Tests for attraction to prey and predator avoidance by chemical cues in spiders of the beech forest floor

2012·4.4 MB·English

by  Melissa B Wetter

2
Arachnologische Mitteilungen43:84-89 Nuremberg,July2012
Tests for attraction to prey and predator avoidance by chemical cues in spiders
ofthe beech forest floor
Melissa B.Wetter,BettinaWernisch & SorenToft
doi:10.5431/aramit4308
Abstract:Spidersleavedraglines,faecesandothersecretionsbehindwhentravelingthroughtheirmicrohabitat.
The presence ofthese secretions may unintentionally inform other animals, prey as well as predators,about a
recentand possible current predation riskorfood availability.Fora wolfspider,otherspiders including smaller
conspecifics,form a substantial part oftheir prey,and largerwolfspiders,again including conspecifics,are po-
tential predators.Wetested two hypotheses:that largewolfspiders maylocate patchesofpotential spiderprey
throughthepresenceofsilkthreadsand/orothersecretions;andthatpreyspidersmayusesecretionsfromlarge
wolfspiderstoavoid patcheswith high predation risk.Weused large(subadultoradult)Pardosasaltanstoprovide
predatorcuesandmixeddwarfspidersorsmall(juvenile)Rsaltanstoprovidepreycues.Subadultwolfspiderswere
significantlyattracted to littercontaminated bydwarfspiders orsmall conspecifics after6 hours but no longer
after 24 hours. In contrast, neither dwarf spiders nor small Rsaltans showed significant avoidance of substrate
contaminated byadultRsaltans.However,small Rsaltansshowed differentactivity patternsonthetwosubstrates.
The resultsindicatethatwolfspidersareabletoincreasetheefficiencyofforaging bysearching preferentiallyin
patcheswith the presenceofintraguild prey.The lackofa clear patch selection response ofthe prey in spiteof
a modified activity pattern maypossibly beassociated with thevertical stratification ofthe beech litterhabitat:
the reduced volume ofspaces in the deeper layers could make downward ratherthan horizontal movementa
fastand safetacticagainsta large predatorthatcannotenterthesespaces.
Keywords:Anti-predatoryresponse,Araneae,cannibalism,IGP,Lycosidae,preydetection
A
kairomone is “a chemical that is produced by one informingotherindividualsaboutthelevelofactivity
organismconveyinginformationtoanotherorganism in the patch and the identity ofprevious occupants.
ofadifferentspecies;itisadvantageoustotherecipi- In many predator-prey contexts, especially in the
entbut detrimentalto the producerofthe chemical” case ofspiders, restriction ofthe kairomone concept
(ReSH 6cCARDE2003).Manyanimals do notleave to interspecific information transfer is unfortunate.
asinglechemicalbutinsteadacomplexofsecretions, Conspecifics may be as important as prey or preda-
and recipients maybenefitbyusing several senses to tors as any heterospecific animal, and unintentional
detectthemultitudeofsimultaneousvisual,mechani- information may benefit these to the same extent.
cal or chemical cues (DiCKE 6c GROSTAL 2001) to Finally,manyexperimentalsetups (includingtheone
getmoredetailedinformationaboutthepreviousoc- used in this paper) are not designed to identify the
cupantofthepatchthancanbeobtainedfromasingle exactcue(s)towhichtheanimalsrespond.Therefore,
chemical cue. Spiders leave draglines as witnesses of inthis studywe analysetowhatextentsecretionsleft
theirlocomotoryactivity,butfaeces and othersecre- byspiders are usedbyotherspiders-whetherofthe
tions maybe released aswell. Otheranimals that are same or different species - as sources ofinformation
able to decode this information, whether potential about potential preyand predators.
preyorpredators,maybenefiteitherthroughavoiding Thereissubstantialevidencetosupporttheuseof
a predator byleaving the patch, or by concentrating silkand/orothersecretionsasinformationsourcesal-
their search in patches where prey have recently lowinganti-predatoryresponsesinspiders,especially
been active. Silklines and othersecretions thus have amongwolfspiders(DiCKE6cGROSTAL2001,PER-
“kairomonalfunction”inthesenseofunintentionally SONSetal.2001,BARNESetal.2002,PERSONSetal.
2002,Persons6cLYNAM2004).Muchlessisknown
MelissaB.WETTER,BettinaWERNISCH,S©renTOFT,Department aboutwhetherspiderscanusecuesAfromtheirpreyand
ofBioscience,AarhusUniversity,NyMunkegade116,DK-8000 thus increase foraging efficiency. few studies have
AarhusC,Denmark established that some spiders can detect cues from
Correspondence:[email protected]
otherspiderpreyspecieswhenforaging(KESSLER6c
submitted:25.11.201haccepted:20.4.2012;onlineearly:25.7.201 LAAN 1990, Persons 6cRypSTRA2000),but none
Preyandpredatoravoidancebychemicalcues 85
ofthese have includedatestofwhethercannibalistic Individuals ofPsaltanswere collectedbyhand from
spiders maysimilarlyutilize cues from individuals of abeechstandinthemixedforestLilleringSkovwest
theirown species. ofArhus,Denmark(N 56° 8’32”, E 9° 56’38”,45 m
Non-web building spiders spend a considerable a.s.l.) duringspringmonths.The cohortswere easily
amount oftime in search ofprey, and estimates of distinguishedbytheirbodysize evenwhen the large
average capturerates areusuallyquitelow(e.g.NYF- cohortwas stillnotmature (smalljuveniles: 3-4mm,
FELER &cBenz 1988). As a result, these spiders are subadults: 5-7 mm). Litter from the same area was
often limited by food in nature (WISE 1993).Their subsequently sifted for dwarf spiders. We used the
successwillbe increased iftheycan search non-ran- mixtureofdifferentspeciesobtainedbysifting.After
domlybyrecognizingpatches ofthe habitat that are the experiments they were identified as being from
rich in potential preywhile poor in the spider’s own the family Linyphiidae (including - in order ofde-
predators. The present experiments were intended creasingabundance-Micronetaviaria,Diplocephalus
to gain evidence about the use ofsilk and/or other picinus^ Tenuiphantestenebricola^ Porhommapallidum^
secretions forpreydetection and predator avoidance Bathyphantesgracilis^ Macrargus rufus, Saloca diceros,
amongspiders ofthe beechforest floor.A dominant Walckenaeriaobtusa^ Wcucullata^ and some unidenti-
predator in the habitat is the wolf spider Pardosa fiedjuveniles) andTheridiidae {Robertuslividus).All
j-^//ß?2.yTöpfer-Hofmann,2000,while an assemblage spiderswerekeptinindividualtubesandstoredunder
ofdwarfspiders aswellas smalljuveniles ofPsaltans coolandmoistconditionsuntiltheexperimentswere
are the most abundant potential intraguild prey.We set up. The spiders used for the predator avoidance
expected that large P saltans would be attracted to experiments were fed intermittently during this pe-
litter contaminated by secretions from the potential riod, and they were observed to produce webs both
intraguildprey,whiletheseshouldtendtoavoidlitter in the tubes and, subsequently, in the experimental
contaminatedbysecretions from largewolfspiders. petri dishes.
Materialsandmethods Procedure
Spiders We used the same experimental design, taken from
The wolfspider Pardosasaltans is a common species Kessler&LAAN (1990),inboththeattractionand
ofdeciduous forests in Denmark. Like otherPardosa avoidanceexperiments.Thecollectedbeechlitterwas
species, it has a mixed searching and sit-and-wait washedinboilingwatertoremoveanysilkandother
foragingbehaviour, named the“sit-and-move”tactic secretions it may have contained and laid out for
by SAMU et al. (2003); i.e. they spend most ofthe two days to drythoroughly.The cleananddrybeech
timewaitingforpreytocometothem,butfrequently litter, selected to be ofapproximately the same size
change position. Presumablytheyend up in patches and the same amount, was placed in 9 cm diameter
of high prey availability. Apart from insects, wolf petridisheswith a smallpiece ofdamp cotton inthe
spiderspreyonotherspidersincludingindividualsof centre for moisture.The spiders intended to provide
theirowngenus and specieswhich mayform alarge the cues were then placed individually in the beech
proportionoftheirprey(EDGAR1969).Cannibalism litter dishes and allowed to stay there for 48 hours.
amongwolfspidersistypicallycommittedbyalarger Controllitterwas treated the same wayexcept there
spiderpreyingonasmaller,wherethelatterusuallyis was no spider in the petri dishes. In the attraction-
onlyhalforless ofthe mass ofthe former (SAMU et to-prey series ofexperiments dwarfspiders or small
al. 1998). Forthe design ofthe experimentswe took Psaltanswereusedforpre-treatmentofthelitter,and
advantage ofthe fact that Psaltans in Denmark has large (subadult) Psaltans acted as test predators. In
abienniallife cycle (ToFT 1976, as Plugubris) with thepredator-avoidanceseriesofexperiments,adultP
reproductioninlatespring.Animalsofthesecondyear saltansofboth sexeswere usedforthe pre-treatment
cohort(i.e.subadultsandadultsaftertwooverwinter- against which dwarfspiders or small P saltans were
ingsasjuveniles)couldthereforebeusedaspredators/ tested.
cannibals, and smalljuveniles ofthe firstyearcohort Immediatelyafterthecue-providingspiderswere
(i.e.afteroneoverwintering) astheirconspecificprey. removed from the petri dishes, one dish-full ofcon-
Other litter dwelling dwarf spiders (linyphiids and taminatedlitterwastransferredtoonesideofalarger
theridiids) found in the same habitat were used as petri dish (14 cm diameter). This was done with a
intraguild prey. pairofforcepstopreventhumancontamination.The
86 M.B.Wetter,B.Wernisch&S.Toft
pileoflitterfromthe smallpetridishwastransferred different responses to adult andjuvenile contamina-
as a whole in order to prevent breakage ofthe silk tion, since smalljuveniles ofthe same size as the test
lines. A similar amount ofcontrol litter was placed spiders would not pose a similar threat as the larger
in the other side ofthe large petri dish, transferred conspecifics. In all experiments observations ofspi-
with forceps and as awhole pile aswell.Therewas a ders that could notunequivocallybe assigned to one
piece ofdamp cotton in the centre for moisture.We response,e.g. ifthe spiderwas in the spaces between
controlledforsidebiasesbyfacingthe contaminated littertypes orfilterpapers,were ignored.The results
side ofhalfthe petri dishes towards one side ofthe onpatchchoicewereanalysedwiththebinomialone-
&
room and the other halftowards the other side. In tailed tests (Siegel Castellan 1988) and the
the attraction experiments a series ofcontrol dishes duration ofbehaviours bypaired /-tests (JMP 8.0).
with cleanlitterinboth sideswere included in order
to furtherrule out anyside bias due to unrecognised Results
external factors. The test spider was then placed in Attraction topreycues
an inverted glass tube in the centre ofthe petri dish Spiders ofthe control group showed no preference
between the two groups oflitter. It was given one for a particular side ofthe petri dish at any ofthe
minute to settle there before removing the tube and two checktimes (6 or24hours) (Tab. 1).Thus there
allowingthespidertomoveaboutfreely.Theposition was no sidebias inherentinthe experimentalset-up.
ofthelargewolfspiderswas recordedasbeingeither After 6 hours, subadult P saltans were significantly
in the clean litter or in the contaminated litter after positionedinthelittercontaminatedbydwarfspiders
6 hours and again after24 hours. as well as in litter contaminated by small P saltans
Thepredator-avoidanceexperimentsdidnothave (Tab. 1). This pattern persisted at the 24 hours in-
separate double-control replicates. They were run spection butitwas no longer statisticallysignificant,
bothwithlitterasdescribedabove andrepeatedwith indicating a weakening effect ofthe cue that earlier
filterpaperinsteadoflitterinthedishes.Inthelatter attracted the test spiders.
seriesweaddedaninspectionafter2hourstoaccount
forthepossibilitythatpredatoravoidancemightbea Avoidanceofpredatorcues
faster, but less enduring, response than attraction to There was no indication that linyphiids or small P
prey. All experimental series were replicated 20-30 saltans juveniles avoided settling in litter contami-
times. Each spider specimen was used only once. A natedbyapotentialspiderpredator(adultPsaltans)^
furthersupplementaryexperimentwasalsoperformed whether the predator was a female or a male (Tab.
using filter paper as the substrate and with one half 2). Repeatingthe same experimentusingfilterpaper
contaminated by an adult female, an adult male or instead oflitter did not change this result. However,
a small juvenile P. saltans and using small juvenile theactivityofsmalljuvenilePsaltansdifferedbetween
Psaltans as test animals (n = 12 in each group). For contaminatedandcleanfilterpaper(Fig. 1). Overall,
45 minutes the duration of the activities “sit” and thespidersspentmuchmoretimesittingthanrunning
“run”were measuredwith stop-watches forboth the (paired/-test, t^^ = 19.3,P< 0.0001).The time spent
contaminatedandthecleanfilterpaper.Weexpected runningwassignificantlyloweroncontaminatedthan
on clean substrate = 3.49, P
(/^^
Tab.1:PositionsofsubadultPardosasaltansatprescribed inspectiontimesin = 0.0013). The same was true
petridishesinwhichonesidehadclean beech litterandtheotherhad litter for sitting, though this was not
contaminated byexposuretopotential intraguild(dwarfspiders)orconspecific
prey(inthecontrolexperimentbothsideshadclean litter). statisticallysignificant(/^^= 1.84,
P=0.075).Itmadenodifference
No.in
Preycontaminant Time of contaminated No.in p* whethercontaminationwasbya
inspection cleanlitter female, a male or ajuvenile (all
litter
P>
6h 9 11 0.748 0.05).
Control (cleanlitter)
24h 10 10 0.588
6 h 16 4 0.006
Dwarfspiders
24h 13 7 0.132
SmallPsaltans]UYtm\es 6 h 15 5 0.021
24h 13 7 0.132
*One-tailed binomialtest
Preyandpredatoravoidancebychemicalcues 87
Tab.2:Positionsoftestspidersatprescribed inspectiontimesin petridishesinwhichoneside
hadclean beech litterandtheotherhad littercontaminatedbyexposuretopotentialintraguild
(dwarfspiders)orconspecificpredator.
Predator Testspider Timeof #in #in F
contaminant (substrate) inspection contam.litter controllitter
P.saltansad. 9 Dwarfspiders 6h 13 16 0.36
(litter) 24h 16 12 0.29
P.saltansad. d 6h 16 13 0.36
24h 15 13 0.43
Psaltansad. 9 Psaltansjuvs 6 h 10 11 0.50
(litter) 24h 9 12 0.20
P.saltansad. 6 6 h 11 12 0.66
24h 11 11 0.58
Psaltansad. 9 Dwarfspiders 2h 16 13 0.36
(filterpaper) 6h 15 15 0.57
24h 13 16 0.36
Psaltansad. d 2h 12 16 0.29
6 h 13 15 0.42
24h 12 15 0.29
*One-tailed binomialtest
Discussion same-sized conspecifics canbe frequentifthe oppo-
Using the same experimental design (though with nents differin hungerlevel (PETERSEN et al. 2010).
pine needles instead ofbeech litter), KeSSLER &c The contaminating spiders left silklines in the litter
LAAN (1990) obtainedsignificantresponsesindicat- and possiblyother substances. BARNES et al. (2002)
ing attraction of adult females to other conspecific showedthatspiderscandetectthedifferencebetween
females, and ofadult males to adult conspecific fe- old and new chemical cues. This was the case in a
males. They also found one linyphiid species that predator avoidance situationwhere the cues were of
avoided litter contaminated by a potential predator the same type as in our study, i.e. silk and/or other
(P. lugubris!saltans in their case too). However, they secretionsfrom anotherspider.Bothmechanicaland
did not see a significant attraction by wolf spiders chemical cues maybe involved in their detection. In
to linyphiid-contaminated litter. The present study ourexperimentsgreatcarewas takento keep the silk
showsthatwolfspidersmaychoosetosettleinapatch lines intactduringmanipulations.Theweakeningof
which has recently been visited by potential spider theresponseafter24hoursmaythereforesuggestthat
prey,irrespectiveofwhetherthispreyisofadifferent itismorevolatileoreasilydegradablekairomonesas-
familyoraconspecific.Incontrast,wefailedtodocu- sociatedwiththesilkorleftonthe substrate,andnot
mentdirectavoidanceofahabitatpatchinfestedwith the silklines as such, that informed the spider about
predatorcues.However,measurementsoftheactivity thepresenceofpotentialprey.Innaturetheremaybe
ofsmalljuvenile P.saltanswhen released in the petri bothmechanicalandchemicaldegradationofthesilk,
dishes indicated a differential response to clean vs. though also here chemical degradation ofassociated
contaminated filter paper. The fact that they spent chemicals islikelytobe fasterthanphysicaldestruc-
more time sitting on the clean paper and less time tion ofsilk lines. An alternative explanation for the
runningonthecontaminatedpaperisconsistentwith fadingresponseistheaccumulationofthetestspider’s
anavoidanceresponse,i.e.gettingawayquicklyfrom ownsilkandsecretions duringthe testperiod.These
acontaminated area and stayingin a clean area. Our would have made it increasingly more difficult to
expectation ofa weaker response to contamination recognizethepreyspider’scuesasthetestprogressed.
from juveniles of the same size as the test spiders Although this experimentproduced evidence thatP
compared with contamination from adults was not saltansareabletodistinguishbetweenpatcheswhich
fulfilled. This may be because cannibalism between are, or are not, inhabitedbyintraguild orconspecific
)
88 M.B.Wetter,B.Wernisch&S.Toft
prey, it is still unknown if they
can distinguish between patches
with different densities ofthese
prey,andtowhatextenttheycan
detectpatchesinthehabitatwith
high densities of their various
insectprey Itis also unknown to
what extent these abilities help
thespidersimprovetheirforaging
efficiencycomparedwithrandom
unguided searches.
In summary, Pardosa saltans
may use prey cues (silk and/or
other secretions) to locate areas
withintraguildand/orconspecific
prey,buttheyseemnottodirectly
avoid areas with predator cues. Sit Run Sit Run Sit Run
This result is surprising because, Females Males Juveniles
apriori,itwouldbeexpectedthat
Contaminants
selectionforavoidingpredatorsis
stronger than selection to obtain Fig.l:Duration (mean±SE)oftheactivities"sit"and"run"bysmalljuvenileRsaltans
a meal. A possible biological duringthefirst45 minutesafterreleaseintoa petridishwithfilterpaper,one
explanation maybe found in the ohatlhferofswechriecthiownassfrcloemanei(tchoenrtraodlu)ltthfeemoatlheesr,haadlufltcomnatlaesm,inoartsemdalwlijtuhvesinliklaensdo/foRr
verticalstratificationofdeciduous saltans.Pairwisecomparisonsshowedsignificantdifferencebetweencleanand
forestlitterhabitats.Toplitterhas contaminatedfor"run"inthejuvenilecontaminantgroup(P=0.0040;indicated
large spaces, but these become by**)andfor"run"intheadultfemaleand malecontaminantgroupscombined
(P=0.029).
smallerindeeperlayerswherethe
old litter is gradually transformed into amorphous References
humus or mull, and the size of spiders inhabiting BarnesM.C.,M.H.Persons&A.L.Rypstra(2002):
these layers decreases accordingly. WAGNER et al. Theeffectofchemicalcueageonantipredatorbehavior
(2003) found the deepest stratum ofdeciduous for- in the wolfspider Pardosa mihina (Araneae: Lycosi-
est litter dominated bytiny dictynids and linyphiids dae). -Journal ofInsect Behavior 15: 269-281 - doi:
10.1023/A:1015493118836
and all other families were rare.Thus, small spiders &
may seek refuge in deeper layers where they are Dicke M. P. GrOSTAL (2001): Chemical detection of
natural enemies by arthropods: an ecological perspec-
vienratcicceaslsimbolevetomelnartgemrasypidbeerst.heItsiasfeerveonptpioosns,ibbleecatuhsaet tive.-Annual ReviewofEcologyand Systematics 32:
1-23-doi:10.1146/annurev.ecolsys.32.081501.113951
ahorizontal responsewould onlybringthe spiderto
EdgarW.D.(1969):Preyandpredatorsofthewolfspider
anotherpossiblydangerous patch.The possibilityof
Lycosalugubris.-JournalofZoology,London159:405-
escapingbyverticalmovementwasnotavailableinthe 411 -doi: 10.1111/j.l469-7998.1969.tb03897.x
experiments,buthasbeendocumentedbyFOLZetal. FolzH.C.,S.M.Wilder,M.H.Persons&A.L.Ryp-
(2006) in awolfspider from agricultural fields. Due stra(2006):Effectsofpredationriskonverticalhabitat
to the structure ofthis habitat, the escape response use and foraging ofPardosa mihina. — Ethology 112:
here consisted ofclimbing thevegetation. 1152-1158-doi: 10.1111/j.l439-0310.2006.01276.x
KesslerA.&F.P.J.LAAN(1990):Theroleofpheromones
Acknowledgements in the distribution ofspiders in the forest litter layer.
ST was supported by grants from the Danish Research - Bulletin de la Societe europeenne dArachnologie
Council(FNU)andfromtheCarlsbergFoundation.Weare 1: 172-177 (http://www.european-arachnology.org/
indebtedtotwoanonymousreviewersfortheircomments. proceedings/12th/Kessler.pdf
Preyandpredatoravoidancebychemicalcues 89
Nyffeler M. &G. Benz (1988): Feeding ecology and Petersen A., K.T. Nielsen, C.B. Christensen & S.
predatory importance ofwolf spiders {Pardosa spp.) Toft(2010):Theadvantageofstarving:successincan-
(Araneae, Lycosidae) in winterwheat fields. -Journal nibalistic encounters amongwolfspiders.-Behavioral
ofApplied Entomology 106: 123-134- doi: 10.1111/ Ecology21: 1112-1117-doi: 10.1093/beeco/arqll9
j.l439-0418.1988.tb00575.x ReSHV.FI.&R.T.CARDE(2003):Encyclopediaofinsects.
Persons M.H. &E.C. LYNAM (2004): Pardosa milvina AcademicPress,Amsterdam. 1266pp.
(Araneae:Lycosidae) spiderlingmovementin thepres- SAMU F, a. SZIRAnyi & B. Kiss (2003): Foraging in
ence ofconspecific and heterospecific silk and excreta. agricultural fields: local ‘sit-and-move’ strategy scales
-Journal ofArachnology32: 341-344- doi: 10.1636/ up to risk-averse habitatuse in awolfspider.-Animal
S03-18 Behaviour66:939-947-doi: 10.1006/anbe.2003.2265
Persons M.H. &A.L. RypSTRA (2000): Preference for Sv\MU F, S.Toft &B. Kiss (1999): Factors influencing
chemical cues associated with recent prey in the wolf cannibalisminthewolfspiderPardosaagrestis(Araneae,
spider Hogna helluo (Araneae: Lycosidae). - Ethology Lycosidae).-BehaviouralEcologyandSociobiology45:
106:27-35-doi: 10.1046/j.l439-0310.2000.00496.x 349-354-doi: 10.1007/s002650050570
PersonsM.H., S.E.Walker&A.L. Rypstra(2002): Siegel S. &N.J. CastellanJr. (1988): Nonparametric
Fitnesscosts andbenefitsofantipredatorbehaviorme- statistics for the behavioral sciences. 2. Ed. McGraw-
diated bychemotactile cues in the wolfspider Pardosa Hill,NewYork.399pp.
milvina (Araneae: Lycosidae). - Behavioral Ecology ToftS.(1976):Life-historiesofspidersinaDanishbeech-
13:386-392-doi: 10.1093/beheco/13.3.386 & wood.-NaturaJutlandica 19: 5-40
Persons M.H., S.E. Walker, A.L. Rypstra S.D. Wagner J.D., S. Toft & D.H. Wise (2003): Spa-
Marshall (2001): Wolfspider predator avoidance tial stratification in litter depth by forest-floor spi-
tactics and survival in the presence ofdiet-associated ders. - Journal of Arachnology 31: 28-39 - doi:
predator cues. - Animal Behaviour 61: 43-51 - doi: 10.1636/0161-8202(2003)031[0028:SSILDB]2.0.CO;2
10.1006/anbe.2000.1594
WiseD.H.(1993):SpidersinEcologicalWebs.Cambridge
UniversityPress,Cambridge. 328pp.