Differential antennal sensitivities of the generalist butterflies Papilio glaucus and P. canadensis to host plant and non-host plant extracts

2008·3.3 MB·English

by  R J Mercader

84 Journalofthe Lepidopterists’ Society
JournaloftheLepidopterists’Society
62(2),2008,84-88
DIFFERENTIALANTENNAL SENSITIVITIES OFTHE GENERALIST RUTTERFLIES PAPILIO
GLAUCUS AND CANADENSIS TO HOST PLANTAND NON-HOST PLANT EXTRACTS
P.
R. Mercader
J.
DepartmentofEntomology, MichiganStateUniversity,EastLansing, MI48824, USA;email:[email protected].
L. L. Stelinski
EntomologyandNematologyDepartment, UniversityofFlorida,CitrusResearchandEducationCenter,
700ExperimentStation Road,LakeAlfred, FL33850,USA,
AND
M. SCRIRER
J.
DepartmentofEntomology,Michigan StateUniversity,EastLansing, MI48824,USA.
ABSTRACT. Itislikelythatolfactionisusedbysomegeneralistinsectspeciesasapre-alightingcuetoamelioratethecostsofforagingfor
suitablehosts. Inwhichcase,significantlyhigherantennalsensitivitywouldbeexpectedtothevolatilesofpreferredoverlessornon-preferred
hostplants.Totestthishypothesis,antennalsensitivitywasmeasuredbyrecordingelectroantennogram(EAG)responsesfromintactantennae
ofthegeneralistsPapilioglaumsL.andP.canadensisR&J(Papilionidae)tomethanolicleafextractsofprimary,secondary,andnon-hostplants.
EAGsrecordedfromantennaeofP. glaucuswereapproximatelyfourfoldhigherthanthoseofP. canadensisinresponsetoextractsofitsmost
suitablehostplant,Liriodendrontulipifera (Magnoliaceae). Likewise, EAG responsesofP. canandensistoitspreferredhost,Populustremu-
loides(Salicaceae),weresignificantlyhigherthanthoseofPglaucus, Inaddition,P.glaucusexhibitedsignificantlyhigher(approximatelythree
fold) EAGresponsestoitspreferredhost,L. tulipifera,thantoitsless-preferredhosts,Pteleatrifoliate. Sassafrasalhidum,andLinderaben-
zoin. Theresultsfromthisstudyindicateasignificantdivergenceintheolfactorysystemoftwocloselyrelatedgeneralistbutterflyspecies,in-
cludingastrongspecializationintheolfactorysystemofP.glaucus.
Additional keywords: Electroantennogram,olfaction,opposition.decision-making,hostselection.
For insects with larvae that develop on a single host be associated with an increase in polyphagy (Levins &
plant, female ovipositional choice determines larval MacArthur 1969). In recent years this idea has been
habitat and therefore the likelihood of larval survival. updated in terms of neural limitations to include a
However, aclearcorrelationbetween adultovipositional prolonged decision-making time along with an
preference and host suitabilityfor larval growth has not increased error rate as costs of polyphagy (Bernays
been found in many systems (reviewed in Mayhew 2001; Janz2003).
1997), and ‘mistakes’ in which eggs are laid on plants Despite the common assertion that olfaction is an
toxic to the larvae are fairly common (Straatman 1962; importantsensorymodalityfororientationtohostplants
Wiklund 1975; Chew 1977; Berenbaum 1981; Larsson (Renwick & Chew 1994; Dicke 2000; Finch & Collier
& Ekbom 1995; Renwick 2002; Graves & Shapiro 2000), the importance ofolfactorycues foroviposition-
2003). Such 'mistakes’ are believed to be rare for site location in day-flying butterflies has received
phytoehemically specialized species, but generalists, relatively little attention compared with moths
such as Papilio glaucus L. and P. canadensis R & J (reviewedin Hansson 1995; Honda 1995, butsee Feeny
(Papilionidae), are known to regularly place a small et al. 1989; van Loon et al. 1992; Baur & Feeny 1995;
fraction of their eggs on hosts toxic to their larvae in Kroutov et al. 1999). In addition, the role ofolfactory
natural habitats (Brower 1958, 1959) and in controlled cues in butterfly host plant searching and acceptance
environments despite the presence of a suitable behavior has received little attention relative to visual
alternative (Scriber et al. 1991; Scriber 1993). In and/or contact cues (e.g. Rausher 1978; Stanton 1982;
contrast, specialist herbivores may fail to oviposit on Scherer & Kolb 1987; Grossmueller & Lederhouse
readily available suitable hosts; for example, Papilio 1985; Thompson & Pellmyr 1991; Honda 1995; Weiss
palamedes’ geographic range is determined by female 1997; Sehoonhovenetal. 1998).
ovipositionalpreferenceandnotthe availabilityof hosts Olfactorycues mayplayanimportantroleinlongand
suitable for larval development (Lederhouse et al. short-range searching behavior of pre-alighting
1992). One hypothesis that has been proposed to generalist butterflies increasingtheir efficiency. Baur &
explain this observation and the higher abundance of Feeny (1995) found electroantennogram (EAG)
specialist insects is that an increase in error rate should evidence for evolutionary lability in the peripheral
Volume 62, Number 2 85
olfactory system of three specialist butterflies, Papilio spieebush, Lindera benzoin (Lauraceae) were collected
polyxenes, P. machaon Hippocrates, and P. troilus. Ifthe from trees growing in Ingham Co. Michigan, U.S.A. in
peripheral olfactory system is labile, it may allow for areas known to be pesticide free. For simplicity,
adaptations in generalist species that allow for a hereafter hosts will be referred to by their common
functionally specialized behavior in areas where a names. The detailed protocol for preparing plant
primary host is abundant, while maintaining the extracts was described by Gok<je et al. (2005). In brief,
flexibility to accept alternate hosts in areas where the dried and ground plant materials (10 g samples) were
primaryhost(s) are rare ornotpresent. Ifolfactorycues treated with 100 ml of methanol for 24 h. Thereafter,
are used to reduce decision-making time in generalist the suspensions were filtered through two layers of
species, significantly higher sensitivity would be cheesecloth and the resultingextractswere stored until
expected in the peripheral nervous system to primary use in glass containers wrapped in aluminum foil in the
hosts overless preferred hosts. dark at4° C.
We tested this hypothesis for the polyphagous P. Electroantennograins (EAGs). The EAG
glaucus by comparing its antennal sensitivity by EAG apparatus and test protocols were a slight modification
recordingswith that ol its siblingspecies, P. canadensis, of those described in detail by Stelinski et al. (2003).
to extracts ofpreferred, secondary, andnon-hostplants. The odorstimuli usedwere theplant extracts described
These two sister species can readily produce fertile above, methanol as a negative control, and hexanal
hybridoffspring(e.g. Scriber 1998); P. canadensis males (Aldrich ChemicalCo., Milwaukee, WI, U.S.A., > 98 %
preferP. glaucus females (Deering& Scriber2002), and pure) dissolvedin hexane (Aldrich) as apositive control.
until recently, they were considered the same species Hexanal was used as a standard positive control given
(Hagen et al. 1991). However, despite their similarities that synthetic green leaf volatiles are known to elicit
theyexhibit significant differences in host plant use. In EAG responses in Papilio species (Bauer & Feeny
particular, tulip tree, Liriodendron tulipifera 1995). Two milligrams of each plant extract, hexanal
(Magnoliaceae), thepreferredhostofP. glaucus, istoxic solution, and methanol or hexane solvents alone (20 pL
to P. canadensis larvae, while quaking aspen, Populus total solution) were pipetted onto 1.4 x 0.5 cm strips of
trenudoides (Salicaeeae), the preferred host of P. Whatman No. 1 filterpaper. Thesewere aged for5 min
canadensis, is toxic to P. glaucus. For each species, in a fume hood to allow for solvent evaporation.
antennal sensitivity was measured by recording EAG Subsequently, strips treated with extract or volatile
responses to plant extracts of four hosts and one non- treatments were inserted into glass Pasteur pipettes.
hostofP. glaucus,whichincludedtuliptree andquaking EAG measurements were recorded as the maximum
mL
aspen. amplitude ofdepolarizationelicitedby 1 puffsofair
through EAG-cartridges directed over antennae ol live
Materials and Methods butterflypreparations. The time interval to expel 1 mL
Insect source. Butterflies used in EAG studies were of stimulus odororclean airwas ca. 120 ms (Stelinskiet
reared from eggs laid by wild-caught females on their al. 2003). Plant-extract or chemical stimuli were
naturalhostplants. P. canadensis femaleswerecollected deliveredthrough onearm ofaglass Y-tube (eacharm 2
from the first flightin the Battenkill RiverValleyareaat cm in length, base 1 cm long, and 0.5 cm diameter)
the New York/Vermont border, U.S.A. and the larvae positioned approximately 5 mm from the antenna as
were reared to pupae in the field on sleeved tree carbon-filtered and humidified air was delivered at 50
branches ofblackcherry, Primusserotina (Rosaceae). P. mL/min into the second arm and onto the preparation
glaucus females were collected in Lancaster Co. in viaTygon tubing.
southeastern Pennsylvania, U.S.A. and were also field- Male and female butterflies ofeach species and sex
reared on black cherry. After eclosion, butterflies were were 2-6 d post-eclosion when used for EAG assays.
fed a honey-water solution and stored at 4° C for a Butterflies were mounted on 5.0 cm diameter plastic
maximum of 6 days until they were tested. By using Petri dishes with a clay strip (30 x 5 mm) placed over
adults that had not encountered any hosts prior to our theirwings andthorax. EAGrecordingswereconducted
assays and were reared on the same common host we by removing the terminal tip ofthe club (< 0.5 mm) of
preventedanyinfluence dueto adult orlarval induction the antenna used for recording with fine scissors, and
of preference (reviewed in Mereader & Scriber2005). therecordingelectrodewaspositioneddirectlyoverthe
Plant extracts. Leaves of tulip tree, L. tulipifera severed end. The reference electrode was inserted into
(Magnoliaceae), quaking aspen, P. tremuloides the head near the base of the antenna. EAGs were
(Salicaeeae), hop tree, Ptelea trifoliata (Rutaeeae), performedca. 30sfollowingmountingofbutterflies and
sassafras. Sassafras albidum (Lauraceae), and terminated at most 2 min later. For each plant extract
. .
86 Journalofthe Leridopterists’ Society
tested, EAGs were recorded from 8-10 insects ofeach (y2 = 11.3, df = 2, P = 0.003) tothe extracts oftulip tree
sex and species. Plant-extract stimulations were and quaking aspen (Fig. 1). The magnitude of EAG
presentedtoindividualbutterflies in random order, and responses of P. glaucus was significantly greater to
control stimulations (filter paper impregnated with 20 extracts of tulip tree than those of P. canadensis. In
pL ofhexane or methanol) were deliveredpriorto each contrast, the magnitude of EAGs elicited by quaking
plant-extractstimulus presentation. aspen extracts was significantlyhigher for P. canadensis
Statistical analyses. Between-speeies, pairwise than P. glaucus.
comparisons of EAG responses were performed EAG within species comparisons. Within-species
separately for tulip tree and quaking aspen on female odor stimuli had a significant effect for P. glaucus (F =
responses using Mann-Whitney U tests with a 50.1, df=6,102, P< 0.0001), and P canadensis (F =
Bonferroni corrected significance level of a < 0.05. 24.67, df = 6,108, P< 0.0001). There was no significant
Within species, EAG responses for P. canadensis were sex-by-odor stimulus interaction for P. glaucus or P.
logtransformed and P. glaucus responses were square- canadensis therefore, male and female responses were
;
root transformed to normalize the distributions and combined for analysis ofpair-wise differences (Tables
I
homogenize variance. Data were analyzed as repeated Table 1. Mean EAG responses ± SE ofmale and female P.
measures analysis ofvariancewithindividualbutterflyas glaucus. Dataformales andfemaleswere combinedforanaly-
the subject, using Proc Mixed in the SAS System (SAS sis given that therewas no significant sex byodorstimulus in-
Institute 2000). The model included odor stimulus and teraction.
sexas explanatoryvariables. Pairmeanseparationswere Mean ± SE EAGantennalresponse(mV)toplantextracts
performedforP. canadensis andP. glaucus usingTukey’s
Odorsources Males N Females N P <0.05
multiple comparisons test.
Methanol 0.07 ±0.01 8 0.09 ±0.02 10 C°
Results
EAG between species comparisons. There were Hexanal 0.70 ±0.06 8 0.60 ±0.09 10 a
significant differences between EAG responses of P. TulipTree 0.85 ±0.07 8 0.87± 0.07 10 a
canadensis (y2 = 15.6, df= 2, P < 0.001) and P. glaucus QuakingAspen 0.26 ±0.07 8 0.19 ±0.04 10 b
Sassafras 0.23 ±0.05 8 0.22±0.04 10 b
•P. glaucus Spicebush 0.23 ± 0.04 8 0.30 ±0.04 10 b
1 n
BP. canadensis
IlopTree 0.25 ±0.04 8 0.28±0.07 10 b
>£0.8 1) “Significantdifferencesinantennalresponsestoodorantstimuliare
indicatedbydifferentlowercaseletters(P<0.05,Tukey’s 1ISD).
CD
if) *
§0.6 Table 2. Mean EAG responses ± SE ofmale and female P.
if> * canadensis Data for males and females were combined for
lC—D analysisgiventhattherewasnosignificantsexbyodorstimulus
ID interaction.
204 -
-
c 11 Mean ± SEEAGantennalresponse(mV)toplantextracts
03
'td Odorsources Males N Females N P< 0.05
4;0.2 -| B <»
Methanol 0.14 ±0.03 10 0.09 ±0.02 9 d“
Hexanal 0.57 ±0.07 10 0.60 ±0.09 9 a
“1 1 1 1
TT QA QA TulipTree 0.21 ±0.03 10 0.24±0.03 9 c
Leaf
QuakingAspen 0.43 ±0.07 10 0.37 ±0.05 9 a
Fig. 1. Median EAGresponsesofP. glaucus andP. canadensis Sassafras 0.30 ±0.06 10 0.46±0.09 9 abc
faenmdalqeusaktiontgheasepxetnracQtAsofPtoulpiuplutrseetrTemTul(Loiidreiso)denBdarrosnatruoluipnidfetrhae) Spicebush 0.35 ±0.05 10 0.41 ± 0.04 9 ab
(
mediansrepresenttheinter-quartileranges (P. canadensisinter- HopTree 0.28 ±0.05 10 0.29±0.06 9 be
quartilerangefortuliptreeissmallerthansizeofsquare). Pair-
wisedifferenceswereanalyzedforeachextractusingthe Mann- “Significantdifferencesinantennalresponsestoodorantstimuliare
WhitneyUtests.Valueswithinextractwithan “hadasignificant indicatedbydifferentlowercaseletters(P < 0.05,Tukey’sHSD).
differenceat Bonferronicorrecteda < 0.05.
Volume 62, Number 2 87
and2). P. glaucus exhibitedhigherEAG responses to its recorded in this study corresponded well with known
preferred host, tulip tree, than to less-preferred hosts, hostplantpreferences of both species.
hop tree, sassafras, and spicebush, and the non-host It is important to note that although greater EAG
quaking aspen (Table 1). Responses to tulip tree were responses were observed for females of P. canadensis
similar to those elicited by the hexanal positive control for its most common host quaking aspen than all other
(Table 1). Likewise, P. canadensis exhibited a hosts tested, these were not significantlydifferent than
significantlyhigher EAG response to its preferred host, those for the marginal host sassafras and non-host
quaking aspen, than to hop tree or tulip tree (Table 2). spicebush (Table 2). This lower specificity in P.
Once again, responses to the preferred host were not canadensis relative to P. glaucus is not unique to the
different from those elicited by the synthetic standard olfactory system. In ovipositional arenas that primarily
(Table 2). However, EAGs elicited by two ofthe non- testcontactchemoreception, P. canadensis femaleshave
hosts, sassafras and spicebush, were not significantly asignificantlylowerspecificitythanP. glaucus, including
different from those elicited by quaking aspen for P. a high acceptance rate for the non-host tulip tree
canadensis. Therewasnodifferencebetween responses (Scriber et al. 1991; Mercader & Scriber 2007). This
to methanol versus hexane solvents alone; hence, data lower specificity in P. canadensis is likely to be due to
are not shown forthe latternegative control. the absence of plants in the Lauraceae (e.g. sassafras
and spicebush), Magnoliaceae (e.g. tulip tree), and
Discussion
Rutaeeae (e.g. hop tree) where P. canadensis occurs,
Antennal sensitivity ofP. glaucus was approximately greatly reducing the selection pressure for higher
three-foldhigher to extracts of the preferred host, tulip specificity.
tree, than to any other extract tested (Table 1). Interestingly the divergence in antennal sensitivity
Conversely, tulip tree extract elicited aweakerantennal between P. glaucus and P. canadensis was observed in
response from P. canadensis than the others tested both males and females (Tables 1 and 2). As males do
(Table 2). Furthermore, EAGs recorded from P. not oviposit and these species do not mate on host
canadensis to extracts of this species’ preferred host plants, divergence in sensitivitytohostplantodors does
plant, quaking aspen, were greater than those from P. nothave anyclearadvantage for the males ofthese two
glaucus (Fig. 1). These results agreewith theprediction species. This similarity in the antennal sensitivities of
that peripheral sensitivity to primary hosts should be males and females in both species may reflect a
greater than to less preferred hosts in generalist developmental similarity between males and females
butterfly species if olfactory cues play a role in host (with no adaptive function in males) or serve an
finding behavior. It is notable that species-specific unknown function.
responses were recorded to preferred host plants Heightened antennal sensitivityof P. glaucus to tulip
despite the use ofextracts ofdriedleaves in the current tree relative to the other host extracts tested lends
study, which may have limited our assay to higher support to the hypothesis that olfactory cues may be
molecular weight volatiles. This suggests that these used to reduce decision-making time duringhost-plant
butterflyspecies mayuse host-plant volatiles, at least as selection in this species. Pre-alighting cues are more
short-range cues,whileforagingforsuitablehostplants, likelyinvolved in maximizingrates ofovipositionthanin
which agrees with field observations of P. glaucus host acceptance behavior (Thompson & Pellmyr 1991);
females hovering, but not landing, on non-hosts while therefore, it is likelythat olfactorycues maybe used to
searching for oviposition sites (R. }. M. personal maximize P. glaucus' rate of landing on tulip tree
observations). wherever this preferred host is present. Furthermore,
Although P. glaucus is a highly polyphagous the higher sensitivity of P. canadensis to odors of
swallowtail species, females exhibit a distinct quaking aspen relative to the other less-preferred plant
ovipositional preference for tulip tree throughout their species evaluated here adds further support to the
range (Scriber et al. 1991; Mercader & Scriber 2005), hypothesis that host-plant location may be, in part,
eveninpopulationswherethis hostplant does notoccur mediated by chemical signals in these two generalist,
(Bossart & Scriber 1995). Congruently, antennal sister butterfly species. Further laboratory and field
responses to tulip tree were approximatelythree times behavioral assayswill need to be conducted to confirm
greater than to another major host (hop tree), two this hypothesis.
secondary hosts (sassafras and spicebush), and a non-
host (quaking aspen). Although the EAG technique Acknowledgements
cannot distinguish between attractive versus deterrent WewouldliketothankWilliam Iloutzand HowardRomack
forprovidingadultbutterfliesand/orpupae. Earlierdraftsofthe
olfactory stimuli, the heightened antennal sensitivities
—
88 Journal ofthe Lepidopterists’ Society
manuscriptweresignificantlyimprovedbycommentsfrom Ce- Lederhouse, R.C„ M.RAyres, K. Nitao& M.Scriber. 1992.
J. J.
sarRodriguez-SaonaandRufusIsaacs.Thisstudywassupported Differential use of Lauraceous hosts by swallowtail butterflies,
inpartbvMAESproject#01644. Papilio troilus and P. palamedes (Papilionidae). Oikos 63:
244-252.
Literature Cited Levins, R. & R. MacArthur. 1969. Anhypothesistoexplainthein-
Baurs,isR.of&plPa.nFteoednoyr.p1e9r9c5e.ptCioonmpianrfaetmiavleeseloefcttrhorpeheysPiaopliolgiiocaslpeacniaelsy.- Mayhtceiowdp,ehnaPcg.eoJu.osf1im9n9os7ne.cotpsAh.daaOgpiytk.iovsEec7po9a:lto4tg1eyr7n-5s04:2o8f9.1h0os-t9-1p1l.antselectionbyphy-
BereCnhbeamuome,coMl.oRg.y51/968:1.26-A3n6.oviposition"mistake"byPapilioglaucus Mercsaedleecrt,ionR.i]n.&adu|ltM.tiSgcerribsewra.llo2w0t0a5i.ls;PhPeanpoitlyipoicglpaluasctuiscit(yL.o)f,hpops.t
(Papilionidae).J. Lepid. Soc.35:75. 25-57. In Ananthakrishnan,T. ,N. & D.Whitman(eds.), Insects
Bernays,E.A. 2001. Neurallimitationsinphytophagousinsects:Im- andphenotypicplasticity. SciencePublishers, Enfield.
plicationsfordietbreadthandevolutionofhostaffiliation.Ann. 2007. Diversificationofhostuseintwopolyphagousbutter-
BossaRretv.,EJ.ntLo.m&olJ..4M6.: 7S0c3r-i7b2e7r.. 1995. Maintenance ofecologically fElni.ets:omdoiflfeErxepn.ceAsppiln.ov1i2p5o:si8t9i-o1n01sp.ecificityorhostrankhierarchy?
significant genetic variation in the tiger swallowtail butterfly Rausher, M.D. 1978. Searchimageforleafshapeinabutterfly. Sci-
through differential selection and gene flow. Evolution 49: ence200: 1071-1073.
1163-1171. Renwick, A. A. 2002. The chemical world ofcrucivores: lures,
Brower. L. P. 1958. Larvalfoodplantspecificityinbutterfliesofthe treatsaJ.ndtraps. Entomol. Exp.Appl. 104:35^42.
Pa.pi1l9i5o9g.laSupceucsigartoiuopn.inLebpuitdte.rfNleiewsso1f2:th1e0P3a-p1i1l4i.oglaucus group. An.n&. RFe.v.S.ECnhteomwo.l.13999:4.377O—v4i0p0o.sition behavior in Lepidoptera.
II. Ecologicalrelationshipsandinterspecificsexualbehavior.Evo- SAS Institute.2000. SAS/STATUser'sGuide,version6,4thed.,vol.
lution 13:212-228. 1. SAS Institute,Cary, NC.
Chew,F. S. 1977. Coevolutionofpieridbutterfliesandtheircrucif- Scherer,C. &G. Kolb. 1987. Behavioralexperimentsonthevisual
erousfoodplants. II. Thedistributionofeggsonpotential food- processingoicolorstimuliinPierisbrassicaeL.(Lepidoptera).
plants. Evolution31:568-579. Comp. Physiol. A 160:645-656. J.
Deereirnogsp,ecMi.ficD.ma&tiJ.ngM.prSecfreirbeenrc.e a20s0y2m.meFtireyldbbeitowaesesanyshysbhroidwizhient-g Schoboinolhoogvy:enf,roLm. Mp.h,ysTi.oJleogrymyto&eJv.oAl.utvioann.LoNoenw. Y1o9r9k8:. CIhnsaepcmt-apnlan&t
NorthAmericanPapiliobutterflyspecies(Lepidoptera:Papilion- ’
Hall.
idae).J. Ethol.20:25-33. Scriber, M. 1993. Absenceofbehavioralinductioninoviposition
Dicke, M. 2000. Chemicalecologyofhost-plantselectionbyherbiv- prefeJr.enceofPapilioglaucus (Lepidoptera, Papilionidae). Great
orous arthropods: a multitrophic perspective. Biochem. Syst. LakesEntomol.26:81-95.
FeenEyc,olP.,2E8.:6S0t1a-d6l1e7r., I. Ahman, & M. Carter. 1989. Effects of hy.br1i9d9s8.ofPIanphielriiotcaanncaedoefnsdiisagannodstPicgllaaruvcaulstr(aiLtespifodropitnetrear:spPeacpiifli-c
plantodoron ovipositionbythe blackswallowtailbutterfly, Pa- ionidae). GreatLakesEntomol.31: 113-123.
Fincp8hi0,l3iS-o.8p2&o7l.Ry.xeHn.esCo(lLleipeird.opt2e0r0a0,.PHaopislti-opnliadnate)s.elJe.ctIinosnecbtyBiensheacvt.s2: itnh.,eoBvGirpLeo.saitGtiiLeoabnkibeneskhha&vyibDor.ridSonfzioPdnaeperi:.lip1oo9sg9sl1ia.bulcReuessceiaxpn-rldoicPna.klacgaleantaoidtfeuondvsiiniapslosadicitrnioeossns
GOKgasEet,chtseAo.orfLy.c,brauLsc.eifdLe.oronSuts'ealppilpnarnsotkpsir.i&EatnetM/o.imnoEal.p.pWrEohxparpli.aotAnep.plla.n2d09i06n5:g.s9'1B-be1yh0ap2ve.istorianl- Stanpstreolenef,cetrMieo.nncLeb.sy.1CO9oe8l2ci.oalsoSegeraiirapch8hy7il:neg3b6iu0ntt-aer3pf6al8it.ecsh.yEecnovliorgoynm63e:nt8—39-fo8o5d3p.lant
Gravpaelsna,dnteS.leexDct.trraco&tpshA.y.sEinMov.liorgSoihncaa.pliErrneots.opmo2on0ls0.e3s.34o:fExl1eo4at2fi6rco-sl1l4aes3r2hm.oostthpsltaontsseloefcttheed StelilCnahssoktriiin,gstLp.oenLre.iuprh|aerRra.olsMaaicdelaalpnetarata&inodnL.aibn|.setGnhucete.oob2fl0si0uq3c.uhebPaardneadspetedantcieloeanoffriolnloltnehgre-,
GrosCsamluifeolrlneiar,buDtt.erWf.ly&faRu.naC..BLioeld.eCrohnoseursve..11109:854.13O—v4i3p3o.sitionsite 4r0e5d-b8a5n6d.edleafroller,Argyrotaeniavelutinana.J.Chem.Ecol.29:
selection -Anaidtorapidgrowth anddevelopmentindietiger Straatmaan, R. 1962. NotesoncertainLepidopteraovipositingon
swallowtailbutterfly,Papilioglaucus.Oecologia66:68-73. plantswhicharetoxictotheirlarvae. Lepid. Soc. 16:99-103.
Hagen. R. PI.. R. C. Lederhouse. J. L. Bossart & J. M. Scriber. Thompson, N. &O. Pellmyr. 1991. EJv.olutionofovipositionbe-
1991. Papilio glaucus and P. canadensis are distinct species. J. havioraJn.dhostpreferencein Lepidoptera. Ann. Rev. Entomol.
Lepid. Soc.45:245-258. 36:65-89.
Hans1s0o0n3,-1B0.27S.. 1995. Olfaction in Lepidoptera. Experientia 51: vanLtoroona,ntJ.enJ.noAg.,raWm.rHe.spFornesenstzto&pFl.anAt.vVoalnateielueswiijnk.two199s2pe.ciEelseco-f
Honda, K. 1995. Chemicalbasisofdifferentialovipositionbylepi- Pierisbutterflies.Entomol.Exp.Appl.62:253-260.
dopterousinsects.Arch. Insect. Biochem.30: 1-23. Weiss, M. R. 1997. Innate colour preferences and flexible colour
Janz, N. 2003. Thecostofpolyphagy: ovipositiondecisiontimevs. learninginthepipevineswallowtail.Anim.Behav.53: 1043-1052.
errorrateinabutterfly.Oikos 100:493^196. Wiklund, C. 1975. The evolutionary relationship between adult
Kroutov,V„ M. S. Mayer&T. C. Emmel. 1999. Olfactorycondi- oviposition preferences and larval host plant range in Papilio
tioning of the butterfly Agraulis vanillae (L.) (Lepidoptera, machaon L.Oecologia18: 185-197.
Nymphalidae)tofloralbutnothost-plantodors. Insect Behav.
J.
12:833-843.
Larsisnosne,ctSs.—&CoB.nfEuksbioonm.or1a99s5t.epOtvoiwpaorsdistiaonnemiwsthaosktesgiiannth.erObiikvoosro7u2s: DecReemcbeievre2d0f0o7r.publication 3January2007; revisedandaccepted5
155-160.