2007 (2008). The Journal ofArachnology 35:530-534 SHORT COMMUNICATIONS JUVENILE NEPHILA (ARANEAE, NEPHILIDAE) USE VARIOUS ATTACK STRATEGIES FOR NOVEL PREY Linden Higgins: Department of Biology, University ofVermont, Burlington, Vermont 05405, USA. E-mail: [email protected] ABSTRACT. Nephila clavipes (Linnaeus 1767) and N. pilipes (Fabricius 1793)juveniles exposed to a novel and potentially dangerous prey item frequently attack using thrown silk. To quantify the frequency with which N. clavipes opt to use thrown silk, naive hand-reared small N. clavipesjuvenile females were observed attacking a new prey type, stingless bees. Repeated exposure to the stingless bees suggests that the spiders incorporate priorexperience into prey attack strategies, as experienced spiders attacked using themoreusual Nephila long-bite. Keywords: Nephila clavipes, Nephilapilipes, wrap-attack, prey capture, learning Orb-web building spiders utilize a range oftactics capacity to learn, experiments with various species to capture insects that are restrained in the web, and have shown that prior experience affects web the choice of tactic varies among species and with architecture (e.g., Heiling & Herberstein 1999; prey type (Robinson & Mirik 1971; Robinson & Nakata & Ushimaru 2004; Prokop 2006), propensity Robinson 1976; Eberhard 1982; Japyassu & Viera to attack particular prey (Herberstein et al. 1998), 2002). Oneprey-capturetactic, throwingsilkwithleg and searching behavior when captured insects are IV, is widely used in phylogenetic analyses of the removed (Rodriguez&Gamboa2000). Althoughthe families of entelegyne spiders (e.g., Sharff & Cod- conditions under which spiders are stimulated to dington 1997; Griswoldetal. 1999; Agnarsson2004). attack via thrown silk have been studied for several In these analyses, the family Nephilidae (previously species of Araneidae (Robinson 1975; Robinson & a tetragnathid subfamily Nephilinae; Kuntner 2006) Robinson 1976), there are no published records was assumed to lack this behavior and to attack all indicatingaroleofpriorexperienceincapturetactics prey by biting (Robinson & Robinson 1976; Eber- used by orb-weaving spiders. Rather, researchers hard 1982; Kuntner 2005a, 2006, 2007) which, since haveconcludedthatthesequenceofattackbehaviors the wrap attack is synapomorphic for the orbicular- is a direct response to the prey type and size and the ians (Griswold et al. 1999; Agnarsson 2004), made success or failure of a given tactic during an attack this a secondary loss in the Nephilidae. This (Robinson & Mirick 1971; Robinson et al. 1971; character assignment was based upon observations Robinson & Robinson 1976; Herberstein et al. 1998; ofadult female Nephila species. However, failure to Japyassu &Viera 2002). However, these experiments observe attacks using thrown silk in diverse web- involved only adult females, and juveniles may building spiders may reflect limited studies, either exhibit learning as they gain experience with diverse with only mature animals orwith prey that are small prey types. relative to the spider’s size. Here, I report juvenile Casual observations of capture by small juvenile females oftwo Nephilaspeciesattackingwiththrown Nephila clavipes (Linnaeus 1767) (Araneae, Nephili- silk. Importantly, the thrown silk attack was only dae) indicated that prior experience can influence employedbynaivejuvenilesexposed toanovelinsect how a spider attacks a potentially dangerous prey {Trigonastinglessbees); spiderswith priorexperience item (Higgins, unpubl. data). In particular, naive with bees did not use this behavior. juveniles frequently threw silk to subdue novel large Differences in experience can have an effect on prey items, a tactic previously reported as absent avarietyofspiderbehaviors, fromspatialorientation from the prey-capture repertoire ofadult N. clavipes (salticids, Hoefler & Jakob 2006) to mate choice (Robinson & Mirick 1971; Robinson 1975; Eberhard (Hebets 2003). Prior experience alters predatory 1982). To determine the frequency of this tactic in behavior in some non-web building spiders (i.e., juveniles interacting with novel prey items, I con- Salticidae, Edwards & Jackson 1994; Skow & Jakob ductedanexperimentwithlaboratory-rearedjuvenile 2005; and Thomisidae, Morse 2000). Although web- N. clavipes in Panama and casual observations of building spiders are often assumed to have less field-collected V. pilipes (Fabricius 1793) in Papua 530 HIGGINS— ATTACK STRATEGIES 531 — Figure 1. Behavioral sequences of31 juvenile N. clavipes attacking a Trigona bee for the first time. The number in each circle is the percentage of individuals using that behavior (the individuals that repeated “advance, retreat” are only counted once each). The number on each arrow is the number of times that particular transition was observed. New Guinea (previously N. maculata, Kuntner responded. The experienced spiders readily attacked 2005b). Voucher specimens for both species were vibrated dead bees, therefore I interpreted refusal of placed at the National Museum ofNatural History, adeadbee by anaive spideras rejection oftheinsect Smithsonian Institution, Washington, DC, USA. because itwas a bee ratherthan because itwas dead. In December 1985, I collected gravid female N. I recorded each individual’s response to the bee until clavipes on Barro Colorado Island, Panama (9°N, the prey was either definitely rejected (after ap- 80°W) and housed them in an open-air insectary in proachingthebeeandretreating, thespiderremained the laboratory clearing to collect egg sacs. After the at the hub > 5 min) or captured and removed to the emergence ofyoung and their dispersal to individual hub oftheweb. Chilled bees and dead, vibrated bees orb-webs, I placed groups of spiderlings in screen are grouped in these analyses due to small sample cages in the insectary, and fed them ad libitum with sizes. fruit flies collected at banana baits. After molting to Robinson & Mirick (1971) provide descriptions of the fourth instar, about0.3 cm leg I tibia+ patella, I adult Nephila attack behavior, which I here modify placed haphazardly-selected individual juveniles on to reflect observed juvenile behaviors. A “bite 23-25 cmdiametersphericalframes(twointersecting attack” is an approach followed immediately by circles of 0.25 cm fiberglass strips) hung in the adirect bitethatis sustained. “Advance and retreat” insectary, upon which they spun orb and barrier is an approach followed by retreating to the hub of webs. Juveniles were offered 3 fruit flies daily (total, the web. During a “thrown-silk attack,” the spider about 2.7 mg wet weight) until the next molt, after approached the bee, turned around, and placed silk which they measured approximately 0.5 cm leg I over the insect using the fourth pair oflegs. tibia + patella length (TPL). Thejuvenilespidersrespondedvariablytothebees To determine the prey-capture strategies of these (Fig. 1). Thesamplesizewasnot sufficientto testfor juveniles, I offered each of thirty-one fifth-instar small effects of spider weight or days since molting spiders a chilled stingless bee {Trigona sp., 3.67 mg (age within instar). Of the 27 successful attacks on mean wet weight, approximately 20% of these Trigonabees, 15 spidersusedthetypicalNephilalong spiders’ body mass and equal to their TPL). These bite attack (one after advance and retreat), followed chilled bees rapidly recovered enough to move, but by pulling the bee from the orb, wrapping it and not to fly, within the time frame of these observa- returning it to the hub suspended from a silk thread. tions. Occasionally, I over-cooled bees: if dead, I Four spiders immediately attacked the bee using vibrated the bee with a tuning-fork until the spider thrown silk followed by a bite, a “wrap-bite” 532 THE JOURNAL OF ARACHNOLOGY couplet. Many spiders (13) advanced and retreated the attack (often following short bites or cut-silk from the bee. These approaches involved touching attacks). All of the successful attacks included the bee, and may have involved attempted bites. thrown silk at some point ofthe sequence, and 6 of | Threeofthesespiders, afteroneormoreapproaches, the successful attacks included cutting the orb. In all , never attacked; one spider approached the bee three cases, attacks on bees after a successful attack were | times without attacking. On one occasion the bee greatly simplified and usually involved just biting, | escaped after the spider had approached and pulling, and returning to the hub (with or without ’j retreated. Ofthe successful attacks following explor- wrapping after removing the bee from the orb). atory approaches, one spider used a bite attack and The exploratory approaches and the thrown-silk eight spiders threw silk. Usually, no clear bite was and cut-silk attacks are distinct from adult Nephila observed until after the spider had removed the attack behaviors (Robinson & Mirick 1971). “Bite- wrapped bee to the hub. and-back-off' attacks by adults involve retreating I only observed the exploratory approach and only short distances from the prey, whereas these subsequentuse ofthrown silkinnaive spiders. When juveniles retreated to the hub of the orb (16-25 cm 1 repeated the offering ofbees to thesejuveniles, all radius orb-webs), and often the bee was healthy switched tactics to the long bite attack by the second enough to escape. In addition, the observed thrown- or, rarely, the third feeding. This behavioral flexibil- silk attack behavior differs from in situ wrapping ity is distinct from the behavioral sequence plasticity reported for adult females. Robinson & Mirick reportedforNephilengysbyJapyassu&Viera(2002), (1971) describe in situ wrapping by adult female N. where mature females altered the attack sequence clavipes as occurring only after the spider has bitten during attacks, but were not sequentially tested for and attempted to pull the prey free of the web; in shifts in attack strategy. fact, they experimentally elicited this behavior by A thrown-silk attack was seen once in the field in restraining prey when the spiders attempted to pull Panama. In this case, a mature female N. clavipes it free. The juveniles I observed were capable of (1.2 cm leg I tibia + patella length, approximately pulling the bee free and wrapping; in N. clavipes, 2 cm body length) wrapped but did not bite an all direct bite attacks were followed by this behav- approximately 10 mmhemipteranbug. Overaperiod ioral sequence. None of the juveniles that threw of1.5 h, sherepeatedlyplacedsilkoverthebugusing silk attempted to pull the bee free before using this leg IV, capturing it between the thrown silk and the tactic. orb-web mesh, and the bugcontinued strugglingand The thrown-silk attack by Nephila greatly resem- freeingitself. The spiderwasnot observed to bite the bles the wrap-attack of araneoid spiders. After bug, and it eventually escaped. approaching, the juvenilesturned awayfrom AttheChristensen Research Institute(nowclosed) the prey and threw silk using legs IV. Unlike wrap- in Madang, Papua New Guinea (5°13'S, 145°48'E), attacks, the bees did not rotate but became I recorded attack strategies for sixteen fourth- sandwiched between the silk mat thrown by the instar juvenile N. pilipes in an insectary during spiderandthefinemeshoftheorb. Thespiderwould an experiment on developmental plasticity (Higgins then cut the bee free ofthe orb, attach a silk thread, 1995). These spiders are the same size as the N. and carry it to the hub hanging from this thread. clavipesjuveniles used in the above-described exper- Biting followed the return to the hub. The main iment. All of these individuals were assigned to difference between the Nephila thrown-silk attack receive either one or two wild-caught Trigona bees andthearaneoidattackwrapisthefailureofNephila daily. I recorded the attack sequence used by these juveniles to rotate the prey “rotisserie-style” around juveniles for the first two bees that were provided. a radius or several viscid spiral strands between two The attack sequences were highly variable, involving radii (Eberhard 1967). This difference may reflect six different responses: advance and retreat, touch, physical constraints imposed by the extremely fine longbite, shortbite,wrapattack, andcut-silkattack, mesh ofthejuvenile Nephila orb web. With a body andwere oftenverylongwith repeatedcouplets such lengthof0.4—0.6 cm, Trigonabeesarelargerthanthe as “wrap and bite,” “touch and bite,” and “touch 0.1 cm orb mesh of Nephila juveniles (Higgins & and wrap.” Only 9 (56.25%) ofthe first attacks were Buskirk 1992). successful; two individuals threw the bee away after It is widely reported that Nephila spiders lack the subduing it, and five allowed the bee to escape. One “attack-wrap” or “wrap and bite” behavior that individualdid notsuccessfullycapture a beeuntil the other orb-web spiders use to deal with large or fourth attempt. Nephila pilipes juveniles exhibited potentially dangerous prey (Robinson & Mirick onenovelattackmodenotseeninN. clavipes. Eleven 1971; Robinson et al. 1971; Robinson 1975; Eber- (68.75%) of the spiders at some point during their hard 1982; Kuntner 2005b, 2006). Most recent attack cut the orb above or beside the bee so that it studieshavefocused on adultfemales, to standardize collapsed over the insect; two individuals cut the silk the assessment of this character for phylogenetic asan initialresponse.Ten(62.5%)ofthespidersused analyses (Kuntner pers. com.). The use ofa thrown- thrown silk to attack the prey at some point during silk attack by juvenile N. clavipes and N. pilipes is HIGGINS—NEPHILA ATTACK STRATEGIES 533 correlated with each individuafs experience with Herberstein, M.E., K.E. Abernethy, K. Backhouse, these relatively large insects. Trigona bees are H. Bradford, F.E. de Crespigny, P.R. Luckock & potentially dangerous prey for small N. davipes M.A. Elgar. 1998. Theeffect offeedinghistory on and I have observed them killjuveniles in Panama. prey capture behaviour in the orb-web spider The observation ofan adult female N. davipes using ArgiopekeyserlingiiKarsch(Araneae: Araneidae). this attack against a bug, perhaps chemically Ethology 104:565-571. defended, indicates that the behavior is not lostwith Higgins, L.E. 1987. Timebudget andprey ofNephiia maturation. It is possible that the rarity of the davipes(Linnaeus) (Araneae: Araneidae) insouth- thrown-silk attack is due to the large size ofmature ern Texas. Journal ofArachnology 15:401M17. females and the relatively small size ofthe common Higgins, L.E. 1995. Direct evidence for trade-offs prey items (Rypstra 1981; Nentwig 1985; Higgins between foraging and growth in ajuvenile spider. 1987; Higgins & Buskirk 1991). Future studies Journal ofArachnology 23:37M3. should address why this tactic is so rarely employed Higgins, L.E. & R. Buskirk. 1992. A trap-building andwhetherthe observed differenceswith theclassic predator exhibits different tactics for different wrap-attack behavior simply reflect a physical con- aspects of foraging behavior. Animal Behaviour straint due to the fine structure of the Nephiia orb 44:485-499. web. Hoefler, C.D. & E.M. Jakob. 2006. Jumping spiders I thank the Smithsonian Tropical Research In- in space: movement patterns, nest site fidelity and stitute (Barro Colorado Island, Panama) and the theuseofbeacons. AnimalBehaviour71:109-116. ChristensenResearchInstitute(Madang, PapuaNew Japyassu, H.F. &C. Viera. 2002. Predatoryplasticity Guinea) for logistical support, arrangement of in Nephiiengys cruentata (Araneae: Tetragnathi- permits anduse offacilities duringthisproject. Field dae): relevance for phylogeny reconstruction. workwassupportedbyNSF(BXR-8413831)andthe Behaviour 139:529-544. Christenson Research Fund. Funding during manu- Kuntner, M. 2005a, ArevisionofHerennia(Araneae: script preparation was provided by NSF (INT- NepMlidae: Nephilinae), the Australasian 'coin 02,33440). Elizabeth Jakob and Matjaz Kuntner spiders.' Invertebrate Systematics 19:391M36. encouraged revision of this long-neglected manu- Kuntner, M. 2005b. Systematics and Evolution of script, and Douglass Morse made several helpful Nephilid Spiders (Araneae, Nephilidae New comments upon the manuscript. Comments from Rank). Doctoral thesis, George Washington twoanonymousreviewershelpedtoclarifytheprose. University, Washington, DC. Kuntner, M. 2006. Phylogenetic systematics of the LITERATURE CITED Gondwanan nephilid spider lineage Clitaetrinae Agnarsson, 1. 2004. Morphological phylogeny of (Araneae, Nephilidae). Zoologica Scripta 35: cobweb spiders and their relatives (Araneae, 19-62. Araneoidea, Theridiidae). Zoological Journal of Kuntner, M. 2007. AmonographofNephiiengys, the the Linnean Society 141:447-626. pantropical ‘hermitspiders' (Araneae, Nephilidae, Edwards, G.B. & R.R. Jackson. 1994. The role of Nephilinae). Systematic Entomology 32:95-135. experience in the development of predatory Morse, D.H. 2000. The effect of experience on the behavior in Phidippus regius, a jumping spider hunting success of newly emerged spiderlings. (Araneae, Salticidae) from Florida. New Zealand Animal Behaviour 60:827-835. Journal ofZoology 21:269-277. Nakata, K. &A, Ushimaru. 2004. Difference in web Eberhard, W.G. 1967. Attack behavior of diguetid construction behavioratnewly occupiedweb sites spidersandtheoriginofpreywrappinginspiders. between two Cydosa species. Ethology 110:397- Psyche 74:173-181. 411. Eberhard, W.G. 1982. Behavioral characters for the Nentwig, W. 1985. Prey analysis of four species of higher classification oforb-weaving spiders. Evo- tropical orb-weaving spiders (Araneae: Aranei- lution 36:1067-1095. dae) and a comparison with araneids of the Griswold, C.D., J.A. Coddington, N.L Platnick & temperate zone. Oecologia 66:580-594. R.R, Forster. 1999. Towards a phylogeny of Prokop, P. 2006. Prey type does not determine web entelegyne spiders (Araneae, Araneomorphae, En- design in two orb-weaving spiders. Zoological telegynae). The Journal ofArachnology 27:53-63. Studies 45:134-131. Hebets, E.A. 2003. Subadult experience influences Robinson, M.H. 1975. The evolution of predatory adultmatechoiceinanarthropod: exposedfemale behavior in araneid spiders. Pp. 292-312. In wolfspiders prefer males ofa familiar phenotype. Function and Evolution in Behavior. (G. Baer- Proceedings of the National Academy of Science ends, C. Beer & A. Manning, eds.). Clarendon USA 100:13390-13395. Press, Oxford, UK. Helling, A.M. & M.E. Herberstein. 1999. Theroleof Robinson, M.H. & H. Mirick. 1971. The predatory experience in web-building spiders (Araneidae). behaviorofthegolden-webspiderNephiiadavipes Animal Cognition 2:171-177. (Araneae: Araneidae). Psyche 78:123-139. 534 THE JOURNAL OF ARACHNOLOGY Robinson,M.H.&B.Robinson. 1976,Theecologyand a Peruvian population of the spider Nephila behaviorofNephilamacuiata:asupplement. Smith- davipes, Oikos 37:179-182. sonianContributionstoZoologyN. 218:1-22. Scharff, N. &J.C. Coddington. 1997. Aphylogenetic Robinson, M.H., B. Robinson & W. Graney. 1971. analysis ofthe orb-weaving spider family Aranei- The predatory behavior ofthe nocturnal orb web dae (Arachnida, Araneae). Zoological Journal of spider Eriophora fuliginea. Revista Peruana de the Linnean Society 120:355-M34. Entomologia 14:304-315. Show, C.D. & E.M. Jakob. 2005. Jumping Rodriguez, S., R.L. & E. Gamboa S. 2000. Memory spiders attend to context during learned avoid- of captured prey in three web spiders (Araneae: ance of aposematic prey. Behavioral Ecology Araneidae, Linyphiidae, Tetragnathidae). Animal 17:34-40. Cognition 3:91-97. Rypstra, A.L. 1981. The effect of kleptoparasitism Manuscript received 17 May 2006, revised8 January on prey consumption and web relocation in 2007.