2013. The Journal ofArachnology 41:25-30 Does behavioral isolation prevent interspecific mating within a parallel ecotypic wolfspider radiation from the Galapagos? C. De Busschere' and F. Hendrickx'-^: 'Terrestrial Ecology Unit, Department of Biology, Ghent University, K. L. Ledeganckstraat 35, B-9000 Ghent, Belgium. E-mail: [email protected]; "Department ofEntomology, Royal Belgian Institute of Natural Sciences, Rue Vautier 29, 1000 Brussels, Belgium Abstract. Behavioral isolation may play an important role in speciation. However, the roles of divergence time and ecological specialization on the evolution of intrinsic barriers to gene flow are poorly understood. On the Galapagos, ecotypic differentiation ofHogna Simon 1885 wolfspiders has led to the repeated evolution ofmorphologically distinct high-elevation and coastal species on Santa Cruz and San Cristobal. This offers a unique opportunity to investigate the importance ofecological factors and evolutionary history on courtship behavior, but also to explore the opportunity for interspecific gene flow. On San Cristobal, both high elevation and coastal Hogna species clearly showed distinct courtship behavior.Thispatterncorresponded primarilywith variationinmalegenital organsratherthan withecotypicclassification or phylogenetic relationship. Despite low acceptance rates, heterospecific mating was observed, suggesting that potential gene flow within as well as among islands should not be neglected when seeking to understand island radiations. Keywords: Gene flow, parallel evolution, natural selection, sexual selection The speciation process necessarily involves the reduction behavioral isolating mechanisms (Andersson 1994; Uetz of gene flow between actually or potentially interbreeding 2000; Rypstra et al. 2009). Moreover, male wolf spider populations (Coyne & Orr 2004; Eutuyma 2005). If popula- courtship, which may involve different sensory channels such tions diverge in allopatry, spatial isolation serves as an initial as visual, vibratory and chemical signals, often leads to isolating barrier (Coyne & Orr 2004). This initial barriermight elaborate species-specific male courtship displays (Miller et al. be re-enforced due to the accumulation of differentially 1998; Hebets & Uetz 2000), enabling delineation of species selected traits thatreduce interspecific attraction and therefore boundaries between morphologically indistinguishable species heterospecific mating; i.e., behavioral/sexual isolation (An- (Den Hollander & Dijkstra 1974; Uetz & Denterlein 1979; dersson 1994; Schluter 2000; Panhuis et al. 2001; Masta & Topfer-Hofmann et al. 2000; Chiarle et al. 2010). Maddison 2002; Rundle & Nosil 2005). In general, mating Within-island habitat specialization was demonstrated to traits are predicted to diverge between populations due to lead to morphologically highly similar Hogna species in similar mechanisms that are not related to the environment, such as habitats on both San Cristobal and Santa Cruz (De Busschere genetic drift and sexual selection (Eutuyma 2005). In contrast, et al. 2010, 2012) (Eig. 1). Hogna galapagoensis Banks 1902 in the light of ecological speciation, the evolution of mating and H. junco Baert & Maelfait 2008 are morphologically traits is predicted to correlate with the environment as a difficult to distinguish (with theexception ofgenital traits) and byproduct ofnatural selection (Boughman 2001; McKinnon & are referred to as “high elevation species” occurring on the Rundle 2002; Rundle & Nosil 2005) and may as such lead to top of Santa Cruz and San Cristobal in the dense pampa assortative mating ofpopulations that have undergone similar vegetation dominated by ferns and sedges (Fig. 1). Similarly, selection pressures. Hence, scenarios wherein allopatric H. hendrickxi Baert & Maelfait 2008 and H. snodgrassi Banks populations each diverged along a similar selection gradient 1902 are morphologically difficult to distinguish and are (i.e., parallel divergence) provide a unique opportunity to test referred to as coastal dry species found in the dry supralittoral the respective roles of sexual and natural selection in the and arid zone along the coast in sparsely vegetated dunes and evolution ofbehavioral isolation (Boughman 2001; Boughman open shrub land on Santa Cruz and San Cristobal (Baert et al. et al. 2005). Although mating traits most frequently evolve 2008c) (Fig. 1). High-elevation species are characterized by independently among species when populations are isolated, darker coloration, smaller body size and smaller eyes than they can also be expected to evolve by species interactions coastal dry species (De Busschere et al. 2012). Contemporary when the diverging populations come into secondary contact gene flow between these species appears absent, based on and suffer reduced hybrid viability (Dobzhansky 1937; Coyne allozyme allele frequencies (Baert et al. 2008a) and on spatial & Orr 1989). isolation of high elevation and coastal dry habitats (Fig. 1). As behavioral isolation is expected to evolve rapidly Nevertheless, on Santa Cruz, ecological divergence between H. between incipient species (Del Solar 1966; Gleason & Ritchie galapagoensis and H. hendrickxi has been shown to occur in 1998) and even faster than intrinsic postzygotic isolation the face oflow levels ofgene flow (De Busschere et al. 2010), barriers (Coyne & Orr 1989, 1997, 2004; Mendelson 2003), we which is in agreement with the very similar shape oftheir male here focus on the role ofbehavioral isolation in the radiation copulatory organs (De Busschere et al. 2012). Moreover, of the ground-dwelling wolf spider genus Hogna Simon 1885 based upon Loosveldt (2004) and our sampling campaigns, from the Galapagos (De Busschere et al. 2010, 2012). As male these Hogna species seem to have a similar seasonal life cycle, wolfspiders need to persuade females by courting, differences in which activity is concentrated in the warm wet season from in courting signals are expected to serve as prezygotic January to May, suggesting no potential role for strong 25 26 THE JOURNAL OF ARACHNOLOGY B. H. hendrickxi jLD] H. hendrickxi pn 1.00 H. galapagoensis H. galapagoensis ^ H. snodgrassi H.junco H.junco H.junco . H. espanola 0.1substitutions/site — Figure 1. (A) Geographical distribution (adapted from De Busschere et al. 2010) and (B) COI-28S phylogeny (De Busschere et al., 2010) of high elevation (green) and coastal (red) Hogiui species on Santa Cruz and San Cristobal (with exception ofH. espcmohi). Node values represent Bayesian posteriorprobabilities; samplinglocalitiesareabbreviated as: Las Palmas(Ip), Media lima (ml), Punta Bassa (pb) and Volcan ElJunco (ej), and estimated minimum and maximum geological ages (MYA) for each island are in parentheses (D. Geist et al. unpubl . data). temporal isolation. The latter observation and the signature of reliable quantitative comparisons can be made concerning the historical gene flow demonstrate that gene flow between H. degree ofinter- and interspecific acceptance rates (see below). — galapagoensis and H. hendrickximight still be possible, despite Analysis of species-specific male courtship behavior. Wild- their ancient initial split (—0.8 My) (De Busschere et al. 2010). caught individuals were used in mating trials to describe the This study system allows us to investigate whether strong species-specific courtship behavior ofthe four species. The use ecological divergence and/or historical gene flow has led to of wild-caught individuals might confound results due to behavioral isolation by addressing the following research mating experience, age and mating status of the female. questions: 1) Are there interspecific differences in male court- Therefore, we restricted our aim to describing the presence or ship behavior, and are thesedifferences related to differences in absence of species-specific male courtship behaviors. Mating ecology,ordotheycorrelatewiththephylogeneticrelationships trials were performed in a plastic arena (30 X 20cm) filledwith among the species? 2) Do interspecific differences serve as 1 cm ofsand. Beforeeach trial, a newfilterpaperwas placed on effective prezygotic isolating barriers? To address these ques- the sand in order to eliminate signals from previous trials and tions, we documented male courtship behavior and performed allowing vibrations. Females were placed in the arena and interspecific mating trials among the four Hogna species for confronted with 1) a conspecific male, 2) a heterospecific whichwithin-island habitat specializationwasdemonstrated on ecotypically similar male (from the other island) or 3) a San Cristobal and on Santa Cruz. heterospecific ecotypically dissimilar male (from the same METHODS island) after a 5 min period of acclimatization. Trials were observed on average for 20 min. This time frame was chosen — Sampling. Juvenile and adult H. galapagoensis and H. based on an initial subset oftrials forwhich weobserved that if hendrickxi from Santa Cruz were sampled at Media Luna matingweretooccur, itwasgenerallycompletedwithinthefirst (pampa, high elevation) and Las Palmas (coastal), respective- 5 min oftheexperiment. Courtshipeventswere recorded with a ly. The species from San Cristobal, H.junco and H. snodgrassi, HDV camera (SONY HV40 Legria). Foreach matingtrial, the were sampled at Volcan El Junco (pampa, high elevation) and longest recorded complete courtship fragment within those Punta Bassa (coastal) in February 2010, respectively (Fig. 1). 20 min was used for further analyses. These fragments were Sampling efforts led to a total of 431 specimens with sample chosen if we observed several stages of the mating process sizes ranging from 64 to 187 individuals per species, 52% of starting from the male detecting the female, approaching her which were juveniles. They were housed individually in the and then elaborately courting her until she responded. quarantine laboratory ofthe Charles Darwin research station A list of five recognizable male courting behaviors was at an average temperature of20°C and fed ad libitum with two defined and used to score male courtship behavior during the to three wild-caught moths per day (adults) or five fruit flies fragment (Table 1; Video 1, online at http://www.bioone.org/ per day (juveniles). Although manyjuveniles exuviated in the doi/suppl/10.1636/K12-49). Females’ reactions were classified lab, none of them reached adulthood, suggesting that our as 1) acceptance of the male (i.e. allowing him to mount), 2) laboratory conditions did not adequately mimic the field aggressive behavior, or 3) no response. For each mating trial, conditions to induce maturation. Since we could not ensure total male courting time (ttot) was assessed, and the absence/ virginity of the females used in the mating experiments, no presence ofmale courting behavior was obtained. DE BUSSCHERE & HENDRICKX—BEHAVIORAL ISOLATION OF WOLF SPIDERS 27 — Table 1. Description ofmale courtship behaviors (see Video 1). PM Pedipalp movements: this involves all movements ofthe male pedipalps and mainly consisted ofdrumming against the substrate. FM Forelegmovements: thisinvolvesallmovementswith thefirsttwopairsoflegsandconsistsofrepeatedly raising,waving, tapping, arching and stretching forelegs. mP Moderate push-ups: this involves a period ofrepeated moderate push-ups ofthe total body invoked by bending the legs. sP Strong push-ups: this involves a period ofrepeated strong push-ups ofthe total body invoked by strongly bending the legs and leading tojumps. Po Poking: repeated pokingofthefemalewith theforelegs; forelegsarepositioned in front ofthemaleand parallelwith thesubstrate. First, we tested whether males courted differently; i.e., distinguished, based on some unique male courting behaviors. expressed different courting behaviors to heterospecific Hogna snodgrassi males often court for extremely long periods females than to conspecific females. This was done for each (up to 12 min) by combining palpal drumming with strong species by comparing the probability of occurrence of each push-ups (sP) (Table 3). In comparison, H. jimco males male courtship behavior toward the females of different generallycourt formuch shorterperiods andcombine pedipalp species by means ofexact Pearson Chi-Square tests (StatXact- drumming and quick movements toward the female, and if 5). Ifno differences were observed in the presence or absence distance is small, males poke the females repeatedly with their of male courtship behavior with respect to the species of the forelegs(Po). Differencesincourtshipbetweenmalesofthetwo female, data for heterospecific and conspecific trials were SantaCruzspecies; i.e., H. galapagoensisand H. hendrickxi, are pooled to describe species-specific male courtship. Second, much less evident, and both species combine palpal drumming, interspecific differences in the probability of expressing a elaborate movements of the forelegs and moderate push-ups specific courtship behavior among males of the different while courting. The latter courtship trait was not observed for species were tested with exact Pearson’s Chi-Square tests to the two species from San Cristobal. Although our quantitative look for the presence ofspecies-specific courtship behaviors. measurements ofthecourtshipofboth Santa Cruz specieswere — Inter- vs. intraspecific courting. Investigating the potential not significantly different (Table 3), some subtle differences for reproductive isolation should ideally be based upon were observed, wherein H. hendrickxi males tended to make heterospecific and conspecific choice and no-choice trials more use ofthe second pair of forelegs than H. galapagoensis using virgin adults. Here, the lack ofvirgins impedes us from males and often moved their pedipalps sidewards while investigating mate preferences directly. However, heterospe- drumming (C. De Busschere pers. observ.). In sum, the species cific mating trials allowed us to observe whether heterospecific on San Cristobal, H. junco and H. snodgrassi, are clearly females elicited male courting behavior and whether females distinguishable, based upon unique ma—le courtship behaviors. could distinguish and reject heterospecificmales. Observations Inter and intraspecific copulations. Males apparently did of heterospecific acceptances under laboratory conditions not prefer conspecific females, as the number of courtship might indicate the presence of weak premating barriers. By events a male displayed was not significantly different when means ofexact Pearson’s Chi-Square tests, we tested whether exposed to heterospecific females (P > 0.45). Although the the probability ofmalecourtingdiffered with respect to female acceptance rate of courting males was in general very low (10%), few heterospecific mating events were observed, and species. the acceptance rate among species did not differ from random RESULTS (A“ = 2.6, P = 0.46) (Fig. 2). Remarkably, despite clear — Interspecific comparison in male courtshipbehavior. Table 2 differences in morphology, H. galapagoensis females accepted gives an overview ofthe total number oftrials performed and heterospecific males ofPI. hendrickxi Moreover, H. galapoen- the number of trials used in the analysis of interspecific sis females also accepted heterospecific males from the comparisons ofmale courtship behavior. Given that particular distantly related H. junco, which has a distinctively different courtship traits were consistently expressed irrespective of the male courtship (Poking) (Fig. 2). species of the female to which the male was exposed to (P > DISCUSSION 0.17), male courtship data were pooled across female species. — Movementsofthepedipalps(PM)were observed in males in all Interspecific differences in male courtship behavior. The four species (Table 3). For the other courtship traits, large mating trials revealed that the high-elevation species H. junco differences were observed among species (Table 3). Courtship and thecoastal species H. snodgrassi, both from San Cristobal, of males of both species from San Cristobal can be clearly show distinct male courtship behaviors. In contrast, our quantitative analysis based on five male courtship traits did — not reveal any significant differences between H. hendrickxi Table 2. Sample sizes of total trials and, in parentheses, trials and H. galapagoensis on Santa Cruz. The lack of difference used in courtship analysis. between these species is in strong concordance with earlier Females studies (Baert et al. 2008b; De Busschere et al. 2012) that noted almost identical male genital traits forH. hendrickxiand Males gala hend snod juiic H. galapagoensis, which clearly differ from those of the San gala 37 (12) 9 (4) - 13 (5) Cristobal species. Furthermore, De Busschere et al. (2012) lieml 13 (5) 13 (6) 3(1) - observed clear interspecific differences in male genital traits .snod - 4(1) 16(9) 6(0) between H. junco and H. snodgrassi. Hence, the variation in June 11 (9) - 5 (4) 16 (10) male courtship behavior appears congruent with the diver- THE JOURNAL OF ARACHNOLOGY 28 — Table 3. Interspecific comparison ofmale courtship behaviors. Occurence gala heml June snod P Total trials 21 12 23 10 PM 21 12 22 10 1.90 1 FM 21 12 2 10 57.58 <0.0001 mP 19 9 0 0 49.38 <0.0001 sP 1 0 0 10 59.14 <0.0001 Po 0 0 21 0 57.58 <0.0001 Total courting time t,ot (s) (± SE) 101 ± 16 147 ± 36 50 ± 10 301 ± 72 gence pattern in two male genital structures involved in the influence the outcome of these mating trials (Uetz & Roberts copulation process. In contrast, the variation in male 2002; Roberts & Uetz 2004). Indeed, the latter might be courtship behavior contrasts with the ecological divergence expected, as palpal drumming dominates male courtship, and into morphologically distinct high elevation and coastal dry both drumming and push-up movements might result in species. Therefore, species with a similar habitat preference, vibrations being transmitted through the substrate. In general, which are highly similar in color pattern and in non-genital female acceptance rate of courting males was low (10%), traits (De Busschere et al. 2012), share no similarities in male which was probably due to the use ofwild-caught individuals. courtship behavior, and hence this observation does not The latter refers to the possibility that wild-caught females suggest that these mating traits evolved as a byproduct of were already fertilized in the field, leading to a possible natural selection. Additionally, the variation in male courtship rejection of courting males in the laboratory (Fernandez- behavior is rather in disagreement with the phylogenetic Montraveta & Ortega 1990; Rypstra et al 2003). Despite the relationships, as the more recently diverged H. jiinco and H. generally small volume of data, which does not permit us to MY snodgrassi (—0.1 ago) are much more different in male test for species-specific acceptance rates, females of H. courtship behavior than H. hemirickxi and H. galapagoensis, galapagoensis accepted heterospecific males. Remarkably, MY which diverged approximately 0.8 ago, albeit under low despite clear differences in morphology, females of H. levels of gene flow (De Busschere et al. 2010). In sum, on galapagoensis accepted heterospecific males of the genetically Santa Cruz and San Cristobal, parallel within-island specia- closely related sister species H. liendrickxi. Moreover, H. tion is only refiected in ecologically relevant traits and not in galapagoensis females also accepted heterospecific males from male courtship behavior. This incongruence indicates that, the distantly related but morphologically highly similar H. beside the similar and strong natural selection, different junco, which has a distinct malecourtship (Poking). Therefore, processes influenced the divergence ofmating traits. despite clear interspecific differences in male courtship — Weak prezygotic mating barriers. Whether the above- behavior and/or morphology, within and between-island mentioned interspecific differences have the potential to acceptances were observed, which suggests that interspecific function as premating isolating mechanisms was investigated prezygotic mating barriers are weak for H. galapagoensis. This by performing interspecific mating trials. Beside the differ- contrasts sharply with other wolf spider studies, in which encesdenoted in malecourtship and morphology, we notethat distinct differences in courtship behavior serve as a strong hitherto undescribed chemical and vibratory cues might also prezygotic mating barrier (Den Hollander & Dijkstra 1974; — Figure 2. Numberoftrialswith nomalescourting(white),acceptanceofcourtingmale(black), aggression againstcourtingmale(light gray) and no response to courting male (dark gray). Total number oftrials are noted above bars. DE BUSSCHERE & HENDRICKX—BEHAVIORAL ISOLATION OF WOLF SPIDERS 29 Uetz & Denterlein 1979; Topfer-Hofmann et al. 2000; Chiaiie support of the Belgian Science Policy (BELSPO, research et al. 2010). Furthermore, it is remarkable to note potentially project MO/36/025), FWO-Flanders (PhD- grant to CDB and weak prezygotic mating barriers, taking into account the deep research project G.0202.06), Leopold III fund and the Royal divergence times [H. galapagoensis-H. hendrickxi: ~0.8MY Belgian Institute ofNatural Sciences (RBINS). and H. galapagoensis-H. jimco: ~1.6MY ago (De Busschere LITERATURE CITED et al., 2010j] and strong ecotypic divergence. The lack of premating barriers has also been found between allopatric Andersson, M. 1994. Sexual Selection. Princeton University Press, lineages of warbler finches on Galapagos, despite differences Princeton, New Jersey. in song and a long divergence time (1.5-2 MY) (Grant & Baert, L., F. Hendrickx & J.-P. Maelfait. 2008a. Allozyme Grant 2002). Weak prezygotic barriers might be explained by characterization of species of the lycosid genus Hogna of the Galapagos archipelago. Journal ofArachnology 36;411-417. the lack ofor weak selection against hybridization due to low Baert, L., J.-P. Maelfait & F. Hendrickx. 2008b. The wolf spiders levels of gene flow, and hence a predominant divergence in (Araneae, Lycosidae) from the Galapagos Archipelago. Bulletin allopatry (Coyne & Orr 1989). Therefore, females of H. from theRoyal Belgian InstituteforNatural Sciences, Entomology galapagoensis were not forced to recognize heterospecific 78:5-37. males. The lack of frequent interactions has also been Baert, L., J.-P. Maelfait, F. Hendrickx & K. Desender. 2008c. suggested for the absence of strong mating isolation between Distribution and habitat preference of the spiders (Araneae) of stream and lake sticklebacks (Raeymaekers et al. 2010). Galapagos. Bulletin from the Royal Belgian Institute for Natural Indeed, range overlap between both ecotypes on Santa Cruz Sciences, Entomology 78:39-111. and San Cristobal might have been limited to periods of Barrett, R.D.H. & D. Schluter. 2008. Adaptation from standing environmental and climatological change (De Busschere et al. genetic variation. Trends in Ecology and Evolution 23:38-44. Boughman, J.W. 2001. Divergent sexual selection enhances repro- 2010 ). ductive isolation in sticklebacks. Nature 411:944-948. Despite the current spatial isolation, the potential for weak Boughman, J.W., H.D. Rundle & D. Schluter. 2005. Parallel prezygotic mating barriers points out that Hogna species evolution ofsexual isolation in sticklebacks. Evolution 59:361-373. boundaries, especially ofH. galapagoensis. could be fragile in Coyne, J.A. & H.A. Orr. 1989. Patterns ofspeciation in Drosophila. the case offuture secondary contact. Moreover, the potential Evolution 43:362-381. forweak prezygotic reproductive barriers for H. galapagoensis Coyne, J.A. & H.A. Orr. 1997. “Patterns ofspeciation in Drosophila" in combination with the detection of ancient hybridization revisited. Evolution 51:295-303. events between H. galapagoensis and H. hendrickxi (De Coyne, J.A. & H.A. Orr. 2004. Speciation. Sinauer, Sunderland, Busschere et al. 2010) and of inter-island dispersal of H. Massachusetts. galapagoensis (Fig. 1), suggest a potential role of within and De Busschere, C., L. Baert, S.M. van Belleghem, W. DeKoninck& F. between-island gene flow in the Hogna radiation. Further Hendrickx. 2012. Parallel phenotypic evolution in a wolf spider radiation on Galapagos. Biological Journal ofthe Linnean Society exploration of the potential role of gene flow should not be 106:123-136. neglected in understanding the Hogna radiation on Galapa- De Busschere, C., F. Hendrickx, S.M. van Belleghem, T. Backeljau, gos, as hybridization among diverging populations might L. Lens & L. Baert. 2010. Parallel habitat specialization within the enhance the spread of adaptive genetic variation and as such wolfspider genus Hogna from the Galapagos. Molecular Ecology catalyze adaptive divergence (Seehausen 2004; Barrett & 19:4029-4045. Schluter 2008; Schluter & Conte 2009) and facilitate recurrent Del Solar, E. 1966. Sexual isolation caused by selection for positive phenotypic evolution. However, the possibility of other and negative phototaxis and geotaxis in Drosophilapscudoohscura. mating barriers such as assortative mating related to habitat Proceedings of the National Academy of Sciences of the United preference (Rundle et al. 2000; Boughman 2001), natural States ofAmerica 56:484M87. selection against migrants and hybrids (Hendry 2004; Nosil & Den Hollander, J. & H. Dijkstra. 1974. Pardosa vlijinisp. nov., a new Crespi 2004) and the role of mechanical and postzygotic ethospecies sibling Pardosa proxinui (C.L. Koch, 1948) from France, with description of courtship display (Araneae, Lycosi- isolation mechanisms, definitely needs further assessment. dae). Beaufortia 22:57-65. In sum, this study provided an initial view of the role of Dobzhansky, T. 1937. Genetics and the Origin ofSpecies. Columbia behavioral isolation among habitat-specialized wolfspiders on University Press, New York. the Galapagos and emphasizes the need for further assessment Fernandez-Montraveta, C. & J. Ortega. 1990. Some aspects of the ofthedegree ofreproductive isolation and thepotential role of reproductive behavior of Lycosa tarentidafasciiventris (Araneae, within and between-island gene flow to understand the Hogna Lycosidae). Journal ofArachnology 18:257-262. radiation on the Galapagos. Futuyma, D.J. 2005. Evolution. Sinauer Associates, Sunderland, Massachusetts. ACKNOWLEDGMENTS Gleason, J.M. & M.G. Ritchie. 1998. Evolution ofcourtshipsongand We gratefully acknowledge H. Herrera, L. Cruz and station reproductive isolation in the Drosophila willistoni species complex: workers for arranging the logistics, administrational support dosexualsignalsdivergethemostquickly?Evolution 52:1493 1500. and field assistance. Scientific research permits for the Grant, B.R. & P.R. Grant. 2002. Lack ofpremating isolation at the collecting trips were obtained from the Barque Nacional Hebbeatsse,oEf.Aa.p&hyGlo.gWe.neUteitcz.tre2e0.00A.meLreigcoarnnaNmateunrtaaltiisotn1a6n0:d1-t1h9e.efficacy Galapagos through the Department of Terrestrial Inverte- of courtship display in four species of wolf spider (Araneae: brates at the Charles Darwin Research Station (Santa Cruz). Lycosidae). Behavioral Ecology and Sociobiology 47:280-286. We thank two anonymous reviewers for their constructive Hendry, A.P. 2004. Selection against migrants contributes to the comments, which helped us to improve the manuscript.This rapid evolution of ecologically dependent reproductive isolation. work could not have been conducted without the financial Evolutionary Ecology Research 6:1219-1236. THE JOURNAL OF ARACHNOLOGY 30 Loosveldt, K.2004. Biometrischeenpopulatiegenetischestudievanhet Rundle, H.D. & P. Nosil. 2005. Ecological speciation. Ecology genus Hogna (Araneae, Lycosidae) op Isla Santa Cruz (Galapagos. Letters 8:336-352. Ecuador). M.Sc. thesis. Ghent University, Ghent, Belgium. Rypstra, A.L., A.M. Schlosser, P.L. Sutton & M.H. Persons. 2009. Masta, S.E. & W.P. Maddison. 2002. Sexual selection driving Multimodal signaling: the relative importance of chemical and diversification in jumping spiders. Proceedings of the National visual cues from females to the behaviour of male wolf spiders AcademyofSciencesoftheUnitedStatesofAmerica99:4442-4447. (Lycosidae). Animal Behaviour 77:937-947. McKinnon, J.S. & H.D. Rundle. 2002. Speciation in nature: the Rypstra, A.L., C. Wieg, S.E. Walker & M.H. Persons. 2003. Mutual threespine stickleback model systems. Trends in Ecology and mate assessment in wolf spiders: differences in the cues used by Evolution 17:480^88. males and females. Ethology 109:315-325. Mendelson, T.C. 2003. Sexual isolation evolves faster than hybrid Schluter, D. 2000. The Ecology of Adaptive Radiation. Oxford inviability in a diverse and sexually dimorphic genus of fish University Press, New York. (Percidae: Etheostoma). Evolution 57:317-327. Schluter, D. & G.L. Conte. 2009. Genetics and ecological speciation. Miller, G.L., G.E. Stratton, P.R. Miller & E.A. Hebets. 1998. Proceedings of the National Academy of Sciences of the United Geographical variation in male courtship behaviour and sexual States ofAmerica 106:9955-9962. isolation in wolf spiders of the genus Schizocosa. Animal Seehausen, O. 2004. Hybridization and adaptive radiation. Trends in Behaviour 56:937-951. Ecology and Evolution 19:198-207. Nosil, P. & B.J. Crespi. 2004. Does gene flow constrain adaptive Topfer-Hofmann, G., D. Cordes & O. Von Helversen. 2000. Cryptic divergence or vice versa? A test using ecomorphology and sexual species and behavioural isolation in the Pardosa lugubris group isolation in Timeina cristinaewalking-sticks. Evolution 58:102-112. (Araneae, Lycosidae), with description oftwo newspecies. Bulletin Panhuis,T.M.,R.Butlin,M.Zuk&T.Tregenza.2001. Sexualselection ofthe British Arachnological Society 11:257-274. and speciation. Trends in Ecology & Evolution 16:364-371. Uetz, G.W. 2000. Signals and multi-modal signaling in spider Raeymaekers, J.A.M., M. Boisjoly, M. Boisjoly, L. DeLaire, D. communication. Pp. 387^05. In Animal Signals: Signaling and Berner, K. Rasanen & A.P. Hendry. 2010. Testing for mating Signal Design in Animal Communication. (Y. Espmark, T. isolation between ecotypes: laboratory experiments with lake, Amundsen & G. Rosenqvist, eds.) Tapir, Trondheim, Norway. stream and hybrid stickleback. Journal of Evolutionary Biology Uetz, G.W. & G. Denterlein. 1979. Courtship behavior, habitat, and 23:2694-2708. reproductive isolation in Schizocosa rovneri Uetz and Dondale Roberts, A.J. &G.W. Uetz. 2004. Chemical signalinginawolfspider: (Araneae, Lycosidae). Journal ofArachnology 7:121-128. A test ofethospecies discrimination. Journal ofChemical Ecology Uetz, G.W. &J.A. Roberts. 2002. Multisensorycues and multimodal 30:1271-1284. communication in spiders: insights from video/audio playback Rundle, H.D., L. Nagel, J.W. Boughman & D. Schluter. 2000. studies. Brain, Behavior and Evolution 59:222-230. Natural selection and parallel speciation in sympatric sticklebacks. Science 287:306-308. Manuscript received9 July 2012, revised5 December2012.