proximal 1.0.1; distal 1.0.0;tb4v spines arranged in two bands: proximal 1.3–2.0; distal1.1.1; with two terminal spines. Dorsal side of anterior legssmooth; ventral side of pedipalp smooth.

Abdomen 3.47 mm long; whitish; cylindrical. Abdominaldorsal hairs 0.06 mm long; thin, curved, compressed, pointed;uniformly, thickly distributed. Vulva DA (Fig. 10E) not distin-guishable from VA; arch-like, anteriorly pointed; DA slightlywider than long; DF wide in dorsal view. MF margins fused,sheet-like, well developed, completely sclerotised (Fig. 10G).VA anterior region completely sclerotised (Fig. 10F); posteriorregion sclerotised except for most internal area; sclerotisedridge at ventral VA external margin, as long as VA, bent to inter-nal side; AVD hardly visible. S attachment not projected underVA; arms as long as DA, straight; tips not projected; neck aswide as arms. TB usual shape. ALS without PS (Fig. 11G);remaining piriform spigots more external than MS, arranged inone row; 3+0 piriform gland spigots; PMS, PLS (Fig. 11H);with fewer than five aciniform gland spigots.

Intraspecific variation

The carapace ranges in length from 2.42–2.68 mm. Some speci-mens have lost some eyes (PME, PLE). Spination variability isshown in Table 8.

Distribution

This species is exclusively known from MSS localities in theTeno region (north-west Tenerife).


RemarksIn most MSS traps, this species was found together withD. iguanensis, D. gibbifera and D. brevisetae. Dysdera madaiclosely resembles D. iguanensis.

DNA sequencesMitochondrial rrnL (GenBank accession numbers EU068076,CRBA579, EU068061, CRBA583) and cox1 (EU068043,CRBA579) DNA sequences from Cabecera Barranco deCochinos, MSS are here reported.


A

C D

B

Fig. 9. Dysdera labradaensis right male bulb. A, retrolateral view; B, anterior view; C, prolateral view; D, posterior view.


Phylogenetic relationshipsParsimony and Bayesian inference analyses support D. madai, sp.nov. as the sister-species of D. iguanensis, although with weaksupport. Both analyses also support the inclusion of D. madai, sp.nov. + D. iguanensis in a clade that includes, among others, twoother troglobitic species (D. ambulotenta, D. labradaensis) and aspecies mostly found in MSS traps (D. gibbifera).

EtymologyThe term, ‘madai’, a noun in apposition means ‘deep’ in the lan-guage of the Guanches (Alvarez Rixo et al. 1991), the inhabi-tants of Tenerife at the time of the Castilian conquest, and refersto the habitat of this species.

0.5

0.2

2

0.5

A B

C D

E

F

G

Fig. 10. Dysdera madai, sp. nov. A, carapace, dorsal view; B, left chelicera, ventral view; C, left male palp, ante-rior view; D, left male palp, prolateral view; E, vulva, dorsal view; F, vulva, ventral view; G, vulva, lateral view.


A

C

B

D

F

G H

E

Fig. 11. Dysdera madai, sp. nov. right male bulb. A, retrolateral view; B, anterior view; C, prolateral view; D, prolateral view; E, posterior apophysis, prolateral view; F, walking leg tarsus. Spinnerets, G, right anteriorlateral spinnerets spigots; H, right posterior lateral spinnerets spigots.


Dysdera ratonensis Wunderlich(Fig. 5C–D)

Dysdera ratonensis Wunderlich, 1991: 306–307, figs 99–100. – Arnedo& Ribera, 1996: 109–122, figs 1–14.

Material examined.Canary Is: 1�, La Palma: Barlovento: Cueva Honda de Gallegos (CRBA1097); 1�, 4 juv., (CRBA 1099). Cueva de Los Palmeros; 1� (CRBA1098); 1 juv. (UL 2756). 1�, El Paso, Cueva de Tacande (RG 3176). 1�, 2juv., Fuencaliente: Cueva Arreboles (RG 3177). 2 juv., Cueva de losPalmeros (CRBA 2946/2956); remains (CRBA 2830). 3 juvs, Cueva delRatón (UL 2753–2755); 3 carapaces (UL2702, 2711–2712). 1� subadult,Garafía, Cueva Honda de la Fajanita (RG 3178). 1� subadult, 1 juv., Villade Mazo, Cueva Callejones (RG 3175). 1 juv., Cueva del Canal (RG 3179).1�, 1 juv., Cueva del Salto de Tigalate (RG2807); remains (RG 2830).

DistributionThe species has been reported in several lava tubes throughoutLa Palma.

RemarksThis large species (carapace length ranges from 5.74 mm to6.51 mm) shows polymorphic eye size, apparently related togeographical location. Southern populations have eyes reducedin size (yet all eyes are present), whereas in the northernmostlocalities (Cueva Honda de La Fajanita and Cueva Honda deGallegos) eye reduction is moderate or non-existent (Fig. 5).

DNA sequencesMitochondrial rrnL (GenBank accession numbers AF244214,AF244213) and cox1 (AF244305, AF244304) DNA sequencesfrom Cueva de Tacande, Mazo, and Cueva de los Palmeros,Fuencaliente, were reported in Arnedo et al. (2001). Additionalcox1 sequences (GenBank accession numbersEU068048–EU068055) from Cueva Los Palmeros and CuevaHonda de Gallegos are here reported.

Phylogenetic relationshipsAll analyses agree (parsimony bootstrap support 60, Bayesian PP<0.95) in joining D. ratonensis to the clade formed byD. liostethus from Gran Canaria and D. curvisetae from Tenerife.

Dysdera sibyllina, sp. nov. Arnedo(Figs 5L, 12A–F, 13A–H; Table 9)

Material examinedHolotype. 1�, Canary Is, Tenerife, Galería de Ingleses, Icod de los

Vinos, 595 m; 17.v.1999, coll. G. I. E. T. (ULL 3166). Paratypes. Canary Is: 1�, same data as holotype, 21.i.2000; coll. G.

I. E. T. (ULL 1046) (spinnerets removed for SEM). 1�, Tenerife, Icod de losVinos, Galería de Breveritas, 10.vi.1999, coll. G. I. E. T. (CRBA 961) (rightleg1 removed for SEM). 1 juv., Cueva de Felipe Reventón, 595 m, 9.iii.2000,coll. G. I. E. T. (CRBA 3185) (freezer-collection).

DiagnosisDysdera sibyllina, sp. nov. differs from other troglobitic speciesof Dysdera inhabiting the lava tubes in northern Tenerife by thefollowing characters: from D. unguimmanis and D. esquiveli byits spinated anterior femora, and from D. ambulotenta and

D. labradaensis by its much smaller size (average carapacelength 2.30 mm versus average carapace length >7.30 mm).

DescriptionHolotype maleFigs 5L, 12A–C, 13A–G. Carapace (Fig. 12A) 2.19 mm long;

maximum width 1.70 mm; minimum width 1.17 mm. Orange,anteriorly darker, becoming lighter posteriorly; smooth withsome small black granules mainly anteriorly; hairy, coveredwith black hairs mainly at lateral and posterior borders. Anteriorborder roughly round, from 1/2 to 3/5 carapace length; anteriorlateral borders convergent; rounded at maximum dorsal width,posterior lateral borders rounded; posterior margin narrow,straight. Eyeless (Fig. 5L). Labium trapezoid-shaped, basewider than distal part; as long as wide at base; straight tipmargin. Sternum yellow, darkened on borders; smooth; uni-formly covered in slender black hairs.

Chelicerae (Fig. 12B) 1.02 mm long, ~1/3 of carapace lengthin dorsal view; fang medium-sized, 0.69 mm; basal segmentproximal dorsal side scantly covered with setiferous granula-tions. Chelicera inner groove medium-size, ~2/5 chelicerallength; armed with three teeth and lamina at base; D>M>B; Dtrapezoid, located roughly at centre of groove; B close to basallamina; M close to B. Anterior legs dark orange, posterior legsyellow. Lengths of male described above: fe1 2.29 mm (all mea-surements in mm); pa1 1.27; ti1 2.09; me1 1.84; ta1 0.66; total8.16; fe2 2.19; pa2 1.17; ti2 2.04; me2 1.84; ta2 0.61; total 7.85;fe3 1.99; pa3 0.87; ti3 1.38; me3 1.68; ta3 0.56; total 6.48; fe42.19; pa4 1.02; ti4 1.84; me4 2.14; ta4 0.71; total 7.90; fe Pdp1.20; pa Pdp 0.51; ti Pdp 0.61; ta Pdp 0.61; total 2.93; relativelength: 1>4>2>3. Spination: all femora spinated, fe1 3 terminalspines on forward margin; fe2 2 terminal spines on forwardmargin. Fe3d spines in one row; 2; pa3 spineless; tb3d spinesarranged in two bands: proximal 1.0.1; distal 1.0.1.; tb3v spinesarranged in two bands: proximal 1.0.1; distal 1.0.1; with two ter-minal spines. Fe4d spines in one row; 6; pa4 spineless; tb4dspines arranged in two bands: proximal 1.0.1; distal 1.0.1; tb4vspines arranged in two bands: proximal 1.0.1; distal 1–0.0.0;with two terminal spines. Dorsal side of anterior legs smooth;ventral side of pedipalp smooth. Claws with more than 15 teeth;length twice claw width.

Abdomen 2.04 mm long; grey; cylindrical. Abdominaldorsal hairs 0.09 mm long; medium-sized, roughly straight,not compressed, blunt, tip not enlarged; uniformly, thicklydistributed.

Table 9. Intraspecific spination variability in D. sibyllina, sp. nov.


Tibia 3 dorsal 1.0.1 0 0 1.0.1Tibia 4 dorsal 1.0–1.1 0 0 1.0–1.1Tibia 3 ventral 0–1.0.1 0 0 0–1.0.0–1Tibia 4 ventral 0–1.0.1 0 0 0–1.0.0–1

Number of rows Number of spines

Femur 1 prolateral 1 3Femur 2 prolateral 1 1–2Femur 3 dorsal 1 0–2Femur 4 dorsal 1 5–6


Male copulatory bulbus (Fig. 12C) T slightly longer thanDD; external distal border straight; internal sloped backwards.DD bent ~45° in lateral view; internal distal border notexpanded. IS, ES equally developed; IS continuous to tip. DDtip (Figs 13A–E) straight in lateral view; posterior (lower) sheetprojected under anterior (upper) one, pointed. C substituted bya row of small teeth. AC absent. LF absent. L well developed;external border not sclerotised, not folded; distal border diver-gent, continuous. LA absent. F absent. AL absent. P fused to T;markedly sloped on its proximal part, perpendicular on distal;lateral length from 2/3 to as long as T width; ridge present, per-pendicular to T; not expanded, upper margin markedly toothed,on its distal part; few teeth (4–6); not distally projected; poste-rior margin not folded.

Paratype femaleFigs 12E–F, 13G–H. All characters as in male except: cara-

pace 2.35 mm long; maximum width 1.84 mm; minimum width1.17 mm. Sternum yellow, anteriorly darker, becoming lighterposteriorly.

Chelicerae 1.15 mm long, ~2/5 of carapace length in dorsalview; fang medium-sized, 0.37 mm. Chelicera teeth: D=B>M;D triangular. Legs orange. Lengths of female legs describedabove: fe1 2.40 mm (all measurements in mm); pa1 1.27; ti12.09; me1 1.99; ta1 0.71; total 8.47; fe2 2.35; pa2 1.27; ti2 2.04;me2 1.96; ta2 0.71; total 8.34; fe3 1.89; pa3 0.92; ti3 1.53; me31.84; ta3 0.66; total 6.83; fe4 2.35; pa4 1.07; ti4 1.84; me4 2.27;ta4 0.76; total 8.29; fe Pdp 1.17; pa Pdp 0.61; ti Pdp 0.56; ta Pdp0.82; total 3.16; relative length 1>2>4>3. Spination: fe1 3 ter-minal spines on forward margin; fe2 one terminal spine on theforward margin. Fe3d spines in one row, 0–1; pa3 spineless;tb3d spines arranged in two bands: proximal 1.0.1; distal 1.0.1.;tb3v spines arranged in one band: proximal 0.0.1; with two ter-minal spines. Fe4d spines in one row: 5; pa4 spineless; tb4dspines arranged in two bands: proximal 1.1.1; distal 1.0.1; tb4vspines arranged in one band: proximal 0.0.1; with two terminalspines. Dorsal side of anterior legs smooth; ventral side of pedi-palp smooth. Claws with more than 15 teeth; hardly larger thanclaw width (Fig. 13F–G).

Abdomen 2.81 mm long; whitish; cylindrical. Abdominaldorsal hairs 0.13 mm long; medium-sized, curved, compressed,pointed; uniformly, thickly distributed. Vulva DA not distin-guishable from VA (Fig. 12E); arch-like, anteriorly rounded; DAslightly wider than long; DF wide in dorsal view. MF marginsfused, sheet-like, well developed, completely sclerotised. VAanterior region completely sclerotised; posterior region sclero-tised in most anterior area (Fig. 12F); tooth-shaped expansionfrom internal posterior border; joined to lateral sclerotisation,along its lateral border, as long as DF lateral margins; AVDcleraly recognisable. S attachment not projected under VA; armsgreatly reduced almost absent; neck as wide as arms. TB usualshape. ALS without PS (Fig. 13G); remaining piriform spigotsmore external than MS, arranged in one row; 2+0 piriform glandspigots; PMS, PLS with fewer than five aciniform gland spigots(Fig. 13H).

Intraspecific variationThe carapace ranges in length from 2.19–2.35 mm. Spinationvariability is shown in Table 9.

DistributionOnly recorded from deep lava tube galleries in the Icod de losVinos region (north central Tenerife)

DNA sequencesMitochondrial rrnL (GenBank accession numbers EU068079,ULL3185; EU068063, UL3166) and cox1 (EU068047,ULL3185) DNA sequences from the Galería de los Ingleses andBreveritas lava tubes.

Phylogenetic relationshipsParsimony and Bayesian analyses of the combined morpho-logical and molecular data matrix support D. sibyllina, sp. nov.as the sister-species to D. andamanae. A previous phylogeneticanalysis also including taxa only scored for morphological char-acters (Arnedo et al. 2001) supported D. andamanae as thesister-species to D. minutissima from Tenerife, suggesting thatD. sibyllina, sp. nov. could also be closely related to the latterspecies. External relationships of this clade are problematic.Parsimony suggests it is related to a clade including endemicspecies from the eastern Canary Islands, although with lowsupport, whereas a clade including D. sibyllina, sp. nov.,D. andamanae and the eastern endemics along with an endemicspecies from Madeira is assigned a high (0.99) Bayesian poste-rior probability.

EtymologyThe name refers to the sibyls (Latin Sibylla), prophetesses of theancient Greek world that lived in caves. One of them, theCumaean Sibyl, led Aeneas to the Underworld to visit hisdeceased father.

Dysdera silvatica Schmidt(Fig. 5A–B)

Dysdera silvatica Schmidt, 1981: 89–90. – Arnedo, 2003: 146.Dysdera rugichelis sensu Arnedo & Ribera, 1997: 267–270, figs 23A–F,

24A–D, 25A–C. Not Dysdera rugichelis Simon, 1907(Misidentification).

Dysdera ?rugichelis sensu Wunderlich, 1987: 57, fig. 18(Misidentification).

Material examinedHolotype. 1 juv. Canary Is, La Gomera, Mte. del Cedro, vi.1976, coll.

G.Schmidt (SMF 34583). Additional material examined. Canary Is: 1�, Cueva de Longueras,

Frontera, El Hierro (ULL 3842); 1�, (ULL 3917); 1 juv. (ULL4280). 1�, El Golfo, Fuente Mancafite (ULL 1825).

Tissue collection. Canary Is: 1�, Hoya del Pino, Frontera, El Hierro(CRBA 1341).

DistributionFormerly recorded in epigean localities in the islands of LaGomera and La Palma. Here we document the first record of thespecies in El Hierro, both in epigean habitats and lava tubes.

RemarksDifferences between the new population in El Hierro and thosein La Gomera and La Palma are mainly somatic: specimens



from El Hierro are larger (compare range in carapace length:El Hierro 6.00–6.30 mm to La Palma and La Gomera3.50–5.90 mm). Cave-dwelling specimens show eye reduction(roughly half the size of epigean specimens, Fig. 5A–B), anddiffering relative leg sizes: 1 and 2 are the longest in cave-dwellers and 1 and 4 longer in epigean specimens. Cheliceraeare also relatively longer. Spination patterns in El Hierrospecimens fall within the variability reported for other islandpopulations, and no apparent differences are observed in thegenitalia.

DNA sequencesMitochondrial rrnL (GenBank accession numbers AF244178,AF244177) and cox1 (AF244274, AF244273) DNA sequencesfrom La Palma and La Gomera were reported in Arnedo et al.(2001). Here we report new sequences of these genes for ElHierro population (rrnL EU068072, cox1 EU068038).

Phylogenetic analysesParsimony and Bayesian inference analyses agree to supportD. enghoffi from La Gomera as the sister-species to D. silvatica,and suggest, although with low support (parsimony bootstrapsupport 55, Bayesian PP <0.95), that the population from ElHierro is more closely related to the population from La Gomerathan to the population from La Palma.

DiscussionCanarian troglobitic members of DysderaAs many as 10 species of Dysdera endemic to the Canaries havebeen exclusively collected in subterranean habitats. TroglobiticDysdera species are restricted to Tenerife and La Palma. Thepaucity of troglobitic populations in La Gomera and GranCanaria Islands is a general pattern shared with other endemicgroups of arthropods and it is most likely the result of thescarcity of volcanic caves owing to the geological features ofthese islands, MSS being the only suitable subterranean habitats(Oromí et al. 1991; Oromí and Izquierdo 1994; Oromí 2004).The eastern Canary Islands have many lava tubes resulting fromrecent posterosive volcanism. The scarce fauna collected in thesubterranean habitats of the eastern islands is almost devoid oftroglobites, probably owing to the dry environment in thesecaves (Martín 1991). In El Hierro island, where a rich troglobiticfauna has been described (Medina 1991; Martín 1992), Dysderaspecimens belonging to species extensively collected in epigeanhabitats (D. silvatica and D. gomerensis) have been collectedfrequently in lava tubes and the MSS. At least two specimensidentified as D. silvatica collected from a lava tube in El Hierroshowed reduction in eye size (Fig. 5), but no significant dif-ferences in morphological measurements compared withepigean specimens. El Hierro is the youngest of the CanaryIslands, with an estimated age of 1.12 Mya (Guillou et al. 1996),which has been frequently cited as a reason to explain thesmaller number of troglobites reported in this island whencompared with the older La Palma and Tenerife. A similarexplanation, namely recent colonisation of the undergroundenvironment, may account for the eye size polymorphismreported in D. ratonensis from La Palma (2 Mya; Carracedo andDay 2002), the second youngest island, as suggested by the low

genetic divergence observed in the cox1 among localities(Table 2). In this case, however, eye reduction was accompaniedby significant differences in the length of appendages.

Most troglobitic Dysdera species have been collected inmore than one lava tube, frequently tens of kilometres apart.Extreme examples are the localities of Cueva de la Fajanita andCueva del Ratón in La Palma (D. ratonensis), and Cueva Grandede Chío and Cueva Honda de Güímar (D. chioensis) in Tenerife.In both cases the caves were located more than 40 km apart.Dysdera gollumi is the only species reported from a single lavatube, although D. sibyllina has been collected in localities thatbelong to the same lava tube system (Icod). Some of the lavatubes inhabited by troglobitic Dysdera species are extremelyyoung (e.g. Cueva del Ratón was formed during the eruption ofthe San Antonio volcano in 1677 AD). In general, lava tubeshave short geological life spans, usually collapsing or filling upwith sediments after 0.03 to 0.5 Mya (Howarth 1973). Even inthe absence of caves, all the islands have a rich undergroundenvironment in the form of shallow, intermediate-sized, inter-connected voids (i.e. MSS) (Oromí et al. 1991; Oromí andIzquierdo 1994), where troglobitic taxa have been collected. Inthe case of Tenerife and La Palma the MSS develops into a con-tinuous habitat that covers most of these islands’ surface, prob-ably with the sole exceptions of the older terrains where it hasbeen completely silted up by sediments and erosive processes(Martín 1992). Epigean Dysdera species have frequently beencollected in MSS traps (Medina 1991). Nevertheless, we reporthere the first direct evidence that troglobitic Dysdera species arealso found in the MSS. A single specimen of D. esquiveli, aspecies formerly known from several lava tubes in Tenerife(Table 3), was collected in an MSS pitfall trap. In addition,D. madai is the first Dysdera species exclusively known fromMSS localities. These data suggest that troglobitic species ofDysdera inhabit and are able to disperse through the network ofcracks and voids between lava tubes, as has been proposed forother troglomorphic taxa (Howarth 1983; Oromí et al. 1991;Hoch and Howarth 1999). Subterranean migration throughmicro- and mesocaverns has also been demonstrated in theAustralian cave pseudoscorpion genus Protochelifer, althoughit is probably restricted to closely located caves (up to 6 km)(Moulds et al. 2007). Sampling of the Canarian MSS has beenmostly limited to shallow depths (~1 m from the surface), owingto the inherent logistic difficulties. Restricted sampling mayexplain the paucity of captures of troglobitic members ofDysdera and the overrepresentation of epigean specimens inmost MSS samples analysed to date.

Cave colonisationColonisation of the underground environment is the first steptowards the evolution of an exclusive subterranean life(Christiansen 1992). Habitat preferences in Dysdera, nocturnalhunters in damp but warm ground habitats, fit in well with theconditions found in the subterranean environment. EpigeanDysdera specimens are frequently collected in cave entrancesand in the MSS. In the Canaries, about half of the endemicDysdera species have been, either sporadically or extensively,collected in subterranean habitats (Fig. 2). There are manyexamples of troglobitic species reported in the family andseveral genera are, in fact, exclusively known from cave-


dwelling representatives (e.g. Minotauria Kulczynsky, 1903;Stalita Schiödte, 1948; Folkia Kratochvil, 1970; StalagtiaKratochvil, 1970; Speleoharpactea Ribera, 1982; or Sardo-stalita Gasparo, 1998). These data suggest that the Dysderidaegroundplan includes preadaptations that may facilitate colonisa-tion and survival in the underground environment. It has beenobserved, for instance, that the epigean species Dysdera eryth-rina (Walckenaer, 1802) displays a remarkable reduction of the

brain optical lobes when compared with other spider groups(A. Lopez, pers. comm.). In addition, air bubbles held in the silknest of Dysdera species have been shown to act as physical gillswhen submerged, which may have helped members of the genussurvive flooding in both edaphic and subterranean habitats(Rovner 1986).

The existence of morphological, physiological or behavi-oural characters in the Dysdera body plan that facilitate cave


2

0.5

0.2

0.5

A

B

C

D

E

F

Fig. 12. Dysdera sibyllina, sp. nov. A, carapace, dorsal view; B, left chelicera, ventral view; C, left male palp, pro-lateral view; D, vulva, dorsal view; E, vulva, ventral view; F, vulva, lateral view.


A

C

B

D E

F G

H I

Fig. 13. Dysdera sibyllina, sp. nov. right male bulb. A, prolateral view; B, posterior apophysis, prolateral view; C,anterior view; D, posterior view; E, retrolateral view; F, walking leg tarsus, anterior view; G, walking leg tarsus,ventral view. Spinnerets, G: right anterior lateral spinnerets spigots; H: left posterior lateral spinnerets spigots.


colonization is further supported by the recurrent colonisationof the subterranean habitats in the Canaries. Phylogeneticanalyses support eight independent colonisations of the under-ground environment by troglobitic species. Multiple inde-pendent invasions of caves by closely related taxa have beenreported in springtails (Christiansen and Culver 1968), Bathy-sciinae beetles (Sbordoni 1982), Astyanax fishes (Dowling et al.2002; Strecker et al. 2003), and plethodontid salamanders(Wiens et al. 2003).

Evolution of cave lifeThe patterns of geographical distribution of subterranean andsurface taxa allow a preliminary test of the predictions of thetwo competing hypotheses on the evolutionary origin of troglo-bites. The climatic relict hypothesis (CRH) assumes fragmenta-tion of a widespread epigean species driven by climatic changes,and isolation of small populations in cave refugia, where diver-gence is mainly driven by genetic drift (Barr 1968; Sbordoni1982; Barr and Holsinger 1985). On the other hand, under theadaptive shift hypothesis (ASH) the extirpation of epigean pop-ulations is not a necessary condition to attain genetic isolation,inasmuch as gene flow is prevented by resource segregation andecological requirements (Howarth 1981, 1986, 1987; Howarthand Hoch 2005). Specific predictions on the geographical distri-bution patterns of epigean and subterranean sister-taxa areeasily derived from the former formulations: under CRH, sisterepigean and subterranean sister-taxa are necessarily allopatricand subterranean species frequently have a relictual distri-bution, whereas under ASH epigean and subterranean species,pairs may show parapatric distributions.

All Dysdera species from the western and central Canariesshare an exclusive common ancestor and, hence, they cannot beconsidered as relict lineages. The species D. andamanae,sp. nov. and D. sybyllina constitute an exception. Although theirphylogenetic position is uncertain – they are shown as related tothe species endemic to the eastern Canary Islands, though withvery low support – they are clearly the closest relatives to eachother (probably also including D. minutissima, see below) butstill show contrasting patterns of habitat specialisation, that is,epigean versus subterranean. Removal of above-ground popula-tions does not seem to be a requirement for speciation of cavespecies, since eight out of ten Canarian troglobitic species ofDysdera have their closest epigean relatives on the same island.In fact, one of the two exceptions, D. sybyllina may also have itsclosest relative in the same island, since morphologically closeand island neighbour D. minutissima could not be included inthe molecular analyses. It could be argued that the allopatry ofthe epigean and subterranean population was a result of intrais-land vicariant processes. Fine scale distributional patterns ofepigean and subterranean taxa seem for the most part to denythis possibility. Dysdera gibbifera partially overlaps distribu-tional ranges with D. labradaensis and D. ambulotenta – it hasactually been collected in the same lava tube. The same patternholds for D. iguanensis and D. madai; both species have beencollected in the same MSS trap. Dysdera brevisetae andD. guayota show partially parapatric distributions with regard totheir subterranean sister-taxa D. esquiveli/D. hernandezi andD. chioensis, respectively. On the other hand, the species pairmade up by D. levipes and D. gollumi is a genuine example of

epigean/subterranean sister-taxa with allopatric distributions:D. levipes is distributed through the north-eastern part ofTenerife, whereas D. gollumi is restricted to the Cueva de LosRoques at more than 2000 m altitude on the Teide volcano. Still,this pattern may prove to be an artefact resulting from under-sampling. The troglobitic species D. ambulotenta has been col-lected in Cueva de Los Roques but also in other lava tubes on thenorthern part of Tenerife, not far from where D. levipes has beencollected. The distributional range of D. ambulotenta suggeststhat high altitude caves could be connected through the MSSwith caves in the lower parts of the northern region and, hence,that D. gollumi could have been originally present in thenorthern caves. The presence of many other troglobites like thespider Troglohyphantes oromii, the woodlouse Venezillo tener-ifensis and several species of the beetle genera Wolltinerfia,Canarobius and Domene both in Cueva de los Roques and in thelow altitude caves of the northern part also supports this hypoth-esis (Martín and Oromí 1986; Arechavaleta et al. 1998).

Dysdera ratonensis and D. unguimmanis are the only speciesthat depart from the general model of parapatric distribution ofepigean and subterranean sister-taxa, although probably owingto different circumstances. The closest relatives of D. ratonensisinhabit Tenerife and Gran Canaria. The species was originallydescribed as a troglobite restricted to the lava tubes of the drysouthern and south-western slopes of La Palma (Wunderlich1991; Arnedo and Ribera 1996), which are also the youngestand show more active volcanism (Ancochea et al. 1994).Subsequently, the species has also been found in lava tubes inthe humid northern slopes of the island, which are the oldest andmore humid. Interestingly, specimens from the northern locali-ties have eyes in the size range of other epigean species, whereasindividuals from the south show distinct eye reduction (Fig. 5).It seems safe to assume that the ancestor of D. ratonensis hadregular eyes by the time the colonisation of La Palma took placeand, hence, that eye reduction was a postcolonisation process.The geographical structure of eye polymorphism in D. ratonen-sis may suggest different ecological adaptations: the populationin the south, where the climate is dryer, would be confined tounderground habitats, whereas in the north, where surface con-ditions are more suitable for Dysdera species, the populationwould be less dependent on subterranean conditions. In contrastto former suggestions (Arnedo et al. 2001), D. unguimmanisclearly belongs to the western-central clade, where it is shownas the sister-taxa to a large clade of species, albeit with very lowsupport. Therefore, in the specific case of D. unguimmanis, spe-ciation by extinction of the epigean relative cannot be com-pletely ruled out.

Based on geographical distributions, our data seemingly pro-vides support for the ASH. Similar results have also been foundin Canarian beetles of the genus Trechus (Contreras-Díaz et al.2007), as well as in cave-dwelling isopods from the HawaiianIslands (Rivera et al. 2002), beetles in the Galapagos (Peck andFinston 1993), and land snails in Borneo (Schilthuizen et al.2005). Conversely, phylogenetic studies on troglobionts in con-tinental areas have tended to favour the CRH (Leys et al. 2003;Allegrucci et al. 2005). These results are to be expected, giventhe latitudinal gradient in the influence of climate oscillationson the Earth’s ecosystems. Molecular evidence, however,suggest that diversification of some troglobitic lineages may



have preceded the Pleistocene, which casts some doubts on therole of climatic changes on cave speciation (Hedin 1997b).Glaciations had a mild effect on the Canaries, mainly restrictedto forest regression during drier periods (Criado 1984).Frequent fluctuations of wet and arid conditions related toglaciations have been reported for the region (Sarthein andKoopman 1980; Sarthein et al. 1981; Petit-Maire et al. 1986;Rognon et al. 1989; Criado and Hansen 1994). Desutter-Grandcolas and Grandcolas (1996) have convincingly arguedthat the criteria used to define CRH and ASH are questionablesince only the present-day distribution of troglobitic taxa andepigean relatives is observable. In oceanic islands volcanicactivity has most likely had a profound impact on the distri-bution of endemic taxa. Volcanism may have played the role ofclimate change in temperate regions by promoting cave specia-tion through removing epigean lineages. However, subterraneanresources are ultimately dependent on energy inputs from exter-nal sources, and the complete destruction of surface ecosystemswould have had a major impact on the corresponding subter-ranean milieu (Howarth 1983; Culver et al. 2006). Alternatively,we suggest that the main role of volcanism is the renewal of thesubterranean habitat, creating new lava tubes and volcanic MSSthat would be subsequently colonised through active dispersalof troglobites from adjacent areas. The aforementioned patternsexhibited by D. ratonensis may illustrate this proposition. Thisis an important consideration since the absence of epigean pop-ulations should therefore not be interpreted as a proof forallopatric speciation, but as the result of subterranean rangeexpansion after parapatric speciation into regions where surfaceenvironment is still too harsh for epigean populations to settle.

Adaptation to subterranean lifeTroglobitic taxa are frequently characterised by a set of morpho-logical and life history features supposedly shared because of thestrong and similar selective pressures acting in subterraneanhabitats (Christiansen 1992). Morphological adaptations to lifein caves include a progressive elongation of body andappendages, global increase of body size, and reduction of bodypigmentation and eye structures (Culver 1982; Sbordoni 1982).Fage (1931) listed the reduction in the number of eggs (but theirincrease in size), along with eye regression, depigmentation, andappendage elongation as among the traits that would characterisecave spiders. However, a cause-effect relationship between cavelife and the evolution of the former characters is far from beinguniversally accepted (Desutter-Grandcolas 1994; Desutter-Grandcolas 1997b). Desutter-Grandcolas (1997a) used a phylo-genetic test of troglobiomorphism to demonstrate that in thecricket clade Amphiacustae, eye reduction and depigmentationwere strictly associated with cave life but refuted this link forappendage elongation (hindleg size). On the other hand, Miller(2005) found that loss of eyes, elongation of the legs and reduc-tion of the tracheal system characterised troglobitic erigoninaespiders of the genus Anthrobia. Canarian Dysdera species showsimilar patterns of character association to cave life, when com-pared with Anthrobia. Eye regression is the only somatic charac-ter shared across all troglobitic Canarian Dysdera species(Fig. 5). The reduction in the diameter of the eye lens is alsoevident in cave specimens of species otherwise mostly found inepigean habitats, e.g. D. silvatica. Although, we did not perform

quantitative analyses of eye size, it is self-evident that there aredifferent levels of eye regression. Dysdera chioensis,D. labradaensis and D. madai show clearly reduced eyes,although all of eyes are present. On the other hand, in D. ambu-lotenta, D. esquiveli, D. hernandezi, D. gollumi, D. sybyllina, andD. unguimmanis, the eyes are smaller or have been lost.Dysderidae species have six indirect eyes, but lack the direct(anterior median) eyes found in most spiders. The indirect eyesusually have a tapetum to increase visual sensitivity, giving theseeyes a silvery appearance (see Fig. 5A). Unlike epigean species,troglobitic species’ eyes are never reflective, suggesting that themodifications also affect the internal anatomy of the eye. A mod-erate reduction in eye size can also be observed in some speciesfrequently, but not exclusively found in caves, such as, forexample, D. gibbifera, whose eyes are similar to those found insome populations of the troglobitic D. ratonensis. The presenceof eye regression in all troglobitic species is not surprising.Molecular genetic estimates of divergence suggest that the lossor substantial reduction of eyes and pigmentation can take placein just one hundred years (Caccone and Sbordoni 2001).

Contrary to the observed pattern of eye regression, body sizeand appendage elongation do not always increase in troglobiticCanarian Dysdera species. In only two out of the nine compar-isons (one involving populations of the same species), troglobitictaxa were significantly larger than their epigean counterparts.The troglobitic species D. esquiveli, D. hernandezi and D. madaiwere, in fact, smaller, although the values were significant onlyin the latter case. The PCA plot of morphological measurementsshows a clear pattern of separation of epigean and troglobiticsister pairs along the main axis, which roughly corresponds toappendage length. In fact, most of the subterranean/epigeansister lineage comparisons showed significant support forappendage elongation in troglobites, especially in the first leg. Insome species, however, the differences were not significant andin one example the epigean species had longer anterior legs (e.g.D. guayota). This pattern adds support to former propositionsthat certain features traditionally associated with adaptation tothe subterranean environment are in fact plesiomorphies, andthus better considered as exaptations (Desutter-Grandcolas1994; Desutter-Grandcolas 1997a, 1997b).

Other characters, which were not analysed quantitatively,also seem, however, to be exclusive to troglobitic CanarianDysdera species. Kuntner and collaborators (1999) showed thattroglobitic continental Dysderidae species show a significantreduction of their respiratory system. A strong reduction in thenumber of tracheoles has also been observed in troglobiticAnthrobia spiders (Miller 2005). The examination of singleindividuals of the species D. ambulotenta, D. esquiveli, D. rato-nensis, and D. unguimmanis revealed a reduction of the respira-tory system similar to that observed in the former examples(Kuntner personal communicatiion). Trogobites usually havelower respiratory rates, which has been interpreted as evidenceof a decrease in metabolic rates associated to limited energyresources or high levels of CO2 (Kuntner et al. 1999). In addi-tion, we have observed a consistent reduction in the number ofspigots in all spinnerets in troglobitic Canarian Dysdera species,when compared with epigean sister-taxa. A reduction in theamount of the silk used to wrap the eggs has been observed inthe cave linyphiid spider Anthrobia (Poulson 1981). The reduc-


tion in the costly production of silk is consistent with a lowmetabolic rate. The tarsal claws of D. unguimmanis and D. sibyl-lina show a remarkable development in size and number ofteeth, a character also reported in troglobitic continental dys-derids (Arnedo and Ribera 1999), collembolans (Christiansen1961) and cixiid planthoppers (Howarth 1991).

Troglobitic Canarian Dysdera species show different combi-nations of eye regression, body size, appendage elongation, andtarsal claw development. Oromí and collaborators (1991) sug-gested that the former pattern might be the result of independentcolonisations of the underground environment over time,leading to morphologically close, yet different species withvarying degrees of troglomorphism. There is no consensusamong researchers as to the correlation between the degree ofmorphological modifications exhibited by a troglobite and thetime elapsed since this taxon evolved to become a cave dweller(Howarth 1986, 1987; Wilkens and Hüppop 1986). If we acceptthat genetic divergence among subterranean and epigean sister-taxa provides a rough approximation to the time of cave coloni-sation, then troglobite Canarian Dysdera species do not supporta close relationship between the time of cave colonisation andthe level of troglobiomorphism. Similar patterns have beenreported in Oliarus planthoppers in the Hawaiian Islands (Hochand Howarth 1999). In the Canaries, the troglobitic cockroachesLoboptera anagae and L. cavernicola from Anaga, are lesstroglobiomorphic but probably older than some of the cave-dwelling species from central Tenerife (Izquierdo 1997).Additional examples may be found in the beetle generaDomene, Wolltinerfia and the planthopper Tachycixius, whichhave eyeless species in central Tenerife and different specieswith reduced eyes in Anaga (Oromí et al. 1991). The existenceof heterogenous selective regimes could provide an alternativeexplanation for the different levels of troglobiomorphismreported. The underground environment is far from beinguniform. Darkness, stability, humidity, percentage of CO2 andthe amount of available nutrients, among other parameters,divide the subterranean compartment into a series of several dif-ferent habitats (Howarth 1991, 1993). We suggest that the dif-ferent levels of troglobiomorphism do not reflect successivestages towards a better adaptation to the underground con-ditions, but are better interpreted as adaptations to particularhabitats of this environment. Selective partitioning of coexistingspecies in different ecological niches has been proposed toexplain recurrent sympatry among cave species belonging todifferent genera in Bathysciine beetles (Sbordoni 1982;Caccone and Sbordoni 2001). Additional examples of Canariantroglobitic taxa with syntopic species showing different levels oftroglobiomorphism include Loboptera troglobia (more troglo-biomophic than sympatric L. subterranean), L. chioensis, andL. penirobusta; or the beetle Domene vulcanica, which overlapsin two lava tubes with the less troglobiomorphic D. alticola.

Community assembly in the undergroundCanarian cave-dwelling Dysdera species have largely overlap-ping geographical distributions (Table 3). Up to five troglobiticDysdera species are found in Cueva del Viento-Sobrado, andlava tubes with a single species, such as Cueva Grande de Chío,are the exception to the rule. The genus Dysdera shows a widerange of body sizes and chelicerae modifications (Deeleman-

Reinhold and Deeleman 1988). Dysdera species are activehunters that grasp their prey with the chelicerae and, hence,body and chelicerae sizes are important indicators of the rangeof potential prey. Morphological (Bristowe 1954; Bristowe1958; Cooke 1965) and physiological (Hopkin and Martin1985) evidence suggests that woodlice constitute the naturaldiet of members of Dysdera, although prey preference experi-ments have failed to prove this point (Pollard et al. 1995). Theelongation of chelicerae in the continental species D. erythrinaand D. crocata is considered an adaptation to capture terrestrialisopods (Bristowe 1958; Pollard 1986). Sympatric troglobiticspecies of Dysdera show a segregation of body size groups(Tables 3 and 6), for example, D. ambulotenta, D. chioensis,and D. gollumi in Cueva de los Roques. Multi-species lava tubesare generally inhabited by large, medium-sized and smallspecies (e.g. D. ambulotenta, D. chioensis and D. gollumi inCuieva de los Roques). When more than one species of thesame body size group coexist in the same locality, the speciesare further differentiated by cheliceral modifications (Fig. 3B).Examples of size segregation in other Canarian arthropodsinhabiting lava tubes include Loboptera troglobia (which issmaller than sympatric species L. subterranean, L. chioensisand L. penirobusta), and the staphylinid beetle Alevonota out-ereloi (which is larger than syntopic A. canariensis and muchlarger than A. oromii in Tenerife), whereas in La Palma the sym-patric A. junoniae, A. tanausui and A. hephaestos also showthree different size categories.

Differentiation in body size and cheliceral modificationssuggest that prey segregation plays a key role in shaping under-ground communities in Canarian Dysdera species. Phylogeneticinformation provides insights into the role played by evolutionin shaping communities (Losos et al. 1998; Gillespie 2004). Thespecies D. madai and D. sibyllina are the only Canarian Dsyderatroglobites that show significant changes in cheliceral morphol-ogy compared with their epigean sister-taxa. Cheliceral modifi-cations in these species may be the result of competitivecharacter displacement (Schluter 2000). Dysdera madai is theonly troglobite that coexists with its epigean sister-taxa,whereas the epigean sister-species of D. sibyllina, D. anda-manae, share similar cheliceral morphology with troglobiticD. esquiveli, which coexists with and has similar body size asD. sibyllina. In Canarian Dysdera species, competition may playan overriding role in dictating body size because of the strongcorrelation between body size and resource use. Character dis-placement through interspecific competition may explain thesignificant reduction in size of D. esquiveli and D. hernandeziwhen compared with their epigean sister-taxa, an effect thatruns counter to the general trend towards larger size in cavespecies. These two partially sympatric sister-species havealmost identical morphologies except for one character: D. her-nandezi has flattened cheliceral fangs. This character hasevolved independently in several Canarian and continentalspecies (Deeleman-Reinhold and Deeleman 1988; Arnedo et al.2001) and seems to be related to a different prey capture strat-egy (M. Rezac, pers. comm.).

ConclusionsThe spider genus Dysdera has undergone a remarkable processof local diversification in the Canary Islands, offering an ideal



model for the study of speciation, morphological adaptation andcommunity assembly. In this paper, we summarise the availableinformation on Canarian troglobitic Dysdera species anddiscuss its relevance for understanding cave life evolution. Wesuggest that the Dysderidae groundplan represents a preadapta-tion to colonise and inhabit the subterranean environment. Wefind evidence for a predominantly parapatric mode of speciationin our study groups, although we cannot entirely rule out anallopatric process due to the confounding effects of postspecia-tional range. All troglobitic species display a certain level of eyeregression, whereas an increase in body size and appendageelongation is common but not universal. We suggest that dif-ferent levels of troglobiomorphism are better interpreted interms of adaptation to local environmental conditions based onthe heterogeneous nature of subterranean habitats, rather than asthe result of differences in the time of colonisation. TroglobiticCanarian species of Dysdera have colonised the undergroundenvironment on at least eight independent occasions. Wehypothesise that trophic segregation is responsible for the highlevel of sympatry between troglobites, and that changes in preycapture strategy may explain the only supported example of spe-ciation in caves.

The study of troglobites in oceanic islands provides invalu-able data for a better understanding of the origins and evolutionof cave life, and improves our knowledge of the generation ofdiversity and the interaction between speciation, ecologicaladaptation and morphological change.

AcknowledgementsWe wish to thank Manuel Arechavaleta, Eduardo Muñoz, Nieves Zurita,Heriberto López, Salvador de la Cruz, Helena Morales and Antonio J. Pérez,who collected many of the specimens included in the present study, andRafael García who provided additional specimens. The paper deeply bene-fited from patient explanations and insightful discussions with FrankHowarth. We are in debt to Laure Desutter-Grandcolas, whose commentsand suggestions on an early draft of the manuscript greatly helped toimprove the final version. Gustavo Hormiga kindly provided access to theSEM facility at George Washington University. The Cabildos (local author-ities) in Tenerife, La Palma and El Hierro granted collecting permits, and thelatter two provided logistic facilities. MA was supported by the Ramón yCajal Programme funded by the Spanish Ministry of Education and Scienceand the European Regional Development Fund, and NM was supported bya graduate grant from the Autonomous Government of the Canary Islands.This project was funded by the Spanish Ministry of Education and Sciencegrants PB97–0937 (CR&PO) and BOS2003–05876 (MA&PO) and addi-tional financial support was provided through project 2005SGR00045 of theAutonomous Government of Catalonia. Part of the fieldwork was supportedby a LIFE-Nature project on the conservation of cave fauna, co-sponsoredby the Consejería de Medio Ambiente (Environmental Department) of theCanarian Autonomous Government, and project P1042004/047 also of thethe Canarian Autonomous Government.

ReferencesAkaike, H. (1973). Information theory as an extension of the maximum like-

lihood principle. In ‘Second International Symposium on InformationTheory’. (Eds B. N. Petrov and F. Csaki.) pp. 1–434. (Akademiai Kiado:Budapest, Hungary.)

Allegrucci, G., Todisco, V., and Sbordoni, V. (2005). Molecular phylogeog-raphy of Dolichopoda cave crickets (Orthoptera, Rhaphidophoridae): ascenario suggested by mitochondrial DNA. Molecular Phylogeneticsand Evolution 37, 153–164. doi:10.1016/j.ympev.2005.04.022

Alvarez Rixo, J. A., Díaz Alayón, C., and Tejera Gaspar, A. (1991).‘Lenguaje de los Antiguos Isleños.’ (Centro de la Cultura PopularCanaria: La Laguna, Tenerife, Spain.)

Ancochea, E., Hernán, F., Cendrero, A., Cantagrel, J. M., Fuster, J. M.,Ibarrola, E., and Coello, J. (1994). Constructive and destructive episodesin the building of a young oceanic island, La Palma, Canary Islands, andthe genesis of the Caldera de Taburiente. Journal of Volcanology andGeothermal Research 60, 243–262.

Arechavaleta, M., Zurita, N., Camacho, A. I., and Oromí, P. (1998). La faunainvertebrada de tres cavidades volcánicas del Parque Nacional del Teide(Tenerife): Los Roques, Cuevas Negras y Chavao. Revista de laAcademia Canaria de Ciencias 10, 65–78.

Arnedo, M. A., and Ribera, C. (1996). Dysdera ratonensis Wunderlich, 1991(Arachnida, Araneae) a troglomorphic species from La Palma, CanaryIslands: description of the male and redescription of the female. RevueArachnologique 11, 109–122.

Arnedo, M. A., and Ribera, C. (1997). Radiation of the genus Dysdera(Araneae, Haplogynae, Dysderidae) in the Canary Islands: the island ofGran Canaria. Zoologica Scripta 26, 205–243. doi:10.1111/j.1463-6409.1997.tb00413.x

Arnedo, M. A., and Ribera, C. (1999). Radiation of the genus Dysdera(Araneae, Dysderidae) in the Canary Islands: The island of Tenerife. TheJournal of Arachnology 27, 604–662.

Arnedo, M. A., Oromí, P., and Ribera, C. (1996). Radiation of the genusDysdera (Araneae, Haplogynae, Dysderidae) in the Canary Islands: TheWestern Islands. Zoologica Scripta 25, 241–274. doi:10.1111/j.1463-6409.1996.tb00165.x

Arnedo, M. A., Oromí, P., and Ribera, C. (2000). Systematics of the genusDysdera (Araneae, Dysderidae) in the eastern Canary Islands. TheJournal of Arachnology 28, 261–292. doi:10.1636/0161-8202(2000)028[0261:SOTGDA]2.0.CO;2

Arnedo, M. A., Oromí, P., and Ribera, C. (2001). Radiation of the spidergenus Dysdera (Araneae, Dysderidae) in the Canary Islands: Cladisticassessment based on multiple data sets. Cladistics 17, 313–353.doi:10.1111/j.1096-0031.2001.tb00129.x

Arnedo, M. A., Coddington, J., Agnarsson, I., and Gillespie, R. G. (2004).From a comb to a tree: phylogenetic relationships of the comb-footedspiders (Araneae, Theridiidae) inferred from nuclear and mitochondrialgenes. Molecular Phylogenetics and Evolution 31, 225–245.doi:10.1016/S1055-7903(03)00261-6

Ashmole, N. P. (1993). Colonization of the underground environment in vol-canic islands. Memoires de Biospeologie 20, 1–11.

Baker, R. H., and DeSalle, R. (1997). Multiple sources of character infor-mation and the phylogeny of Hawaiian Drosophila. Systematic Biology46, 654–673. doi:10.2307/2413499

Baker, R. H., Yu, X., and DeSalle, R. (1998). Assessing the relative contri-bution of molecular and morphological characters in simultaneousanalysis trees. Molecular Phylogenetics and Evolution 9, 427–436.doi:10.1006/mpev.1998.0519

Barr, T. C. (1967). Observations on the ecology of cave. AmericanNaturalist 101, 475–492. doi:10.1086/282512

Barr, T. C. (1968). Cave ecology and the evolution of troglobites.Evolutionary Biology 2, 35–102.

Barr, T. C., and Holsinger, J. P. (1985). Speciation in cave faunas. AnnualReview of Ecology and Systematics 16, 313–337. doi:10.1146/annurev.es.16.110185.001525

Bristowe, W. S. (1954). The chelicerae of spiders. Endeavour 13, 42–49. Bristowe, W. S. (1958). ‘The World of Spiders.’ (Collins: London, UK.)Caccone, A., and Sbordoni, V. (2001). Molecular biogeography of cave life:

a study using mitochondrial DNA from bathysciine beetles. Evolution55, 122–130.

Carracedo, J. C., and Day, S. (2002). ‘Canary Islands.’ (Terra Publishing:Harpenden, UK.)

Chapman, P. (1982). The origin of troglobites. Proceedings of the Universityof Bristol Speleological Society 16, 133–141.


Christiansen, K. (1961). Convergence and parallelism in caveEntomobryinae. Evolution 15, 288–301. doi:10.2307/2406229

Christiansen, K. (1962). Proposition pour la classification des animaux cav-ernicoles. Spelunca 2, 76–78.

Christiansen, K. (1992). Biological processes in space and time: cave life inthe light of modern evolutionary theory. In ‘The Natural History ofBiospeleology’. (Ed. A. I. Camacho.) pp. 453–480. (Museo Nacional deCiencias Naturales (CSIC): Madrid, Spain.)

Christiansen, K., and Culver, D. C. (1968). Geographical variation and evo-lution in Pseudosinella hirsuta. Evolution 22, 237–255.doi:10.2307/2406522

Coleman, C. O. (2003). “Digital inking”: how to make perfect line drawingson computers. Organisms, Diversity and Evolution 3, 303–304.doi:10.1078/1439-6092-00081

Contreras-Díaz, H. G., Moya, O., Oromi, P., and Juan, C. (2007). Evolutionand diversification of the forest and hypogean ground-beetle genusTrechus in the Canary Islands. Molecular Phylogenetics and Evolution42, 687–699. doi:10.1016/j.ympev.2006.10.007

Cooke, J. A. L. (1965). A contribution to the biology of the British spidersbelonging to the genus Dysdera. Oikos 16, 20–25. doi:10.2307/3564861

Criado, C. (1984). El relieve erosivo: las formas de modelado. In ‘Geografiade Canarias’. pp. 105–142. (Interinsular Canarias: Santa Cruz deTenerife, Spain.)

Criado, C., and Hansen, A. (1994). Morfodinamica litoral, torrencial y vol-canica durante el Pleistoceno final y Holoceno de Jinamar (GranCanaria, Islas Canarias). In ‘Geomorfologia en Espana’. (Eds J. ArnáezVadillo; J.M. García-Ruiz and A. Gómez Villar.) pp. 369–389. (SociedadEspanola de Geomorfologia: Logrono, Spain.)

Culver, D. C. (1982). ‘Cave Life.’ (Harvard University Press: Cambridge,MA, USA.)

Culver, D. C. (2001). Subterranean ecosystems. In ‘Encyclopedia ofBiodiversity’. pp. 527–540. (Academic Press: San Diego, CA, USA.)

Culver, D. C., Deharveng, L., Bedos, A., Lewis, J. J., Madden, M., Reddell,J. R., Sket, B., Trontelj, P., and White, D. (2006). The mid-latitude bio-diversity ridge in terrestrial cave fauna. Ecography 29, 120–128.doi:10.1111/j.2005.0906-7590.04435.x

Deeleman-Reinhold, C. L., and Deeleman, P. R. (1988). Revision desDysderinae. Tijdschrift voor Entomologie 131, 141–269.

Desutter-Grandcolas, L. (1993). The cricket fauna of Chiapanecan caves(Mexico): Systematics, phylogeny and the evolution of troglobitic life(Orthoptera, Grylloidea, Phalangopsidae, Luzarinae). InternationalJournal of Speleology 22, 1–82.

Desutter-Grandcolas, L. (1994). A phylogenetic test of troglobitic adap-tation in crickets (Insecta, Orthoptera, Grylloidea). Comptes Rendus del'Academie des Sciences. Serie III, Sciences de la Vie 317, 907–912.

Desutter-Grandcolas, L. (1997a). Are troglobitic taxa troglobiomorphic?A test using phylogenetic inference. International Journal of Speleology26, 1–19.

Desutter-Grandcolas, L. (1997b). Studies in cave life evolution: a rationalefor future theoretical developments using phylogenetic inference.Journal of Zoological Systematics and Evolutionary Research 35,23–31.

Desutter-Grandcolas, L., and Grandcolas, P. (1996). The evolution towardtroglobitic life: A phylogenetic reappraisal of climatic relict and localhabitat shift hypotheses. Memoires de Biospeologie 23, 57–63.

Dowling, T. E., Martasian, D. P., and Jeffery, W. R. (2002). Evidence formultiple genetic forms with similar eyeless phenotypes in the blindcavefish, Astyanax mexicanus. Molecular Biology and Evolution 19,446–455.

Fage, L. (1931). Araneae, 5e série, précédée d’un essai sur l’évolutionsouterraine et son déterminisme. In Biospeologica, LV. Archives deZoologie Expérimentale et Générale 71, 91–291.

Felsenstein, J. (1985). Confidence limits on phylogenies: An approach usingthe bootstrap. Evolution 39, 783–791. doi:10.2307/2408678

Folmer, O., Black, M., Hoeh, W., Lutz, R., and Vrijenhoek, R. (1994). DNAprimers for amplification of mitochondrial cytochrome c oxidasesubunit I from diverse metazoan invertebrates. Molecular MarineBiology and Biotechnology 3, 294–299.

Gillespie, R. G. (2004). Community assembly through adaptive radiation inHawaiian spiders. Science 303, 356–359. doi:10.1126/science.1091875

Glover, P. E., Glover, E. C., Trump, E. C., and Wateridge, L. E. D. (1964).The lava caves of Mount Suswa, Kenya, with particular reference to theirstudies in ecological role. Studies in Speleology 1, 51–56.

Goloboff, P. A., Farris, J. S., and Nixon, K. C. (2003). TNT: Tree AnalysisUsing New Technologies. Available at http://www.zmuc.dk/public/phylogeny/TNT/ [Verified 30 November 2007]

Gould, S. J., and Vrba, E. S. (1982). Exaptation- a missing term in thescience of form. Palaeobiology 8, 4–15.

Grandcolas, P. (1997). ‘The Origin of Biodiversity in Insects: PhylogeneticTests of Evolutionary Scenarios.’ (Muséum national d’Histoirenaturelle: Paris, France.)

Guillou, H., Carracedo, J. C., Torrado, F. P., and Badiola, E. R. (1996). K-Arages and magnetic stratigraphy of a hotspot-induced, fast grown oceanicisland: El Hierro, Canary Islands. Journal of Volcanology andGeothermal Research 73, 141–155. doi:10.1016/0377-0273(96)00021-2

Hall, T. (1999). BioEdit: a user-friendly biological sequence alignmenteditor and analysis program for Windows 95/98NT. Nucleic AcidsSymposium Series 41, 95–98.

Hedin, M. C. (1997a). Molecular phylogenetics at the population/speciesinterface in cave spiders of the Southern Appalachians (Araneae:Nesticidae: Nesticus). Molecular Biology and Evolution 14, 309–324.

Hedin, M. C. (1997b). Speciational history in a Diverse clade of habitat-specialized spiders (Araneae: Nesticidae: Nesticus): Inferences fromgeographic-based sampling. Evolution 51, 1929–1945. doi:10.2307/2411014

Hoch, H., and Howarth, F. G. (1999). Multiple cave invasion by the speciesof the cixiid planthopper Oliarus in Hawaii. Zoological Journal of theLinnean Society 127, 453–475. doi:10.1006/zjls.1998.0185

Holsinger, J. R. (1988). Troglobites: the evolution of cave-dwelling organ-isms. American Scientist 76, 146–153.

Hopkin, S. P., and Martin, M. H. (1985). Assimilation of zinc, cadmium,lead, copper, and iron by the spider Dysdera crocata, a predator ofwoodlice. Bulletin of Environmental Contamination and Toxicology 34,183–187. doi:10.1007/BF01609722

Howarth, F. G. (1972). Cavernicoles in lava tubes on the island of Hawaii.Science 175, 325–326. doi:10.1126/science.175.4019.325

Howarth, F. G. (1973). The cavernicolous fauna of Hawaiian lava tubes, 1.Introduction. Pacific Insects 15, 139–151.

Howarth, F. G. (1981). Community structure and niche differentiation inHawaiian lava tubes. In ‘Island Ecosystems; Biological Organization inSelected Hawaiian Communities’. (Eds D. Mueller-Dombois,K. W. Bridges and H. L. Carson.) pp.318–336. (Academic Press: NewYork, USA.)

Howarth, F. G. (1983). Ecology of cave arthropods. Annual Review ofEntomology 28, 365–389. doi:10.1146/annurev.en.28.010183.002053

Howarth, F. G. (1986). The tropical cave environment and the evolution oftroglobites. In ‘Proceedings of the 9th International Congress ofSpeleology’. pp. 153–155. (Barcelona, Spain.)

Howarth, F. G. (1987). The evolution of non-relictual tropical troglobites.International Journal of Speleology 16, 1–16.

Howarth, F. G. (1991). Hawaiian cave faunas: Macroevolution on youngislands. In ‘The Unity of Evolutionary Biology’. (Ed. E. C. Dudley.)pp. 285–295. (Dioscorides Press: Portland, OR, USA.)

Howarth, F. G. (1993). High stress subterranean habitats and evolutionarychange in cave-inhabiting arthropods. American Naturalist 142,565–577.

Howarth, F. G., and Hoch, H. (2005). Adaptive shifts. In ‘Encyclopedia ofCaves’. (Eds D. C. Culver and W. B. White.) pp. 17–24. (AcademicPress: Boston, MA, USA.)



Izquierdo, I. (1997). ‘Estrategias adaptativas al medio subterraneo de lasespecies del genero Loboptera Brunner W. (Blattaria, Blattellidae) en lasIslas Canarias.’ (Universidad de La Laguna: La Tenerife, Spain.)

Jarrige, J. (1957). Coleopteres Brachelytra de l’ile de la Reunion. MemoiresInstitut Sciences Madagascar. Serie E 8, 103–118.

Juberthie, C. (1984). La colonisation du mileu souterrain; théories etméthodes, relations avec la spéciation et l’evolution souterraine.Memoires de Biospeologie 11, 65–102.

Kane, T., and Culver, D. C. (1992). Biological processes in space and time:analysis of adaptation. In ‘The Natural History of Biospeleology’. (Ed.A. I. Camacho.) pp. 377–400. (Museo Nacional de Ciencias Naturales(CSIC): Madrid, Spain.)

Kumar, S., Tamura, K., and Nei, M. (2004). MEGA3: Integrated softwarefor molecular evolutionary genetics analysis and sequence alignment.Briefings in Bioinformatics 5, 150–163. doi:10.1093/bib/5.2.150

Kuntner, M., Sket, B., and Blejec, A. (1999). A comparison of the respira-tory systems in some cave and surface species of spiders (Araneae,Dysderidae). The Journal of Arachnology 27, 142–148.

Legendre, P., and Legendre, L. (1998). ‘Numerical Ecology.’ (Elsevier B. V.:Amsterdam, The Netherlands.)

Leleup, N., and Leleup, J. (1968). ‘Mission zoologique belge aux ãilesGalapagos et en Ecuador.’ (Musâee Royal de l’Afrique Centrale; Institutroyal des sciences naturelles de Belgique: Tervuren, Bruxelles,Belgium.)

Leys, R., Watts, C. H., Cooper, S. J., and Humphreys, W. F. (2003).Evolution of subterranean diving beetles (Coleoptera: Dytiscidae:Hydroporini, Bidessini) in the arid zone of Australia. Evolution 57,2819–2834.

Losos, J. B., Jackman, T. R., Larson, A., De Queiroz, K., and RodriguezSchettino, L. (1998). Contingency and determinism in replicated adap-tive radiations of island lizards. Science 279, 2115–2118.doi:10.1126/science.279.5359.2115

Martín, J. L. (1991). ‘Fauna invertebrada del Parque Nacional de Timanfaya(Lanzarote, Islas Canarias).’ (Servicio de publicaciones de la CajaGeneral de Ahorros de Canarias: Santa Cruz de Tenerife, Spain.)

Martín, J. L. (1992). ‘Caracterización ecológica y evolución de las comu-nidades subterráneas en las islas de Tenerife, el Hierro y La Palma.’(Universidad de La Laguna: La Laguna, Spain.)

Martín, J. L., and Oromí, P. (1986). An ecological study of Cueva de losRoques lava tube (Tenerife, Canary Islands). Journal of Natural History20, 375–388. doi:10.1080/00222938600770281

Medina, A. L. (1991). El medio subterráeneo superficial en las IslasCanarias: Caracterización y consideraciones sobre su fauna. Ph.D.Thesis, Universidad de La Laguna, Tenerife, Spain.

Miller, J. A. (2005). Cave adaptation in the spider genus Anthrobia(Araneae, Linyphiidae, Erigoninae). Zoologica Scripta 34, 565–592.doi:10.1111/j.1463-6409.2005.00206.x

Miller, J. S., and Wenzel, J. W. (1995). Ecological characters and phylogeny.Annual Review of Entomology 40, 389–415. doi:10.1146/annurev.en.40.010195.002133

Moulds, T. A., Murphy, N., Adams, M., Reardon, T., Harvey, M. S.,Jennings, J., and Austin, A. D. (2007). Phylogeography of cave pseudo-scorpions in southern Australia. Journal of Biogeography 34, 951–962.

Nixon, K. C. (2002). WinClada. (Published by the Author: New York, USA.)Available at www.cladistics.com [Verified 30 November 2007]

Oromí, P. (2004). Canary Islands: Biospeleology. In ‘Encyclopedia of cavesand karst science’. (Ed. J. Gunn.) pp. 179–181. (Fitzroy Dearborn: NewYork, USA.)

Oromí, P., and Izquierdo, I. (1994). The Canary Islands: Undergroundenvironment and adapted fauna. In ‘Encyclopaedia Biospeleologica’.(Eds C. Juberthie and V. Decu.) pp. 631–639. (Société Internationale deBiospéologie: Moulis, France.)

Oromí, P., Martín, J. L., Medina, A. L., and Izquierdo, I. (1991). The evolu-tion of the hypogean fauna in the Canary Islands. In ‘The Unity of

Evolutionary Biology’. (Ed. E. C. Dudley.) pp. 380–395. (DioscoridesPress: Portland, OR, USA.)

Peck, S. B. (1975). The invertebrate fauna of tropical American caves, PartIII: Jamaica, an introduction. International Journal of Speleology 7,303–326.

Peck, S. B. (1990). Eyeless arthropods of the Galapagos Islands, Ecuador:Composition and origin of the cryptozoic fauna of a young, tropical,oceanic archipelago. Biotropica 22, 366–381. doi:10.2307/2388554

Peck, S. B., and Finston, T. L. (1993). Galapagos islands, troglobites: thequestions of tropical toglobites, parapatric distribution with eyed-sister-species, and the origin of parapatric speciation. Memoires deBiospeologie 20, 19–37.

Petit-Maire, N., Delibrias, G., Meco, J., Pomel, S., and Rosso, J. C. (1986).Paleoclimatologie des Canarie Orientales (Fuerteventura). ComptesRendues Academie Sciences Paris, t. 303, Serie II 13, 1241–1246.

Platnick, N. I. (2006). The world spider catalog. (American Museum ofNatural History: New York.) Available at http://research.amnh.org/entomology/spiders/catalog81-87/INTRO1.html [Verified 30 Novem-ber 2007]

Pollard, S. D. (1986). Prey capture in Dysdera crocata (Araneae:Dysderidae), a long fanged spider. New Zealand Journal of Zoology 13,149–150.

Pollard, S. D., Jackson, R. R., Van Olphen, A., and Robertson, M. W. (1995).Does Dysdera crocata (Araneae Dysderidae) prefer woodlice as prey?Ethology Ecology and Evolution 7, 271–275.

Pons, J., and Vogler, A. (2006). Size, frequency, and phylogenetic signal ofmultiple-residue indels in sequence alignment introns. Cladistics 22,144–156. doi:10.1111/j.1096-0031.2006.00088.x

Posada, D., and Buckley, T. R. (2004). Model selection and model averag-ing in phylogenetics: advantages of Akaike Information Criterion andBayesian approaches over likelihood ratio tests. Systematic Biology 53,793–808. doi:10.1080/10635150490522304

Posada, D., and Crandall, K. A. (1998). MODELTEST: testing the model ofDNA substitution. Bioinformatics 14, 817–818.

Poulson, T. L. (1981). Variations in life history of linyphiid cave spiders. In‘Eighth International Congress of Speleology, Bowling Green,Kentucky’. (Ed. B. F. Beck.) pp. 60–62. (National Speleological Society:Bowling Green, KY, USA.)

Racovitza, E. G. (1907). Essai sur les problemes biospeologiques. ArchivesZoologie Experimentale et Generale (Biospeologica I) 43(6), 371–488.

Rambaut, A., and Drummond, A. J. (2003). TRACER. Available athttp://tree.bio.ed.ac.uk/software/tracer [Verified 30 November 2007]

Reddell, J. R. (1977). preliminary survey of the caves of the YucatanPeninsula. In ‘Studies of the Caves and Cave Fauna of the YucatanPeninsula’. (Ed. J. R. Reddell.) pp. 215–296. (The Speleo. Press: Austin,TX, USA.)

Ribera, C., and Juberthie, C. (1994). Araneae. In ‘EncyclopaediaBiospeologica.’ (Ed. C. Juberthie.) pp. 197–214. (Société Internationalede Biospéologie: Moulis, France.)

Rivera, M. A., Howarth, F. G., Taiti, S., and Roderick, G. K. (2002).Evolution in Hawaiian cave-adapted isopods (Oniscidea: Philosciidae):vicariant speciation or adaptive shifts? Molecular Phylogenetics andEvolution 25, 1–9. doi:10.1016/S1055-7903(02)00353-6

Rognon, P., Coude-Gaussen, G., Le Costumer, M. N., Balouet, J. C., andOchietti, S. (1989). Le massif dunaire de Jandia (Fuerteventura,Canaries): evolution des paleoenvironments de 20.000 BP a l’actuel.Bulletin de l’Association francaise pour l’Etude du Quaternaire 1,31–37.

Ronquist, F., and Huelsenbeck, J. P. (2003). MrBayes 3: Bayesian phylo-genetic inference under mixed models. Bioinformatics 19, 1572–1574.

Rovner, J. S. (1986). Nests of terrestrial spiders maintain a physical gillflooding and the evolution of silk constructions. The Journal ofArachnology 14, 327–338.


Sanger, F., Nicklen, S., and Coulsen, A. R. (1977). DNA sequencing withchain terminating inhibitors. Proceedings of the National Academy ofSciences of the United States of America 74, 5463–5468.

Sarthein, M., and Koopman, B. (1980). Late Quaternary deep-sea record onnorthwest African dust supply and wind circulation. In ‘Palaeoecologyof Africa and the Surrounding Islands’. (Eds E. M. van Zinderen Bakker,Sr. and J. A. Coetzee.) pp. 239–253. (A. A. Balkema: Rotterdam,Germany.)

Sarthein, M., Tetzlaff, G., Koopmann, B., Wolter, K., and Pflaumann, U.(1981). Glacial and interglacial wind regimes over the eastern subtropi-cal Atlantic and North-west Africa. Nature 293, 193–196.

Sbordoni, V. (1982). Advances in speciation of cave animals. In‘Mechanism of Speciation’. (Ed. C. Bigozzi.) pp. 219–240. (A.R. Liss:New York, USA.)

Sbordoni, V., Allegrucci, G., and Cesaroni, D. (2000). Population geneticstructure, speciation and evolutionary rates in cave-dwelling organisms.In ‘Ecosystems of the World’. (Eds H. Wilkens, D. C. Culver and W. F.Humphreys.) pp. 453–477. (Elsevier: Amsterdam, The Netherlands.)

Schilthuizen, M., Cabanban, A. S., and Haase, M. (2005). Possible specia-tion with gene flow in tropical cave snails. Journal of ZoologicalSystematics and Evolutionary Research 43, 133–138. doi:10.1111/j.1439-0469.2004.00289.x

Schiner, J. R. (1854). Fauna der Adelsberger-, Lueger- und Magdalenen-Grotte. In ‘Die Grotten und Höhlen von Adelsberg, Lueg, Planina undLaas’. (Ed. A. Schmidl.) pp. 231–172. (Braunmülkler: Wien, Germany.)

Schluter, D. (2000). ‘The Ecology of Adaptive Radiation.’ (OxfordUniversity Press: Oxford, UK.)

Simmons, M. P., and Ochoterena, H. (2000). Gaps as characters insequence-based phylogenetic analyses. Systematic Biology 49,369–381. doi:10.1080/10635159950173889

Simmons, M. P., Ochoterena, H., and Carr, T. G. (2001). Incorporation, rel-ative homoplasy, and effect of gap characters in sequence-based phylo-genetic analyses. Systematic Biology 50, 454–462. doi:10.1080/106351501300318049

Simon, C., Frati, F., Beckenbach, A., Crespi, B., Liu, H., and Flook, P.(1994). Evolution, weighting, and phylogenetic utility of mitochondrialgene sequences and a compilation of conserved polymerase chainreaction primers. Annals of the Entomological Society of America 87,651–701.

Sorenson, M. D. (1999). ‘TreeRot ver. 2.’ (Boston University: Boston, MA,USA.) Available at http://people.bu.edu/msoren/TreeRot.html [Verified5 December 2007]

Strecker, U., Bernatchez, L., and Wilkens, H. (2003). Genetic divergencebetween cave and surface populations of Astyanax in Mexico(Characidae, Teleostei). Molecular Ecology 12, 699–710. doi:10.1046/j.1365-294X.2003.01753.x

Swofford, D. (2001). ‘PAUP Phylogenetic Analysis Using Parsimony (andOther Methods), Version 4.’ (Sinauer Associated: Sunderland, MA,USA.)

Thompson, J. D., Gibson, T. J., Plewniak, F., Jeanmougin, F., and Higgins,D. G. (1997). The CLUSTAL_X windows interface: flexible strategiesfor multiple sequence alignment aided by quality analysis tools. NucleicAcids Research 25, 4876–4882. doi:10.1093/nar/25.24.4876

Trajano, E. (1995). Evolution of tropical troglobites: applicability of themodel of quaternary climatic fluctuations. Memoires de Biospeologie22, 203–210.

Wheeler, W. C., and Hayashi, C. Y. (1998). The phylogeny of the extant che-licerate orders. Cladistics 14, 173–192. doi:10.1111/j.1096-0031.1998.tb00331.x

Wiens, J. J., Chippindale, P. T., and Hillis, D. M. (2003). When are phylo-genetic analyses misled by convergence? A case study in Texas cavesalamanders. Systematic Biology 52, 501–514.

Wilkens, H., and Hüppop, K. (1986). Sympatric speciation in cave fishes?Studies on a mixed population of epi- and hypogean Astyanax(Characidae, Pisces). Zeitschrift fuer Zoologische Systematik undEvolutionsforschung 24, 223–230.

Wunderlich, J. (1991). Die Spinnen-fauna der Makaronesischen Inseln.Beiträge zur Araneologie 1, 1–619.

Young, N. D., and Healy, J. (2003). GapCoder automates the use of indelcharacters in phylogenetic analysis. BMC Bioinformatics 4, 6.

Manuscript received 24 April 2007, accepted 22 October 2007



Appendix 1. Species included in the cladistic analysis and GenBank accession numbers for the cox1 and rrnL gene sequencesAll accession numbers starting with EU are new sequences obtained in the present study

Species Locality cox1 rrnL

Harpactea hombergi Iberian Peninsula, Montseny AF244233 AF244148Stalita stygia Slovenia, Mrzla Jama AF244234 AF244149Harpactocrates radulifer Iberian Peninsula, La Rioja, Vargañon AF244235 AF244150Dysderocrates silvestris Croatia, Gracac, 5 km ENE Zadar AF244236 AF244151CONTINENTALDysdera adriatica Slovenia, Kozina EU068026 EU068064Dysdera atlantica Morocco, Essaouira, S from the city EU068029 EU068057Dysdera crocata Canary Is, Tenerife, Buenavista, Teno Alto AF244237 AF244152Dysdera edumifera Iberian Peninsula, Madrid, Peralejo de Fresnedillas EU068032 EU068069Dysdera erythrina Iberian Peninsula, Montseny AF244252 AF244162Dysdera fuscipes Iberian Peninsula, Huelva, Sierra Aracena. Fuenteheridos EU068039 EU068073Dysdera lucidipes melillensis Morocco, Ras el Ma EU068042 EU068060Dysdera mauritanica Morocco, Essaouira EF458138 EF458093Dysdera mucronata Morocco, Tangier EU068044 EU068077Dysdera ninni Slovenia EU068045 EU068062Dysdera scabricula Iberian Peninsula, Castelló, Desert de les Palmes EU068046 EU068078Dysdera cf. seclusa Morocco, Barrage Nakhla, rd. Tetouan to Xauen EU068035 EU068071Dysdera sp. MA Morocco, High Atlas, Tazouguerte AF244244 AF244155Dysdera sp. MB Morocco, High Atlas, Tazouguerte AF244245 AF244156Dysdera sp. MC Morocco, High Atlas, Tazouguerte AF244246 AF244157Dysdera sp. MD Morocco, High Atlas, Tizi-n-Ait Ouirra AF244247 AF244158Dysdera sp. MF Morocco, Middle Atlas, Azrou AF244249 AF244159Dysdera sp. MH Morocco, Draa River’s mouth AF244250 AF244160Dysdera coiffaiti Madeira, Levada do Cedro AF244251 AF244161CANARIANDysdera alegranzaensis Canary Is, Alegranza AF244257 AF244167Dysdera ambulotenta Canary Is, Tenerife, La Orotava, Cueva del Bucio AF244276 AF244181Dysdera ambulotenta Canary Is, Tenerife, Icod de los Vinos, Cueva de Felipe Reventón EU068055 EU068065Dysdera andamanae Canary Is, Gran Canaria, Santa María de Guía, Brezal del Palmital EU068027 EU068056Dysdera andamanae Canary Is, Gran Canaria, Santa María de Guía, Brezal del Palmital EU068028 EU068066Dysdera arabisenen Canary Is, Gran Canaria, Vega de Santa Mateo, La Avejerilla AF244291 AF244198Dysdera bandamae Canary Is, Gran Canaria, Agüimes, Barranco de Guayadeque AF244286 AF244193Dysdera brevisetae Canary Is, Tenerife, Santa Cruz, El Bailadero AF244299 AF244207Dysdera brevispina Canary Is, Tenerife, Santa Cruz, El Bailadero AF244316 AF244227Dysdera calderensis Canary Is, La Gomera, Hermigua, Barranco de Juel AF244308 AF244217Dysdera calderensis Canary Is, La Palma, Garafía, Juan Adalid AF244310 AF244219Dysdera chioensis Canary Is, Tenerife, Guía de Isora, Cueva Grande de Chío AF244281 AF244188Dysdera chioensis Canary Is, Tenerife, Güímar, Cueva Honda de Güímar EU068030 EU068067Dysdera chioensis Canary Is, Tenerife, La Orotava, Cueva de los Roques AF244282Dysdera cribellata Canary Is, Tenerife, Santa Cruz, Cruz del Carmen AF244224Dysdera curvisetae Canary Is, Tenerife, Buenavista, Playa del Barranco de Natero EU068031 EU068068Dysdera enghoffi Canary Is, La Gomera, Hermigua, Barranco Aramaqué AF244271 AF244175Dysdera esquiveli Canary Is, Tenerife, Icod de los Vinos, Cueva de Felipe Reventón AF244298 AF244206Dysdera gibbifera Canary Is, Tenerife, Icod de los Vinos, Cueva de Felipe Reventón AF244277 AF244182Dysdera gibbifera Canary Is, Tenerife, Buenavista, Cabecera Barranco de Cochinos, Teno, MSS EU068033Dysdera gibbifera Canary Is, Tenerife, Buenavista, Monte del Agua, MSS EU068034 EU068070Dysdera gollumi Canary Is, Tenerife, La Orotava, Cueva de los Roques AF244297 AF244204Dysdera gollumi Canary Is, Tenerife, La Orotava, Cueva de los Roques EU068036 EU068058Dysdera guayota Canary Is, Tenerife, Adeje, Roque del Conde AF244283 AF244190Dysdera gomerensis Canary Is, El Hierro, Frontera, Pista del Derrabado AF244318 AF244229Dysdera gomerensis Canary Is, La Gomera, Hermigua, Pajarito AF244317 AF244228Dysdera hernandezi Canary Is, Tenerife, Tacoronte, Cueva Labrada EU068037 EU068059Dysdera iguanensis Canary Is, Tenerife, Santa Cruz, Cabezo del Tejo AF244279 AF244186Dysdera iguanensis Canary Is, Gran Canaria, Mogán, Inagua AF244280 AF244187Dysdera insulana Canary Is, Tenerife, Santa Cruz, El Bailadero AF244314 AF244225Dysdera labradaensis Canary Is, Tenerife, Icod de los Vinos, Cueva de Felipe Reventón EU068040 EU068074Dysdera labradaensis Canary Is, Tenerife, Icod de los Vinos, Cueva del Viento-Sobrado AF244275Dysdera lancerotensis Canary Is, Fuerteventura, Pájara, Barranco del Ciervo AF244242 AF244154Dysdera levipes Canary Is, La Gomera, Vallehermoso, Ermita de Santa Clara AF244296 AF244203

(continued next page)


Appendix 1. (continued)

Species Locality cox1 rrnL

Dysdera levipes Canary Is, Tenerife, Los Realejos, Palo Blanco AF244295 AF244202Dysdera levipes Canary Is, Tenerife, Santa Cruz, Cruz del Carmen, Cabeza Zapata EU068041 EU068075Dysdera liostethus Canary Is, Gran Canaria, Gáldar, Pico del Viento AF244303 AF244212Dysdera liostethus Canary Is, Gran Canaria, Santa Nicolás de Tolentino, Degollada de Tasartico AF244302 AF244211Dysdera longa Canary Is, Fuerteventura, Pájara, Barranco del Ciervo AF244255 AF244165Dysdera macra Canary Is, Tenerife, La Orotava, Izaña, AF244300 AF244209Dysdera madai, sp. nov. Canary Is, Tenerife, Buenavista, Barranco de Cochinos, MSS (K290) EU068061Dysdera madai, sp. nov. Canary Is, Tenerife, Buenavista, Barranco de Cochinos, MSS (K421) EU068043 EU068076Dysdera montanetensis Canary Is, Tenerife, El Rosario, Las Raíces AF244278 AF244183Dysdera nesiotes Canary Is, Montaña Clara AF244263 AF244170Dysdera orahan Canary Is, El Hierro, Frontera, Mirador de Bascos AF244313 AF244222Dysdera paucispinosa Canary Is, Gran Canaria, Mogán, Degollada de las Brujas AF244306 AF244215Dysdera ramblae Canary Is, La Gomera, Hermigua, Barranco de Juel AF244311 AF244220Dysdera ratonensis Canary Is, La Palma, Mazo, Cueva Tacande AF244305 AF244214Dysdera ratonensis Canary Is, La Palma, Fuencaliente, Cueva Palmeros AF244304Dysdera ratonensis Canary Is, La Palma, Fuencaliente, Cueva Palmeros EU068049Dysdera ratonensis Canary Is, La Palma, Barlovento, Cueva Honda de Gallegos EU068048Dysdera ratonensis Canary Is, La Palma, Barlovento, Cueva Honda de Gallegos EU068050Dysdera ratonensis Canary Is, La Palma, Barlovento, Cueva Honda de Gallegos EU068051Dysdera ratonensis Canary Is, La Palma, Barlovento, Cueva Honda de Gallegos EU068052Dysdera ratonensis Canary Is, La Palma, Barlovento, Cueva Honda de Gallegos EU068053Dysdera ratonensis Canary Is, La Palma, Barlovento, Cueva Honda de Gallegos EU068054Dysdera rugichelis Canary Is, Gran Canaria, Agaete, Pinar de Tamadaba AF244293 AF244200Dysdera sibyllina, sp. nov. Canary Is, Tenerife, Icod de los Vinos, Galería de Ingleses EU068063Dysdera sibyllina, sp. nov. Canary Is, Tenerife, Icod de los Vinos, Cueva de Felipe Reventón EU068047 EU068079Dysdera silvatica Canary Is, La Gomera, Hermigua, Barranco de Juel AF244273 AF244177Dysdera silvatica Canary Is, La Palma, Garafía, Pista de Machín AF244274 AF244178Dysdera silvatica Canary Is, El Hierro, Frontera, Cueva de Longueras EU068038 EU068072Dysdera sanborondon Canary Is, Fuerteventura, Tuineje, Cuchillos de Jacomar AF244256 AF244166Dysdera spinidorsum Canary Is, Fuerteventura, Puerto del Rosario, La Matilla AF244269 AF244173Dysdera tilosensis Canary Is, Gran Canaria, Santa María de Guía, Brezal del Palmital AF244288 AF244195Dysdera unguimmanis Canary Is, Tenerife, Icod de los Vinos, Cueva del Viento-Sobrado AF244285 AF244192Dysdera unguimmanis Canary Is, Tenerife, Icod de los Vinos, Cueva de Felipe Reventón AF244284Dysdera verneaui Canary Is, Tenerife, Los Silos, Monte del Agua AF244321 AF244232Dysdera volcania Canary Is, Tenerife, Icod de los Vinos, Cueva de Felipe Reventón AF244184Dysdera yguanirae Canary Is, Gran Canaria, Santa María de Guía, Brezal del Palmital AF244290 AF244197


http://www.publish.csiro.au/journals/is

Appendix 2. Clade support values for parsimony and Bayesian inference analysesClade numbers as in Fig. 2. Only parsimony bootstrap support values above 50% are reported. Total (BS) and partition (morphology, cox1, rrnL and gaps)

Bremer supports for each gene and genome partition. ns, contradicted (clade did not appear in that analysis)

PBSBP PP morpho cox1 rrnL gaps BS

PBSBP PP morpho cox1 rrnL gaps BS

1 90 1.00 12.00 –4.50 1.50 2.00 112 92 0.99 9.50 –3.00 3.50 1.00 113 – 0.79 8.00 –13.50 8.50 0.00 34 85 1.00 –7.50 –8.50 13.00 12.00 95 – ns –1.50 4.00 –1.50 0.00 16 84 1.00 9.00 3.00 –6.00 5.00 117 – 0.74 2.50 –4.50 2.00 1.00 18 – ns 1.67 0.83 –0.17 –0.33 29 85 0.99 3.83 11.67 –8.17 –1.33 610 56 0.68 0.50 –3.00 3.50 0.00 111 58 0.94 –0.50 –1.00 0.50 3.00 212 87 1.00 1.00 8.17 4.50 –0.67 1313 84 1.00 3.50 –1.50 4.50 1.50 814 – ns 3.50 3.50 –5.00 –1.00 115 86 1.00 7.50 –4.00 5.50 0.00 916 74 1.00 5.25 –4.50 4.75 1.50 717 52 1.00 5.00 1.00 –2.00 –2.00 218 – 0.83 –1.50 4.00 –1.50 0.00 119 – 0.98 0.50 0.00 0.50 0.00 120 – 0.57 7.00 –2.00 –2.00 1.00 421 88 1.00 2.50 –0.67 1.17 0.00 322 – ns 5.50 –2.33 –1.50 2.33 423 69 1.00 –2.36 1.86 8.64 1.86 1024 – 0.95 10.50 –7.00 –5.50 4.00 225 – 1.00 6.00 –4.00 0.50 0.50 326 – 0.56 10.00 –2.00 –6.50 –0.50 127 99 1.00 5.36 7.71 –6.36 2.29 928 – 0.96 4.83 1.67 –3.83 –0.67 229 – ns –2.00 4.00 0.00 0.00 230 97 1.00 –2.00 4.00 0.00 0.00 231 – ns 4.00 6.50 –7.50 4.00 732 – ns 9.50 5.00 –11.50 –2.00 133 84 1.00 6.33 –1.67 0.67 –0.33 5

34 100 1.00 7.72 6.41 1.21 4.67 2035 55 0.74 6.33 –2.00 –3.22 1.89 336 – 0.95 6.58 –8.17 0.92 1.67 137 – ns 1.23 –6.27 5.17 2.87 338 99 1.00 7.94 –9.00 7.56 7.50 1439 – ns 6.58 –8.17 0.92 1.67 140 – ns 6.58 –8.17 0.92 1.67 141 – ns 2.22 –8.94 3.72 4.00 142 – ns 6.83 –10.17 –2.33 7.67 243 – ns 6.58 –8.17 0.92 1.67 144 96 1.00 10.30 –7.40 0.10 2.00 545 – 0.99 9.28 –0.63 –7.85 3.19 446 – 1.00 6.00 2.50 –11.50 6.00 347 73 0.99 13.50 4.00 –8.50 –2.00 748 99 1.00 –0.06 6.56 6.31 5.19 1849 61 0.47 3.50 4.00 –14.50 8.00 150 – 1.00 3.57 –3.00 0.63 2.80 451 56 0.99 0.00 2.00 1.00 0.00 352 – ns –1.67 –7.67 6.67 3.67 153 – ns –3.37 –6.52 3.92 6.97 154 76 1.00 –2.17 –3.67 –1.83 10.67 355 98 1.00 1.33 –9.87 8.87 11.67 1256 64 0.75 –3.17 –7.33 7.83 3.67 157 97 1.00 3.53 1.73 0.87 1.87 858 – ns 3.03 –7.05 –0.95 5.97 159 – 0.70 –3.17 –3.17 –2.33 9.67 160 62 0.99 10.82 1.25 –9.17 1.10 461 57 0.99 10.54 –0.52 –7.62 1.61 462 100 1.00 8.27 9.93 10.23 2.57 3163 71 1.00 0.67 –5.33 2.67 4.00 264 – 0.87 –0.37 –4.87 4.17 2.07 165 54 0.93 2.22 –5.67 0.96 5.49 3

258.62 –104.10 –0.07 165.55 320

The dark side of an island radiation: systematics and evolution of troglobitic spiders of the genus...

Documents

Transcript of The dark side of an island radiation: systematics and evolution of troglobitic spiders of the genus...