Bull.Br.arachnol.Soc.(2010)15(2),55–58 55 Predation by three species of spiders on a cave fish crickets and a species of veliid water strider at high in a Mexican sulphur cave abundance.Duringourannualfieldworkinthecave,we alsorepeatedlynoticedseverallargerspeciesofAraneae Joachim Horstkotte1, Rüdiger Riesch2*, Martin Plath2,3 sitting on stones near the water’s edge with the tarsi of and Peter Jäger4 their front legs sometimes touching the water surface, 1Biocenter Grindel, University of Hamburg, Martin-Luther-King- potentially also preying on surfacing fish. Further, on Platz3,20146Hamburg,Germany one of our trips into the cave in 2008, we observed a 2Department of Zoology, University of Oklahoma, 730 Van Vleet largeunidentifiedctenidthatenteredthewaterbodyand Oval,Norman,OK73019,USA actively swam across a 3-m pool of water to emerge on 3DepartmentofEcologyandEvolution,Goethe-UniversityFrankfurt theotherside.Clearly,atleastsomeofthelarge-bodied amMain,Siesmayerstrasse70a,60054FrankfurtamMain,Germany 4Research Institute Senckenberg, Senckenberganlage 25, 60325 Araneae observed in the cave do not completely shun FrankfurtamMain,Germany the toxic waters. Furthermore,otherspidersarealsoknowntohunton thewatersurfaceandsomeofthemfeedbothonfishor Summary amphibians and insects, e.g. Dolomedes Latreille, 1804 Previously,giantwaterbugs(Belostomasp.,Belostoma- (Pisauridae; worldwide) (Breene, Sweet et al., 1988; tidae) were thought to be the only predators of the cave- Suter, 1999; Bleckmann & Bender, 1987), Thalassius dwelling fish Poecilia mexicana (Poeciliidae) in a southern Simon,1885(Pisauridae;Africa,Asia)(Abraham,1923; Mexicansulphurcave(CuevadelAzufre).Here,wereport on three species of Araneae [Ctenidae gen. sp., Tinus sp. Sierwald,1988),AncylometesBertkau,1880(Pisauridae; (Pisauridae), and Hemirrhagus pernix (Theraphosidae)] South America) (Schiapelli & Gerschman, 1970), Tre- from that cave, which also prey on P. mexicana. To our chalea Thorell, 1869 (Trechaleidae; South America) knowledge,thisisthefirstreportofspiderpredationona (Berkum, 1982), Pirata Sundevall, 1832 and Pardosa cavefish. C. L. Koch, 1847 (both Lycosidae; worldwide) (Breene, Sweet et al., 1988), Heteropoda natans Jäger, 2005 (Sparassidae;Sabah,Borneo)(Airamé&Sierwald,2000; Introduction Jäger, 2005). Diving has been reported for Dolomedes and Thalassius spinosissimus Simon, 1891 (Sierwald, Caves have often been assumed to be predator-free 1988; Jäger, unpubl. data). Hence, spider predation on environments for fish living in them (Romero & Green, cave mollies seems highly likely, and the present study 2005; Tobler, Schlupp et al., 2006). Only recently we represents the first attempt to investigate the presumed have demonstrated a previously unreported case of a role of fish predation by several species of Arachnida in predator–prey interaction between an aquatic insect of this particular cave system. thegenusBelostomaLatreille,1807(Heteroptera:Belos- tomatidae) and a cave fish (i.e. the cave molly), a cave-adapted form of the Mexican livebearer Poecilia Material and methods mexicana Steindachner, 1863 (Teleostei: Poeciliidae; During September 2008 we were able to collect ten Tobler, Schlupp et al., 2007; Tobler, Franssen et al., individuals of large-bodied Arachnida in various parts 2008;Tobler,2009).InthesouthernMexicanCuevadel of the Cueva del Azufre (location of cave entrance: Azufre (also known as Cueva de las Sardinas or Cueva 17(26#32$N, 92(46#31$W). An additional experiment de Villa Luz), cave mollies not only live in permanent wasconductedwithonespiderinSeptember2004(Table darkness,butalsohavetocopewithhighconcentrations 1, Fig. 1). We used an experimental approach as de- of hydrogen sulphide (H S; Gordon & Rosen, 1962; 2 scribed in previous studies (see Tobler, Schlupp et al., Tobler, Schlupp et al., 2006), which are acutely toxic to 2007; Tobler, Franssen et al., 2008 for methodological metazoans (Grieshaber & Völkel, 1998). Hydrogen sul- details) to test which Arachnida would prey on mollies. phide leads to extreme hypoxia in the water (Tobler, Briefly,individualarachnidswereinitiallyhousedinside Schlupp et al., 2006); hence cave mollies perform so- the cave in empty, perforated 1.5-l plastic bottles and calledAquaticSurfaceRespiration(ASR)toexploitthe were not fed for a 48h acclimation period. At the start more oxygenated (and thus less sulphidic) top-most of the experiment, the arachnids were transferred into layer of the water column (Plath, Tobler et al., 2007; individual perforated 5-l plastic bottles, which were Tobler, Riesch et al., 2009). This behaviour exposes the submerged in the water to approximately one fourth of mollies to predation by the sit-and-wait predator Belos- their depth. We also introduced four mollies per bottle toma sp., which waits for passing, surfacing fish at the edge of the water (Tobler, Schlupp et al., 2007; see also as potential prey items [2 small (mean!SD length: Kramer,Manleyetal.,1983forincreasedpredationrisk 18.75!2.22mm) and two large (28.20!2.86mm)]. Each bottle was checked for fish-predation after 24 h for fish exhibiting ASR). and, if predation had occurred, the experiment was The high abundance of Goeldichironomus fulvipilus terminatedandthearachnidcollectedforspeciesidenti- (Rempel, 1939) (Diptera: Chironomidae) in the Cueva fication. If no predation was observed, the bottles were del Azufre may be the reason for a rich araneofauna left on site for an additional 24h. Unfortunately, we with mostly small-sized species (Lavoie & Evans, 2002). were unable to recover three bottles after the second We also observed two species of unidentified ensiferan 24-htrialperiod,possiblybecausetheywereremovedby *Correspondingauthor’se-mail:[email protected] other visitors to the cave. 56 Predationbyspidersoncavefish Results Macrobrachium Bate, 1868) on Tobago at sites where prawnsoccuratsimilarhighdensitiestothecavefishin Among the remaining 8 bottles predation had oc- our study system (Ladle & Velander, 2003). curred in three cases: We used 5-l bottles for our experiment, and through- (1) One adult female Hemirrhagus pernix (Ausserer, outthecave,alargenumberofsmallpoolsareapproxi- 1875)(Orthognatha:Theraphosidae),whichwasthefirst mately similar in size to our experimental setting; record of a female ever known of this species, had however, the central pools of each cave chamber are consumed all four fish [prosoma (PS) length 13.3mm, quite large and are several square metres in diameter. opisthosoma (OS) length 18.3mm; Table 1, Fig. 1D]. Does this mean that our experimental conditions were (2) An adult female of the genus Tinus (Pisauridae) not an adequate representation of the situation in the had consumed two fish and when collected still held a cave?Wedonotthinkso,becausealthoughmostofthe half-consumed third molly between the chelicerae (PS arachnidsweusedforthisexperimentwerecaughtatthe 6.0mm, OS 8.1mm; Table 1, Fig. 1A). According to edgeofthemain(large)poolsofeachcavechamber,this illustrationsinCarico(1976)thisspiderisclosetoTinus is exactly where cave mollies can be found in highest palictlusCarico,1976,knownonlyfromitstypelocality, densities (even higher than the densities we simulated San Luis Potosi, c. 1000km NW of Cueva del Azufre. here) close to the water surface feeding on mosquito From this single female it is hard to judge whether it larvae attached to the peripheral rocks at the air-water represents a new species, although some differences in interface (Tobler, Schlupp et al., 2006; Tobler, 2008). the female copulatory organ suggest so. Therefore, despite the fact that fish may be able to (3) One juvenile ctenid had caught and half- escapemoreeasilyinthenaturalsettingsofthecave,we consumed a cave molly (PS 7.4mm, OS 7.7mm; Table think our microcosms were of appropriate size, because 1, Fig. 1C). we did not try to quantify capture rates, but rather to In contrast, another ctenid, a subadult male, had not determinewhetherthesearachnidscouldpreyonthefish preyed on the cave mollies (PS 12.8mm, OS 15.0mm; or not. Table 1, Fig. 1B). A further Tinus sp. and two un- Despite the small number of individuals used for the identifiedctenidshadnoteatenanyfish,norhadanyof experiments,ourfindingsareimportantforatleastthree the three specimens of the very abundant amblypygid reasons: (1) this cave in particular is clearly not a Phrynus sp. (Amblypygi: Phrynidae) consumed a molly predator-free environment for the cave fish. Although (Table 1, Fig. 1E). lack of predation has been assumed for other cave systems housing fishes (Romero & Green, 2005), the Discussion apparent predation by Belostoma sp. (Plath, Parzefall et al., 2003; Tobler, Schlupp et al., 2007; Tobler, Our results indicate that giant water bugs, Belostoma Franssen et al., 2008; Tobler, 2009) and three different sp., are not the only predators of P. mexicana in the speciesofspiders(thisstudy)clearlynegatesthisnotion. Cueva del Azufre. At least three species of large spiders (2)Araneae,andnotBelostomasp.,mayactuallybethe (Tinus sp., Ctenidae gen. sp., and Hemirrhagus pernix) top-predators in this subterranean ecosystem. Presum- are also (at least partly) fish predators in this system. ably,thethreespeciesofspidersdescribedherealsoprey Furthermore,thisisprobablyaconservativeassessment, on other organisms living in the cave, possibly even because several of our test specimens may simply not includingBelostomasp.(averagelengthfromtipofhead have been motivated to forage (i.e. were not hungry) to end of abdomen: 20mm; Tobler, Franssen et al., duringourrelativelyshortexperiment.Forexample,we 2008), but future experiments will need to investigate did not find evidence for predation by the abundant this further. Finally, (3) predation by spiders may have amblypygid(Phrynussp.),althoughanotheramblypygid profound influences on behavioural and life-history species [Heterophrynus cheiracanthus (Gervais, 1844)] traits in cave-dwelling P. mexicana (see Riesch, Plath is reported to prey on freshwater prawns (genus Species(stage,sex) Family Fieldcode Result(after24h) Notes Hemirrhaguspernix(adultfemale) Theraphosidae JH-10 4fishconsumed Tinussp.(adultfemale) Pisauridae JH-1 3fishconsumed 1fishstillheldincheliceraewhencollected sp.indet.(juvenile) Ctenidae JH-3 1fishcaught(dead) sp.indet.(subadultmale) Ctenidae JH-2 – sp.indet. Ctenidae JH-6 – Tinussp. Pisauridae JH-5 – Builtwebinbottle Phrynussp. Phrynidae(Amblypygi) JH-4 – Phrynussp. Phrynidae(Amblypygi) JH-7 – Phrynussp. Phrynidae(Amblypygi) JH-8 – sp.indet. Ctenidae JH-9 Noresult ‘‘Stolen’’beforeevaluationwaspossible (sameasJH-2) Ctenidae MP-1 – September2004 Table1: Summary results of the predation experiments in the Mexican sulphur cave (Cueva del Azufre). – refers to all fish alive after 24hours. J.Horstkotte,R.Riesch,M.Plath&P.Jäger 57 A.Ramm,K.Scharnweber,M.SchulteandM.Ziege)as well as T. Schulz-Mirbach (University of Munich) for their help in the field in 2008 and M. Tobler (Texas A&M University) for help in 2004. We thank P. Weygoldt, Freiburg (Germany), and E. Hebets, University of Nebraska (USA) for valuable discussion and help with species identification. Hemirrhagus pernix was determined by J.-P. Rudloff/Rosslau (Germany). The Mexican Government (Permiso de Pesca de Fomento No. DGOPA.06192.240608.-1562) and the Municipal of Tacotalpa (SM/1133/208) kindly provided permitsfortheworkattheCuevadelAzufre.Financial support came from the DFG (PL 470/1–2) and from NSF (DEB-0743406). References ABRAHAM,N.1923:Observationsonfishandfrog-eatingspidersof Natal.Ann.NatalMus.5:89–95. AIRAMÈ,S.&SIERWALD,P.2000:Huntingandfeedingbehavior of one Heteropoda species in lowland rainforest on Borneo (Araneae,Sparassidae).J.Arachnol.28:251–253. BERKUM, F. H. van 1982: Natural history of a tropical, shrimp- eatingspider(Pisauridae).J.Arachnol.10:117–121. BLECKMANN, H. & BENDER, M. 1987: Water surface waves generated by the male pisaurid spider Dolomedes triton (Walckenaer)duringcourtshipbehavior.J.Arachnol.15:363– 369. BREENE, R. G., SWEET, M. H. & OLSON, J. K. 1988: Spider predatorsofmosquitolarvae.J.Arachnol.16:275–277. CARICO, J. E. 1976: The spider genus Tinus (Pisauridae). Psyche, Camb.83:63–78. GORDON,M.S.&ROSEN,D.E.1962:Acavernicolousformofthe poeciliidfishPoeciliasphenopsfromTabasco,Mexico.Copeia 1962:360–368. GRIESHABER,M.K.&VÖLKEL,S.1998:Animaladaptationsfor tolerance and exploitation of poisonous sulphide. A. Rev. Physiol.60:33–53. JÄGER, P. 2005: A ‘swimming’ Heteropoda species from Borneo (Araneae:Sparassidae:Heteropodinae).J.Arachnol.33:715– 718. KRAMER,D.L.,MANLEY,D.&BOUGEOIS,R.1983:Theeffects of respiratory mode and oxygen concentration on the risk of aerialpredationinfishes.Can.J.Zool.61:653–665. LADLE,R.J.&VELANDER,K.2003:Fishingbehaviourinagiant Fig.1: Photographs of four species of Arachnida used in the preda- tionexperiment.ATinussp.(Pisauridae);B,CCtenidaegen. whipspider.J.Arachnol.31:154–156. sp.; D Hemirrhagus pernix (Theraphosidae); E Phrynus sp. LAVOIE,K.&EVANS,K.2002:Distributionandabundanceofthe midge Goeldichironomus fulvipilus, in Cueva de Villa Luz, (Phrynidae,Amblypygi).FishwereconsumedbyA,CandD. Tabasco,Mexico.J.CaveKarstStud.64:181. NotethatAhadjustkilledthefishshowninthephoto,while PLATH, M., PARZEFALL, J. & SCHLUPP, I. 2003: The role of Chadalreadypartiallyconsumedthefishshown. sexualharassmentincave-andsurface-dwellingpopulationsof the Atlantic molly, Poecilia mexicana (Poeciliidae, Teleostei). Behav.Ecol.Sociobiol.54:303–309. etal.,inpress,foradiscussion).Forexample,predation PLATH,M.,TOBLER,M.,RIESCH,R.,GARCuIADELEÓN,F. by piscivorous fishes on the related Trinidadian guppy, J., GIERE, O. & SCHLUPP, I. 2007: Survival in an extreme Poecilia reticulata, has been shown to profoundly affect habitat:therolesofbehaviourandenergylimitation.Naturwis- age-structures, size at maturity, and overall fecundity senschaften94:991–996. (e.g. Reznick & Endler, 1982). The question of whether REZNICK,D.&ENDLER,J.1982:Theimpactofpredationonlife history evolution in Trinidadian guppies (Poecilia reticulata). andhowpredationalsoaffectslife-historyevolutionand Evolution36:160–177. behavioural traits (e.g. escape behaviour) in the cave RIESCH,R.,PLATH,M.&SCHLUPP,I.Inpress:Toxichydrogen molly clearly warrants further research. sulfideanddarkcaves:lifehistoryadaptationsinalivebearing fish(Poeciliamexicana,Poeciliidae).Ecology. ROMERO, A. & GREEN, S. M. 2005: The end of regressive Acknowledgements evolution: examining and interpreting the evidence from cave fishes.J.FishBiol.67:3–32. Wewouldliketothanktheparticipantsofthe‘‘Tropi- SCHIAPELLI,R.D.&GERSCHMAN,B.S.1970:Consideraciones cal Ecology’’ course of the University of Potsdam (A. sobre el género Ancylometes Bertkau 1880 (Araneae: Pisauri- Böttger, N. Bunzel, V. Duwe, N. Herrmann, L. Padur, dae).Actazool.lilloana27:155–175. 58 Predationbyspidersoncavefish SIERWALD, P. 1988: Notes on the behavior of Thalassius spinosis- Lifeontheedge:hydrogensulfideandthefishcommunitiesof simus (Arachnida: Araneae: Pisauridae). Psyche, Camb. 95: a Mexican cave and surrounding waters. Extremophiles 10: 243–252. 577–585. SUTER, R. B. 1999: Cheap transport for fishing spiders (Araneae, TOBLER,M.,SCHLUPP,I.&PLATH,M.2007:Predationofacave Pisauridae): the physics of sailing on the water surface. J. fish (Poecilia mexicana, Poeciliidae) by a giant water-bug Arachnol.27:489–496. (Belostoma, Belostomatidae) in a Mexican sulfur cave. Ecol. TOBLER, M. 2008: Divergence in trophic ecology characterizes Ent.32:492–495. colonization of extreme habitats. Biol. J. Linn. Soc. 95: 517– TOBLER, M., FRANSSEN, C. & PLATH, M. 2008: Male-biased 528. predationofacavefishbyagiantwaterbug.Naturwissenschaf- TOBLER,M.2009:Doesapredatoryinsectcontributetothediver- ten95:775–779. gencebetweencave-andsurface-adaptedfishpopulations?Biol. TOBLER,M.,RIESCH,R.,TOBLER,C.M.&PLATH,M.2009: Lett.5:506–509. Compensatorybehaviorinresponsetosulfide-inducedhypoxia TOBLER, M., SCHLUPP, I., HEUBEL, K. U., RIESCH, R., affects time budgets, feeding efficiency, and predation risk. GARCuIADELEÓN,F.J.,GIERE,O.&PLATH,M.2006: Evol.Ecol.Res.11:935–948.