SPECULATIVEHUNTING BY ANARANEOPHAGICSAL TICID SPIDER
by
ROBERT J.CLARK , DUANE P.HARLAND and ROBERT R.JACKSON 1,2) (Departmentof Zoology,University of Canterbury, Private Bag 4800, Christchurch, New Zealand)
(Acc.3-VII-2000)
Summary Portia mbriata ,anaraneophagic jumping spider ( Salticidae),makes undirected leaps ( er- raticleaping with no particulartarget being evident) in the presence of chemicalcues from Jacksonoidesqueenslandicus ,anothersalticid and a commonprey of P. mbriata. Whether undirectedleaping by P. mbriata functionsas hunting by speculation is investigatedexperi- mentally.Our rsthypothesis, that undirected leaps provoke movement by J.queenslandicus , wasinvestigated using living P. mbriata andthree types of luresmade from dead, dry arthro- pods (P. mbriata, J.queenslandicus and Muscadomestica ).When a living P. mbriata made undirectedleaps or aspring-drivendevice made the lures suddenly move up and down, sim- ulatingundirected leaping, J.queenslandicus respondedby wavingits palps and starting to walk.There was no statisticalevidence that the species from which the lure was made in u- enced J.queenslandicus ’responsein these tests. Our second hypothesis, that J.queenslandi- cus revealsits location to P. mbriata bymoving, was investigated by recording P. mbriata’s reaction to J.queenslandicus when J.queenslandicus reactedto luressimulating undirected leaping.In these tests, P. mbriata respondedby turning toward J.queenslandicus and waving its palps.
Keywords: Portia mbriata , Jacksonoidesqueenslandicus ,jumpingspiders, predation, spec- ulativehunting.
1) Correspondingauthor; e-mail address: [email protected] 2) WethankPhil T aylorand David Blest for useful discussion and valuable comments on the manuscript.Financial support was provided by theNational Science Foundation ( GrantBNS 861078)and the Marsden Fund of New Zealand(Grant UOC512). c KoninklijkeBrill NV ,Leiden,2000 Behaviour137, 1601-1612 ® 1602 CLARK,HARLAND&JACKSON
Introduction Ageneralproblem facing predators is howto locate prey(Curio, 1976). Whenvision is relied on,being out of the predator’sline ofsight orbeing camouaged will interfere with detection (Edmunds,1974). ‘ Huntingby speculation’(directing attacks at refugeswhere prey tend to be found or probingareas in whichprey normally hide) is apotential solution(Curio, 1976).Envisaged not as prey-capturebehaviour, but instead as atactic forlocating prey,hunting by speculation mightfunction for a predator byprovoking a responsethat reveals the prey’slocation. W oodstorks,for example,may probe submerged vegetation even when no prey is visible (Kahl& Peacock,1963), Octopuscyanea Grayattacks holes in coral even inthe absenceof prey(Y arnell, 1969)and lions mayrun to the topof hills, apparentlyin anticipation ofstartling unwaryprey on the otherside (Schaller, 1972).However, examples fromarthropods, and experimental studies onany predator,have been scarce. Inthe present paper,we investigate huntingby speculation in Portia m- briata (Doleschall) fromQueensland, Australia, anaraneophagic jumping spider (Salticidae) that preysespecially oftenon other salticids (Jackson& Blest, 1982). Jacksonoidesqueenslandicus Wanless (Salticidae) is especially abundantin the same habitat as P. mbriata (Jackson,1988 )andis probably the salticid species onwhich P. mbriata most oftenpreys. Preliminary stud- ies showthat chemical cues from J.queenslandicus ,evenin the absence of J.queenslandicus , prepare P. mbriata forpredation by stimulating the adoptionof a special palpposture (retracted palps) characteristic ofstalk- ingsequences against salticids as preyand by heightening P. mbriata’s at- tention to visual cues from J.queenslandicus .Preliminary studies indicate that chemical cues from J.queenslandicus also elicit intermittent undirected leaping(erratic leapingwith noparticular target beingevident) by P. mbri- ata.Theabsence of anapparent target suggests that this behaviourfunctions as speculative hunting.T wohypotheses are consideredhere: (1)undirected leaps by P. mbriata stimulate J.queenslandicus to move;(2 )bymoving, J.queenslandicus gives awayits location to P. mbriata.
Methods General Standardmaintenance procedures in a controlled-environmentlaboratory (light-dark cycle, 12L:12D;lights on at0800 h) wereadopted, as detailedelsewhere (Jackson & Hallas,1986). SPECULATIVE HUNTINGBYSPIDER 1603
Adultfemales of P. mbriata and J.queenslandicus (bodylength: P. mbriata 10-11 mm; J.queenslandicus 5-6mm), from laboratory cultures, were used.
Experiment1: J. queenslandicusviewing P .mbriata
Thesetests were used to get baseline information on how J.queenslandicus reactedwhen P. mbriata madeundirected leaps. For a testchamber, we used a transparentperspex box (length ´ width ´ height: 211 ´ 144 ´ 44mm). Three regions of the box were de ned: region1 extended50 mm outfrom one end of thebox; region 2 extended50 mm outfrom theopposite end; region 3 wasthe space between the other two regions. There was a 5-mm widehole ( keptplugged with a cork)in the bottom of thebox in the centre of region 1 and anotherin the centre of region 2. Testspider 1 wasput in thebox 24-28 h beforetesting started and had free access to the threeregions during this pre-test interval. T estingbegan between 0800 and 1000 h. When test spider1 waswithin 50 mm ofoneend of the box, a partition(partition 1) wasput into place, therebyclosing this region ( region1) offfrom the rest of the box. At thesame time, another partition(partition 2) wasput into place, closing off region 2. Testspider 2 wasintroduced 10minlater into region 2 throughthe hole in the bottom of the cage. For transfer, rsttest spider2 wasenticed into a plastictube ( diameter5 mm),then the tube was positioned with oneend against the hole in thetest chamber. When gently prodded by inserting a softbrush throughthe other end of thetube, test spider 2 walkedslowly out into the test chamber. Testingbegan only if J.queenslandicus hadremained quiescent for the previous 5 sfacing region2. With J.queenslandicus quiescent,partition 2 followedby partition1 wasremoved. Thebehaviour of thetwo spiders was observed for the next 10 min. Control tests were the sameas experimental tests except that partition 2 remainedin place ( i.e.J. queenslandicus couldnot see P. mbriata duringthese tests). Beinginterested speci cally in how J.queenslandicus reactedto undirected leaping by P. mbriata,weconsider only those experimental and control tests in which: (1) P. mbriata madea singleundirected leap during the 10-min test interval and (2) J.queenslandicus remainedquiescent for the entire period prior to P. mbriata makingits undirected leap. The testwas aborted if: ( 1) P. mbriata movedduring an experimental test into region 3 before makingan undirectedleap or (2) P. mbriata madea secondundirected leap before the 10- mintest interval elapsed. No individual J.queenslandicus or P. mbriata wasused in more thanone successful test. Different individuals were used in experimentaland control tests.
Experiment2: J. queenslandicusviewing a lure
Ourobjective was to test simultaneously how J.queenslandicus reactedto a moving lure( asimulationof undirected leaps by P. mbriata) and how P. mbriata reacted to J.queenslandicus ’reactionto thelure. The test chamber ( Fig.1) wasa rectangularperspex box (length ´ width ´ height: 147 ´ 51 ´ 51mm) withtwo tubes. Its design permitted viewingof a lureby J.queenslandicus andviewing of J.queenslandicus by P. mbriata. Tube1 (internaldiameter 13 mm), made of transparent glass, tinsidethe box at one end.Except when introducing J.queenslandicus ,thehole opening to the outside was kept stoppered.Initially, a holeat the opposite end of thetube was blocked by anopaquemetal screen( partition1). Partition 1, which tintoa slitin the box, could be moved from side toside( indicatedby dotted lines in Fig.1). Tube 2 (internaldiameter 13 mm), situated on 1604 CLARK,HARLAND&JACKSON
Fig.1. Apparatus used in experiment 2. Jacksonoidesqueenslandicus (inTube 1), but not Portia mbriata (inTube 2), can view the lure. P. mbriata canview the reaction of J.queenslandicus tothelure. T opand bottom of the box are opaque perspex and the sides transparentperspex. Tubes are transparent glass. Before testing begins, the opening of each tubeinto the box is blockedby aslidingpartition, as indicated by dottedlines. During testing bothtubes are unblocked by aligning the hole in thepartition with the tube opening. The lure ison top of a woodenrod connected to anelectric motor ‘ leapgenerator’ ( notshown). thetop of thechamber, housed P. mbriata.Thedistal opening of tube 2 waskept stoppered exceptwhen introducing P. mbriata.Theproximal opening of tube 2 openedinto a wider tube( internaldiameter 20 mm) whichin turn opened into the box. An opaque metal screen (partition2) coveredthe opening between the narrow and wide tube. Partition 2 tintoa slit inthewider tube, and it couldbe movedfrom side to side( indicatedby dotted lines in Fig. 1). Thewide tube was positioned so that its centre was directly above the opening of tube 1 intothe interior of thebox. The top of thebox was opaque. This meant that the only part of thebox’ sinteriorvisible to P. mbriata duringa testwas around the opening of tube 1. At thefar end of the box, positioned in front of tube 1, there was a holein the bottom ofthe box through which a lurecould move. Each lure was made by positioning a dead J.queenslandicus , P. mbriata or house y (Muscadomestica L.)ina lifelikeposture on a corkdisk (diameter 17 mm; height22 mm). The dead arthropod was then sprayed with an aerosolplastic adhesive (Crystal Clear Lacquer, Atsco Australia Pty.) for preservation and to maskchemical traces that might have remained on thedead arthropod. Undirectedleaps were simulated using a ‘leapgenerator’ (a metalstylus moved by an electricmotor and a spring).When activated, the motor pulled the stylus down 10 mm against anelectromagnet,stretching the spring. When a switchwas pushed, the electromagnet was temporarilydisabled, letting the spring suddenly return the stylus to its original position, SPECULATIVE HUNTINGBYSPIDER 1605 afterwhich the electric motor immediately moved the stylus back against the reactivated electromagnet. Awoodenrod (80 mm longand 1 mm thick)connected the lure to theleap generator. One endof therod was glued to thecentre of the bottom of the cork disk that held the lure ( Fig.1). Theother end of the rod was glued at rightangles to the distal end of the stylus. Before testing began,the stylus was held in placeby the magnet and the lure was positioned just below the holein the cage. The lure was oriented so thatit was facing tube 1. The quiescent lure and the leapgenerator remained out of the test spiders’ view even after tube 1 wasuncovered because thebottom of thecage was opaque. Ineach partition there was a holeequal in size to theopening of the tube it blocked.Tubes wereunblocked by rstmoving partition 1 slowlyto where its hole was aligned with the openingof tube 1, providing J.queenslandicus withaccess to the interior of thebox ( ‘partition movedaway’). Next, partition 2 wasmoved away, providing P. mbriata witha clearview oftheend of tube1. Partitions were moved away only when the following criteria were met: (1) J.queenslandicus and P. mbriata wereboth quiescent; (2) neither was standing on the partition;( 3)bothwere facing the proximal opening of the tube ( i.e.,bothwere facing into theinterior of the box); ( 4)bothhad been quiescent for the previous 5 s;(5) J.queenslandicus wasnear the distal end of tube1 ( i.e.,farenough back to beout of P. mbriata’sline of sight whenpartition 2 wasmoved away). If, after partition 2 wasmoved away, both test spiders remainedquiescent for the next 5 s,testing began by pushing the switch to make the lure springupward 10 mm intothe box. The behaviour of thetwo test spiders was recorded for thefollowing 30 s. Testingwas aborted if (1) either spider failed to become quiescent while facing the specied direction within 2 hofbeingplaced into its tube or ( 2)eitherspider moved during theinterval between moving the partitions ( unblockingtubes) and making the lure leap. When testswere aborted, the same two spiders were tested on subsequent days until a successful testwas achieved or foursuccessive days of unsuccessfultesting elapsed. Theseprocedures and the design of the apparatus meant that, when a testbegan, J.queenslandicus couldsee the lure but P. mbriata couldnot see the lure nor could it see J.queenslandicus . J.queenslandicus becamevisible to P. mbriata onlyafter moving to the proximalend of tube1. Controltests were identical to testsduring which lures were made to leap except that the lure(in controls, always a dead,mounted P. mbriata)waspositioned 10 mm furtherbelow thehole in the box where it remainedbelow the opaque surface of the box when the switch waspushed. This meant that, in control tests, potential cues that might have come from sound orsubstrate vibration were still present, but visual cues from the lure were absent. During controltests, the behaviour of both spiders was recorded for 30 s starting5 safterpartition 2 wasremoved. Noindividual J.queenslandicus or P. mbriata wasused in more than one successful experimentaltest, and different individuals were used in testswith each of the three types of lures.Another set of individualswas used in the control tests. 1606 CLARK,HARLAND&JACKSON
Results
Experiment1. J.queenslandicus viewing P.mbriata
Therewere 14 experimental tests inwhich P. mbriata madean undirected leap (Table1). In veof these tests, J.queenslandicus remainedquies- cent forthe remainderof the test (‘noreaction’). P. mbriata showed no recognisablereaction to these ve J.queenslandicus .Inthe othernine tests, J.queenslandicus orientedtoward P. mbriata andbegan to wave its palps upanddown within 2safter P. mbriata leapt. Subsequently,eight ofthese J.queenslandicus beganto walkabout, but the other J.queenslandicus re- mainedin place for145 s, withpalps wavingintermittently ,thenbecame quiescent (P. mbriata orientedtoward this J.queenslandicus ).In one test, J.queenslandicus wavedits palps,then walked about after the undirected leap, but P. mbriata didnot orient towardor otherwise react to J. queens- landicus’movement.In the otherseven tests, P. mbriata orientedwhen J.queenslandicus becameactive. Inthree instances, this was after J. queens- landicus beganto walk. In the otherfour instances, it was while J. queens- landicus was wavingits palps butbefore beginning to walk.In all instances, P. mbriata retracted its palps after orientingtoward J.queenslandicus .
TABLE 1. Results fromExperiment1: Jacksonoidesqueenslandicus viewing Portia mbriata in experimental tests butnot in controls
Experimental Control N 14 9 Jacksonoidesqueenslandicus remainedquiescent and 5 9 Portia mbriata didnot orient toward Jacksonoides queenslandicus
Jacksonoidesqueenslandicus becameactive and Portia 1 0 mbriata didnot orient toward Jacksonoidesqueens- landicus
Jacksonoidesqueenslandicus becameactive and Portia 8 0 mbriata orientedtoward Jacksonoidesqueenslandicus
N :No.of tests in which Portia mbriata madean undirected leap. T estof independence (Fisher’sexact) comparing how many J.queenslandicus becameactive during experimental (9of14) andcontrol (0 of9)tests: p = 0.003. SPECULATIVE HUNTINGBYSPIDER 1607
Therewere nine control tests in which P. mbriata (notvisible to J.queenslandicus )madean undirected leap (Table 1). Jacksonoidesqueens- landicus remainedquiescent in each.Compared to whenin controlcages, J.queenslandicus was signicantly morelikely tobegin walking about and wavingpalps if aleaping P. mbriata was visible (test ofindependence, p < 0.01).
Experiment2. J.queenslandicus viewing alure Spiders (both J.queenslandicus and P. mbriata)rarely left their respective tubes duringtests. J.queenslandicus becameactive (walkedand waved its palps) in experimental tests (leapinglure ofany of the three types visible) signicantly moreoften than in control tests ( p < 0.001for each type of lure, Table2 ).How often J.queenslandicus and P. mbriata becameactive during tests didnot vary signi cantly amongthe experimental tests with different lures. Therewere 36 successful controltests. Jacksonoidesqueenslandicus and P. mbriata bothremained quiescent in 31(Table2 ).In ve, J.queenslandi- cus walkedto the distal endof tube 1. When this happened, P. mbriata retracted andwaved its palps. Thebehaviour of both J.queenslandicus and P. mbriata differeddepend- ingon whetheror not a lure was visible to J.queenslandicus .Therewere 32 successful tests usinga lure madefrom a P. mbriata (Table2). In nine, bothtest spiders remainedquiescent duringthe 30-stesting interval. Inone, P. mbriata becameactive, but J.queenslandicus remainedquiescent. This P. mbriata didnot retract its palps.In the remaining22 tests, J. queens- landicus becameactive about5 safter the lure was madeto leap,after which P. mbriata retracted andwaved its palps. Therewere 34 successful tests usinga lure madefrom a J.queenslandicus (Table2). In 14, both test spiders remainedquiescent duringthe 30-stesting interval. Therewere 20 tests in which J.queenslandicus becameactive within5 safter the lure was madeto leap,after which P. mbriata retracted andwaved its palps. Therewere 35 successful tests usinga lure madefrom a house y (Table2). In 12, both test spiders remainedquiescent duringthe 30-stesting interval. Inone, P. mbriata becameactive, but J.queenslandicus remained quiescent.There were 23 tests in which J.queenslandicus wavedits palps within5 safter the lure was madeto leap,after which P. mbriata retracted andwaved its palps. 1608 CLARK,HARLAND&JACKSON l o g s o t ) n r i u e t r e c c n l 1 1 1 a i n o b 0 0 0 p e i d c 0 0 0 s d . . . m n i h n 0 0 0 o t v a i e c l p s < < < e w e r n m n d u p p p e o l n o , , , r i e s f 2 8 6 i g f u r 5 3 9 e n . . . o a q i r 3 5 1 t p a v . s 2 1 2 o J m e e t o m c ( g ( n C . n e i s s d t n a s w n i o e e t i e e t t i r p v c i o v e l t e a d P r c t a r n a t h n i o c e a e c i f i i r h o m m i n a b r w s o c t 5 3 0 4 e r s e n m 2 2 2 p r i e b e x t s f e t a n t s a m o i a n e i i t t o B r r r f f b r n o o e e s m t . P t t f f l o a o u d N s s n e e e r a l u R l o o 1 1 1 ) e . a r 0 0 0 ) e t m r v s c 0 0 0 - n u . . . ) u n l o p 0 0 0 s c e c . i p d ) a l d h < < < n 1 a t n i e g p a p p p p l w n w , , , e s i o d 0 6 6 r d n n e ( w 3 0 9 e o n v . . . i l e s e a 1 9 i 9 i o f u r 2 1 1 r v w o q a t t p n . s = = = d s J o u m e n 2 2 c 2 t a o c o ( x x i t x h C t d d i e e l n w k b h a l i l c a s w i s i s o h e w v s u r n ( v e w c i e t i h t d o e e i n c d c i v n a o a u n i e s t n 5 2 0 3 a q e t e l c o 2 2 2 r s n s a s i s m u e n k t l a e ) e c e c f 4 e d g a e i o , m u J n b . a i 3 o q o c v , n e o 2 N o b s m w ( k a o 6 2 4 5 t l r c N 3 3 3 3 a ( o a i r t r J t s b n ) e : ) t o e e e y m s 2 c l d d d u a t a a a t c a n e . i i i n n s m m m t P d e e r u n e e e n o o r r r d a m m a i P h u u l u n i h r l l l t s a r e ( ( ( d d d n e ) e e e p s s s e s t t t s t t t p x e u s s s n n n u e x u c e e e u u u c i t t t i q h o o o E d l l l d c s . s n a a a m m m n a t t e t t t a e , , , a s 2 s d l n n n l ) i e e d d d s e e e s t t a a a o a m E n n t l f e e e m m n m o e e L i a i i o r o d d d o e i r r r e r f s r B t u e e e e u k b a n m m m q p p p q p A t c o o o o x x x y m a a . . r r r T T C E f E f J E f J J d SPECULATIVE HUNTINGBYSPIDER 1609
Discussion
Salticids candetect motionless prey(Jackson & Tarsitano, 1993),but move- ment facilitates preydetection andstimulates the salticid to beginpredatory sequencesearlier (Heil, 1936;Crane, 1949; Drees, 1952; Jackson & Tar- sitano, 1993).Our hypothesis is that undirectedleaps functionto enhance P. mbriata’s ability to locate J.queenslandicus :leaps, byattracting atten- tion,elicit palpwaving and walking by J.queenslandicus ,whichin turnpro- vides movementcues P. mbriata canuse to locate J.queenslandicus . Our ndingssupport this hypothesis. Inexperiment 1, J.queenslandicus moreoften waved its palps andwalked whenundirected leaps by P. mbriata couldbe seen.In control tests, when P. mbriata’sundirectedleaps couldnot be seen, J.queenslandicus tended to remain quiescent.Experiment 2 also demonstratedthat J.queenslandicus ’ reaction to undirectedleaps tends to attract P. mbriata’sattention. Seeinglures makesimulations ofundirectedleaps inexperiment 2 elicited comparablereactions from J.queenslandicus ,regardless ofwhetherthe lures weremade from P. mbriata, J.queenslandicus orhouse ies. These nd- ings suggest that J.queenslandicus ’responseis ageneralised investigatory behaviourprovoked by anunidenti ed object movingin the neighbourhood andnot a reaction specically to P. mbriata. P. mbriata appearsto test the environmentfor the presenceof J. queens- landicus byattempting to provoke,with undirectedleaps, aresponsefrom its not-yet-seenprey .Undirectedleaps differ fromhow octopuses and wood- storks ushout prey by attacking the prey’smicrohabitat (Yarnell, 1969; Kahl& Peacock,1973 )because P. mbriata appearsnot to focuson a target whenit leaps. P. mbriata’sundirectedleaps appearto bemore comparable to alion runningup ahill in anticipation ofstartling unwarybut not-yet-seen prey(Schaller, 1972). Curio(1976) with the term ‘speculative hunting’outlined a general mechanism forhow prey might be located. Details aboutthe role ofany particular behaviourin a predator’srepertoire, or precisely howspeculative huntingworked, were not considered. For P. mbriata,wehave details concerningthe cues that provokespeculative huntingand the mannerin whichundirected leaping is usedby P. mbriata to ndits prey. Thestimuli governingspeculative huntingby otherpredators are notwell understood,but P. mbriata’s undirectedleaping is stimulated bychemical 1610 CLARK,HARLAND&JACKSON cues froma specic prey.Theleap itself is notprey-capture behaviour, becauseundirected leaps are notseen after P. mbriata has located its prey. Instead,this tactic canbe envisaged as somethingmore akin to setting atrap. Triggeredby chemical cues,undirected leaping is still speculative because chemical cues donot guarantee the presenceof J.queenslandicus in the immediate vicinity. Undirectedleaps providethe preywith visual cues and preyin turnprovide the predatorwith visual cues.By soliciting visual cues inresponse to chemical cues, P. mbriata appearsto co-ordinate sensory modalities. Still otherfacets of P. mbriata’spredatorystrategy mayqualify as speculative hunting.Females of Euryattus sp.,another salticid onwhich P. mbriata preys,nest ina rolled-updead leaf suspendedby silk guylines fromtree trunks,boulders or the vegetation(Jackson, 1985 ).To catch Euryattus females, P. mbriata simulates the courtshipsignals usedby Euryattus males (Wilcox& Jackson,1998). Upon ndinga conspecic female’s nest, a Euryattus male goesdown the guylinesand signals by suddenlyand forcefully exinghis legs, therebymaking the leaf rockback and forth. Euryattus females react to the male’s signal bycoming out of the nest andeither mating with the male ordrivinghim away.When P. mbriata locates anest, asimilar sequenceis seen. Portia mbriata movesto aposition abovethe suspendedleaf, theneither lowers itself onits owndragline or walks downone of Euryattus’guylines.Once on the leaf, P. mbriata uses a special behaviour,‘ shuddering’,which mimics the courtshipof Euryattus males andinduces the resident to comeout (Jackson & Wilcox,1990). Interestingly, P. mbriata will shuddereven when no Euryattus is present, suggestingthat this behaviourhas aninvestigatory function. When shudders provokea responsefrom a resident inside arolled upleaf, P. mbriata continuesthe predatorysequence. When no replyis forthcoming, P. mbriata desists fromsignalling (Jackson et al., 1997). Besides preyingon salticids, P. mbriata also preyson web-buildingspi- ders fromother families. Portia mbriata enters the otherspider’ sweband, instead ofsimply stalking orchasing down the resident, makes aggressive mimicry signals (Jackson& Wilcox,1998 ).For example, by manipulating the websilk with its legs andpalps, P. mbriata maylure the resident spider to within striking distance byimitating the struggles ofan insect onthe web (Jackson& Blest, 1982).However, P. mbriata will initiate signalling even SPECULATIVE HUNTINGBYSPIDER 1611 whena resident is notvisible (Jackson,1995 ),suggesting that aggressive- mimicry signals maysometimes functionas huntingby speculation. Likeundirected leaping, initiating aggressive mimicry signals inanother spider’s web,or on asuspendedleaf of Euryattus with noresident visible, mayattract the attention ofa preythat P. mbriata has notyet seen.By re- spondingwith investigatory behaviourto P. mbriata’s aggressive-mimicry signals, the preymay be induced to giveaway its location.Undirected leap- ingsends visual cues to potential prey,whereas web signals andshuddering onaleaf sendvibratory cues to potential prey,andthis maybe the primary difference. Notonly might aggressive-mimicry signals sometimes beenvisaged as speculative hunting,but undirected leaping might be envisaged as aggressive mimicry becauseundirected leaping may simulate the kindsof visual cues that normallyelicit investigation by J.queenslandicus .Forexample, the initial cues J.queenslandicus mightnormally get froman insect (potential prey)ora conspecic individual(potential mate orrival) maynot be so different fromthe cues providedby P. mbriata’sundirectedleaping. In P. mbriata’s predatorystrategy, speculative huntingand aggressive mimicry appearto bebroadlyoverlapping topics.
References
Crane,J. (1949).Comparative biology of salticidspiders at Rancho Grande, V enezuela.Part IV:Ananalysis of display.— Zoologica34, p. 159-214. Curio,E. (1976).The ethology of predation.— SpringerV erlag,Berlin. Drees,O. (1952).Untersuchungen ü berdie angeborenen V erhaltensweisenbei Springspinnen (Salticidae).— Z.Tierpsychol.9, p.169-207. EdmundsM. (1974).Defense in animals: a surveyof antipredator defenses. — Longmans, New York. Heil,K.H. (1936).Beiträ ge zurPhysiologie und Psychologie der Springspinnen. — Z.Vergl. Physiol.23, p. 1-25. Jackson,R.R. (1985).The biology of Euryattus sp.indet., a web-buildingjumping spider (Araneae,Salticidae) from Queensland: utilization of silk, predatory behaviour, and intraspecic interactions.— J.Zool.,Lond. B 1.p. 145-173. ——(1988).The biology of Jacksonoidesqueenslandica ,ajumpingspider ( Araneae: Salticidae)from Queensland: intraspeci c interactions,web-invasion, predators, and prey.— N.Z. J.Zool.15, p. 1-37. ——(1995).Cues for web invasion and aggressive mimicry signalling in Portia. — J. Zool. Lond.236, p. 131-149. 1612 CLARK,HARLAND&JACKSON
——&Blest,A.D. (1982a).The biology of Portia mbriata ,awebbuilding jumping spider (Araneae,Salticidae) from Queensland: utilization of websand predatory versatility. — J.Zool.Lond. 196, p. 255-292. ——&Hallas,S.E.A. (1986).Comparative biology of Portiaafricana , P. albimana, P. m- briata, P. labiata, and P. schultzi,araneophagicweb-building jumping spiders ( Araneae: Salticidae):utilization of silk, predatory versatility, and intraspeci c interactions.— N.Z. J.Zool.13, p. 423-489. ——,Li,D. &Robertson,M.B. ( 1997).Cues by whichsuspended-leaf nests of Euryattus (Araneae:Salticidae) females are recognised by conspecic malesand by anaggressive mimicsalticid, Portia mbriata .—J.Zool.Lond. 243, p. 29-46. ——&Tarsitano,M.S. (1993).Responses of jumpingspiders to motionlessprey. — Bull. Brit.Arachnol. Soc. 9, p. 105-109. ——&Wilcox,R.S. (1990).Aggressive mimicry, prey-speci c predatorybehaviours and predator-recognitionin thepredator-prey interactions of Portia mbriata and Euryattus sp.,jumping spiders from Queensland. — Behav.Ecol. Sociobiol. 26, p. 111-119. ——(1998).Spider-eating spiders. — Amer.Scient. 86, p. 350-357. Kahl,M.P .&Peacock,C.J. (1963). The bill-snap re ex: a feedingmechanism in the Americanwood stork. — Nature199, p. 505-508. Schaller,G.B. (1972).The Serengeti lion. — Universityof Chicago Press, Chicago. Wilcox,R.S. &Jackson,R.R. (1998).Cognitive abilities of araneophagicjumping spiders. — In:Animal cognition in nature:the convergence of psychologyand biology in laboratory and eld( R.P.Balda,I.M. Pepperberg & A.C. Kamil,eds). Academic Press, New York, p. 411-433. Yarnell,J.L. ( 1969).Aspects of the behaviour of Octopuscyanea Gray.— Anim.Behav. 17, p. 747-754.