Responses of North American Papilio troilus and P. glaucus to potential hosts from Australia

2008·7.8 MB·English

by  J Mark Scriber

18 Journalofthe Lepidopterists’ Society
JournaloftheLepidopterists'Society
62(1),2008,18-30
RESPONSES OF NORTH AMERICAN PAPILIO TROILUS AND GLAUCUS TO POTENTIAL HOSTS
P.
FROM AUSTRALIA
Mark Scriber
J.
Dept. Entomology,MichiganStateUniversity,EastLansing, MI48824, USA;SchoolofIntegrativeBiology, UniversityofQueensland,
Brisbane,Australia4072;email:[email protected]
Michelle L. Larsen
SchoolofIntegrativeBiology,UniversityofQueensland, Brisbane,Australia4072
And
Myron P. Zalucki
SchoolofIntegrativeBiology,UniversityofQueensland, Brisbane,Australia4072
ABSTRACT. We testedtheabilitiesofneonatelarvaeofthe Lauraceae-specialist, P. troilus, andthegeneralist Easterntigerswallowtail,
Papilioglaucus(bothfrom LevyCounty,Florida)toeat,survive,andgrowonleavesof22plantspeciesfrom7familiesofancientangiosperms
inAustralia, Rutaceae, Magnoliaceae, Lauraceae, Monimiaceae, Sapotaceae,Winteraceae,andAnnonaceae. Clearly,somecommonPapilio
feedingstimulantsexistinAustralianplantspeciesofcertain,butnotall,Lauraceae.ThreeLauraceaespecies (twointroducedCinnamomum
speciesandthenativeLitsealeefeana)wereassuitableforthegeneralistP. glaucusaswasobservedforP. troilus.Whilenoabilitytofeedand
growwasdetectedfortheLauraceae-specializedP. troilusonanyoftheotherslxancientAngiospermfamilies,tiregeneralistP.glaucusdidfeed
successfullyon MagnoliaceaeandWinteraceaeaswellas Lauraceae. In addition,somelarvaeofoneP. glaucus familyattemptedfeedingon
Citrus (Rutaceae) and asmall amountoffeedingwas observedon southernsassafras (Antlierosperma moschatum; Monimiaceae), butallP.
glaucus (from4families) diedonAnnonaceaeandSapotaceae. Surprisingly,the North American Lauraceae-specialist(P. troilus)diedonall
Lauraceaespeciesbyday#12,butsomegeneralistP.glaucuslarvaesurvived. Mostofthegeneralist(P.glaucus)offspringsurvivedandgrewvery
well on all 3 species ofMagnoliaceae assayed (Magnolia virginiana, Michelia champaca, & Michelia doltsopa) and on Tasrrmnnia insipida
(Winteraceae).Theabilityoftheselarvaetofeedandgrowon T. insipida butnotT. lanceolatasuggestssignificantphytochemicaldifferences
mayexistwithintheWinteraceae.TwoMonimiaceae“sassafras”plantspecieswereunsuitabletobothNorthAmericanPapiliospeciesdespite
theirveryclosephylogeneticrelationshipwiththeLauraceae.
Additional key words: Annonaceae, detoxification, Lauraceae, Magnoliaceae, Monimiaceae, neonate survival, Papilionidae, Rutaceae,
Winteraceae,P.glaucus,P. troilus
Rutaceae-feeding is the primary pattern in 75-80 % Annonaceae).
ofthegenusPapilio (Scriber 1984a). Insection IVofthe There are shared groups of key phytochemicals
Papilionidae (Munroe 1961), Papilio Heraclides among the Rutaceae, Lauraceae, Magnoliaceae,
( )
cresphontes Cramer is constrained to Rutaceae, unable Annonaceae, Apiaceae, and Aristolochiaceae
to survive on plants of the Magnoliaceae, Lauraceae, (Berenbaum 1995; Brown et al. 1995; Nishida 1995)
Rosaceae, or Salieaceae (Scriber et ah 1991a,&b). and these can affect opposition (Dethier 1941,1954;
However, in Section III ofthePapilionidae, ancestorsof Feeny 1995) as well as larval survival and growth
the polyphagous North American P.(Pterourus)glaucus (Munroe 1961; Nitao et al. 1992; Johnson et al. 1996).
L. groupandtheirP. troilus L. sistergroup arebelieved Ourgoal herewas to examineneonate larvalsurvival on
to have been Rutaceae feeders (Hancock 1983; Scriber reported host plants of other Australian Papilionidae
et al. 1991a), with subsequent specialization on the and representative species from these chemically-
Lauraceae and Magnoliaceae, as Rutaceae became related plant families, including the Australian
scarce after the Cretaceous (Hancock 1983; Scriber Winteraceae and Monimiaceae which are ancient
1995). With theTroidinitribebelievedtohaveoriginsin angiosperms very closely related to the Lauraceae,
remnant Gondwana 65-90 mya (Brabyet al. 2005), the Magnoliaceae, and Annonaceae (Bremer et al. 2003)
phylogenetic distances and geological timing (late with presumed similarity in phytochemicals. The
Jurassic and early Cretaceous; Soltis et al. 2005) ofthe ancient Doryphora sassafras Endl.(Monimiaceae) and
evolutionarydivergence ofthese plant groups has been Tasmannia =Drimys insipida R.Br. ex DC.
( )
recentlysuggestedtobe30-50 millionyearsago (Gaunt (Winteraceae) are reported in Australia as host plants
and Miles 2002; Zakharov et al. 2004). Such for Graphium sarpedon (L.) and G. macleayanum
diversification of the roots of Papilionidae lineages in (Leach) butterflies along with the Lauraceae and
the Leptoeircini (=Graphiini) andPapilioninitribesalso Rutaceae (Braby2000).
corresponds to plate tectonics and subsequent Papilio troilus L. (spicebush swallowtail) is a
diversification of early Angiosperm families (e.g. Lauraceae-feeding specialist found across the eastern
Volume 62, Number 1 19
half of the USA which naturally feeds on sassafras. 2003). Tasmannia insipida andT. lanceolata (Poir.)A.C.
Sassafrasalbidum (Nutt.) Ness, andspicebush,Lindera Smith are ancient angiosperms in the Winteraceae,
benzoin (L.) Blume, across most of its range, and red which is also closely related to the Lauraceae and
bay, Persea borbonia (L.) Spreng., in Florida and the Monimiaceae. Both of these ancientplant families have
southeast coastal areas (Scriber 2005). Preliminary aromatic species used by Australian swallowtail
bioassays with P. troilus in North Americaconfirm that butterflies, such as Graphium macleayanum. Australia
this species is a host plant family specialist and will not seemed to be the best place to evaluate suitability of
initiate feeding on plants other than members of the ancient Angiosperm species (Bremer et al. 2003; see
Lauraceae, including all other families used by Papilio also Grimaldi & Engel 2005) since this may have been
glaucus L. (eastern tiger swallowtail; Scriber et al. the“cradle”offloweringplantevolution, includingbasal
1991b), which also occurs across the eastern USA. P. families such as the Winteraceae and Monimiaceae
glaucus is the most polyphagous of all 563 species of (both usedbyAustralianswallowtailbutterflyspecies) as
swallowtail butterflies in the world (Scriber 1984a, well as the more widespread Lauraceae, Magnoliaceae,
1995). It feeds occasionally on spicebush and sassafras Rutaceae, Annonaceae, andAristolochiaceae (Bremeret
(Lauraceae; Scriber et al. 1975), but also includes al. 2003). The phylogenetically basal angiosperm
several dozen other host plant species from 9 different families have their origins, when geological plate
families (including the Magnoliaceae, Rutaceae, driftinghadnotfullyseparatedthecontinents (Grimaldi
Oleaceae, Rosaeeae,Tiliaceae, Betulaeeae, Platanaceae, & Engel 2005), and the Papilionidae are believed to
and others; Scriber 1986, 1988). have roots concurrent with these earlyfloweringplants
Plant species for neonate larval survival and growth (Gaunt & Miles 2002; Braby et al. 2005; ef. Miller
bioassayswereselected fromlistsofrecorded hostplant 1987).
species for Papilio aegeus Donovan and Graphiurn The modernphylogenyandsystematicsofthe ancient
species in Australia (Braby 2000; Edwards et al. 2001, Angiosperm families, examined here for their relative
Scriberetal. 2006, 2007). suitability as larval host plants, have recently been
Lauraceae feeding and oviposition in P. troilus are revised basedon manyindependent molecularanalyses
apparently determined by phytochemical (Bremer et al. 2003). The phylogenetically basal
feeding/oviposition stimulants (Lederhouse et al. 1992, angiosperms, including the 4 orders, Laurales,
Carter & Feeny 1999, Carteret al. 1999, Frankfater & Magnoliales, Canellales, andPiperales are allsupported
Scriber 1999, 2003). Sassafras and spicebush are the as monophyletic, and molecular analyses put them
preferredhosts throughout mostofthebutterfly’s range. together in a group called the magnoliids, despite the
In Florida, where these plants are scarce, red bay lack of support using morphological traits alone
(Persea spp.), is used by P. troilus populations. (Bremer et al. 2003). Within this single basal group
Preliminary studies indicated that extracts of Persea (magnoliids), the Laurales includes the Hemandiaceae,
painted on leaves of Lindera depress neonate growth Lauraceae and the closely-related Monimiaceae. The
rates ofnorthern populations ofP. troilus (Nitao et al. Magnoliales includes the Magnoliaceae and
1991). It is clear that among various geographical Annonaceae. The Piperales includes the
populations of this Lauraceae specialized butterfly Aristolochiaceae and Piperaeeae. The Canellales
species, there is variation in the suitability ofdifferent includes the primitive Winteraceae with Tasmannia
plantspeciesforoviposition, larvalacceptanceandlarval (=Drimys) species reportedashosts forotherspecies of
growth (Nitaoetal. 1991; Scriberetal. 1991b; Scriber& swallowtails (Scriber 1984a; Braby 2000). All of these
Margraf2005). families have some swallowtail butterfly species
We wanted to evaluate the abilities ofthe ancestral (Papilionidae) reported as feeding on them, but 75% of
Papilionidae, North American section III, Munroe all swallowtail butterfly species feed on the Rutaceae
(1961) species P. troilus and P. glaucus larvae to (Scriber 1984a; Berenbaum 1995). The Rutaceae
consume, process and growon these ancient Australian (including Flindersia Geijera Citrus & Zieria) are
, , ,
angiosperm species including unique genera of the believed to have survived the extensive worldwide
Lauraceae that differentiated independently of the Cretaceous-Tertiary extinctions in the eastern part of
North American Lauraceae. In Australia, there exist at Gondwana (the Australian landmass) of the southern
least two species ofplants called sassafras, Dori/plwra hemisphere, alongwithotherancientangiosperms, such
sassafras Endl. and Antherosperma moschatum Poir. as theWinteraceae, some Lauraceae, and Monimiaceae
(southern sassafras). These plants are both in the (Raven &Axelrod 1974).
Monimiaceae, which is an ancient angiosperm family This study of the North American P. troilus and P.
very closely related to the Lauraceae (Bremer et al. glaucus was conducted tovalidate host use abilities (or
20 Journalofthe Lepidopterists’ Society
inabilities) of the neonate larvae in their first bites were introduced at 96 hours for continued larval
(Zalucki et al. 2002) on species of7 families ofancient feeding and growth for another 96 hours, when they
angiosperms. In order to determine the relative were weighed again. These assay methods were used
suitability ofeach host for larval consumption, growth, successfully in our previous studies ofhost use by the
and survival, we conducted controlled environment Australian P. aegeus (see Fig. 1; Seriberetal. 2007).
bioassays using neonates from eggs of different wild Numberswereassignedforeachinstar(e.g. 1, 2, 3, 4)
females. Results provided clues to the historical and molts were assigned the midpoint (e.g. 1.5 =
(phylogenetic) or potential (future) abilities of molting from first to second instar). Survivors of P.
geographically-widespread specialist and generalist glaucus andP. troilus at 12dayswere destroyedbecause
species of North American Papilio to use different of constraints imposed by the AQIS import permit
Australianplant families. (#200520165).
Materials and Methods Plant species used for bioassays. Plant species for
neonate lanal survival and growth bioassays were
Adult capture and female oviposition. Femalesof
selected from lists ofrecorded host plant species for P.
P. troilus and P. glaucus were capturedin LevyCounty, aegeus and Graphium species in Australia (Braby2000;
Florida in March and April of2006. Using methods as Edwards et al. 2001; Seriber et al. 2006, 2007). These
described by Seriber (1993), individual females were nativeAustralianplantswereobtainedas seedlings from
placed in clear plastic boxes containing red bay leaves Fairhill Native Plants (Yandina, QL), Barung Landcare
forP. troilus andseveral species ofpotential hostplants Nursery (Maleny, QL). Anthony Hiller at Mount
forP.glaucus (includingsweetbay. Magnoliavirginiana Glorious BiologicalCentre (Mt. Glorious, QL),Turners
L. (Magnoliaceae), black cherry, Prunus serotina Ehrh. Garden Center at Rochdale, near South Brisbane, and
(Rosaceae) and white ash, Fraxinus americana L. Greening Australia Nursery (near The Gap, QL), and
(Oleaceae). Eggs were collected daily, counted, and from the University ofTasmania at Hobart. Seedlings
placed in a controlled environment chamber for 1-5 were brought to the University of Queensland
days at 4-6 ° C, until express mailed to our Australian Glasshouse during mid-October, where they were
quarantine lab in the School of Life Sciences Goddard transplanted into 4-liter pots with standard sterilized
Building (Australian DEH and AQIS permits had been potting soil (half sand). Each tree seedling was then
obtained previously; also with clearance from fertilized with Flowfeed EX7 fertilizer (Grow Force
BiosecurityAustralia). Constraints of the AQIS permit Australia Ltd; N-P-K, 20.8%, 3.3% and 17.4%
and Biosecurity Australia prevented us from sending respectively). New leaves (not fully-expanded) had
adults to Australia for oviposition preference assays on developed on all plants by the time the larval feeding
nativeplants there. bioassays started in mid-March2006.
Larval bioassays and rearing. Eggs from each Some plant species’ leaves (3 species of
Papilio femmalme were kept in sterimlemclear plastic Petri Magnoliaceae, 2 species of Rutaceae) used in these
dishes (20 deep; 100 or 150 diameter) in the studies were field-collected at the Brisbane Botanical
same controlled environment chamber until they Gardens (with theassistanceofDirectorPhil Cameron).
eclosed as neonate larvae. Newly emerged neonates Camphor tree leaves were collected from the UQ
were distributed in a split-brood design across an array Campus nearby the lab. The full list ofplants tested is
ofpotential hostplant species, with 2-3 larvae per dish
given below.
(each dish containing a new leafand a mature or fully- Rutaceae (native unless noted):
expanded leafofone plant species supportedwith their Citrussinensis Osbeck (sweet orange; “Joppa”
petioles immersedin awater-soakedflorist "oasis” foam introduced.);
wrapped tightlywith aluminum foil to retain moisture) Geijera salicifolia Schott (brushwilga);
for each 96-hour period. Neonate larvae were Flindersia australis R.Br (AustralianTeak),
introduced to each species with a fine camelhairbrush Magnoliaceae (all introduced):
bygentlyplacingthem on the aluminum foil atthebase Magnolia virginiana L. (sweetbay; North
ofboth petioles (the new leaf and the mature leaf) in America);
order for the larvae to choose which leafthey crawled Michelia champaca L. (yellowmagnolia; Asia);
onto. Daily survival and growth were monitored and Michelia doltsopa Bueh.-Hum. ex D.C. (silver
recorded. The total number of fecal pellets, the cloud) an Asian species endemic to the
estimated leaf area consumed (mm2), the instar stage Himalayan region ofChinaandTibet,
(or molt), and the larvalweights (for all survivors) were Lauraceae (native unless noted):
recorded at 96 hours. Fresh, new and mature leaves Beilschmedia obtusifolia (F. Muell. ex Meis.)
Volume 62, Number ] 21
(Blushwalnut); Annonaceae (introduced):
Cinnamomumcamphora (L.) Presl (camphor Annonamuricata L. (soursop);
laurel) an introducedtree, abundant in Annona reticulata L. (custard apple),
Queensland and NSW; Winteraceae (native):
Cinnamomum oliveri (F.M. Bailey) (Oliver’s Tasmannia insipida R.Br. ex DC. (purple
sassafras); cherry);
Cinnamomum virens R.T. Baker; Tasmannia lanceolata =Drimys aromatica)
(
Cn/ptocan/a glaucescens R.Br. (jackwood); (Poir.) A. C. Smith (mountain pepper,
Cnjptocanja microneura Meisn. (Murrogun); winterberry) found in TasmaniaandVictoria
Endiandra discolor Benth. (rosewalnut); and NSW, ’
Litsea leefeana (F. Muell.) (bollywood); Sapotaceae (native):
Neolitsea dealbata (R.Br.) Merr. (bollygum), Pouteria = Planchonella australis (R.Br)
( )
Monimiaceae (native): Baehni (blackapple).
Donjphorasassafras Endl. (sassafras);
Results
Antherospermamoschatum Poir. (southern
sassafras) found in TasmaniaandVictoria (the Neonate larvae of the Lauraceae specialist, P. troilus
onlyhost forTasmanian swallowtail butterfly died on all species in all families except Lauraceae.
subspecies, G. m. moggana Couchman), While some of the species within this favored family
—
Monimiaceae specialist feeds on Magnoliaceae & Lauraceae
lie}
G. macleayanum moggana
G. macleayanum moggana
on Micheliadoltsopa on Camphortree
Rutaceae specialist Feeds on Magnoliaceae & Lauraceae
P. aegeus
on
Michelia
doltsopaw
Annonaceae specialist feeds, oviposits, & pupates on Magnoliaceae
Graphium eurypylus
on Micheliachampaca
larva
egg
Fig. 1. Feedingassayswith family-specializedAustralian Lepidopterathatalso survivedon Magnoliaceae; la) Macleay’s swal-
low-tail(Tasmaniansubspecies,G.m. moggana)isaspecialistonthe Monimiaceae,butfeedsandpupateson MagnoliaceaeandLau-
raceae(Scriberetal. 2006), lb) PapilioaegeusisaRutaceaespecialist,butfeedson3speciesofMagnoliaceaeandcamphortreein
theLauraceae(Scriberetal.2007), lc) Graphiumeurypylusisan Annonaceaespecialistthatoviposits, feeds,andpupateson Mag-
noliaceaenear Brisbane,Australia(Larsenetal. 2008).
22 Journalofthe Lepidopterists’ Society
Table 1. Meansurvival andgrowth indices ofneonate larvaeofP.glaucus families (G1-G4) andP. troilus (T1 &T2) rearedonvarious
ancientAngiospermplantspeciesat4days,8days,and12days.
4 Days 8Days 12Days
Larval Family (n) %surv. meanwt. instar %surv. meanwt. instar %surv. meanwt. instar
RUTACEAE
Citrussinensis"joppa"
G1 0 na
G2 3 33.3 3.7 1 died
G3 0 na
G4 0 na
T1 0 na
T2 3 0 died
Flindersiaaustralis
G1 0 na
G2 3 0 died
G3 0 na
G4 0 na
T1 3 0 died
T2 0 na
Geijerasalicifolia°
G1 2 0 died
G2 3 0 died
G3 2 0 died
G4 2 0 died
T1 2 0 died
T2 2 0 died
MAGNOLIACEAE
Magnoliavirginiana
G1 2 100 9.9 1.5 100 39.6 2.5 100 147.1 4
G2 3 66.7 8.4 1.5 66.7 31.8 2.8 66.7 125.1 3.5
G3 2 50 8.4 1.5 50 27 2.5 50 106.7 3
G4 2 50 7.6 2 50 44.7 3 50 214.3 4
T1 2 0 died
T2 3 0 died
Micheliachampaca
G1 2 100 10.3 1.5 50 72.8 3 50 427.1 4
G2 3 66.7 17.7 2 33.3 111.9 3 33.3 450 4
G3 2 100 10.7 1.8 100 41.4 2.8 100 205.6 4
G4 2 0 died
T1 2 0 died
T2 3 0 died
Micheliadoltsopa
G1 2 100 4.9 1 100 9.5 1.8 100 17.1 2
G2 3 66.7 4.6 1 100 7.4 1.3 100 6.7 1,3
°
Volume 62, Number 1 23
Table 1. (continued)
4Days 8Days 12 Days
LarvalFamily (n) %surv. meanwt. instar %surv. meanwt. instar %surv. meanwt. instar
Micheliadoltsopa (cont.)
G3 2 100 7 bo 100 27.2 2.8 100 136.9 3.3
G4 died
T1 died
T2 died
LAURACEAE
Beilschmiediaobtusifolia
G1 2 0 died
G2 3 0 died
G3 2 0 died
G4 2 0 died
T1 2 0 died
T2 2 0 died
Cinnamomumcamphora
G1 2 0 died
G2 3 100 8.7 100 15.1 2 100 41.3 2.8
G3 0 na
G4 2 0 died
T1 2 0 died
T2 3 66.7 3.8 1.3 66.7 4 1.5 0 died
Cinnamomumoliveri
G1 0 na
G2 3 66.7 0 died
G3 0 na
G4 0 na
T1 0 na
T2 3 66.7 3.7 1,3 66.7 4 1.5 0 died
Cinnamomumvirens
G1 0 na
G2 3 33.3 0 died
G3 0 na
G4 0 na
T1 0 na
T2 0 na
Cryptocaryaglaucescens°
G1 2 0 died
G2 3 0 died
G3 2 0 died
G4 2 0 died
T1 2 0 died
T2 2 0 died
Cryptocaryamicroneura
G1 0 na
G2 3 0 died
G3 0 na
G4 3 0 died
T1 0 na
T2 0 na
m
°
24 Journal ofthe Lepidopterists’ Society
Table 1. (continued)
4Days SDays 12Days
LarvalFamily (n) %surv. meanwt. instar %surv. meanwt. instar %surv. meanwt. instar
Endiandradiscolor
G1 2 0 died
G2 3 0 died
G3 2 0 died
G4 2 0 died
T1 2 0 died
T2 2 0 died
Litsealeefeana
G1 2 0 died
G2 3 33.3 2.3 1 33.3 2.9 1 0 died
G3 2 0 died
G4 2 0 died
T1 2 50 3.8 1 50 2.4 2 0 died
T2 2 0 died
Neolitseadealbata°
G1 2 0 died
G2 3 0 died
G3 2 0 died
G4 2 0 died
T1 2 0 died
T2 2 0 died
MONIMIACEAE
Antherospennamoschatu
G1 2 0 died
G2 3 0 died
G3 2 50 1.1 1 0 died
G4 2 0 died
T1 2 0 died
T2 3 0 died
Doryphorasassafras"
G1 2 0 died
G2 3 0 died
G3 2 0 died
G4 2 0 died
T1 2 0 died
T2 3 0 died
WINTERACEAE
Tasmanniainsipida
G1 2 100 4.7 1 100 10 1.8 100 17.1 2
G2 3 66.7 4.6 1 66.7 7.4 1.3 66.7 6.7 1.5
G3 2 100 2.3 1 50 7.4 1 50 6.5 9
G4 3 50 3.7 1 50 7.9 1.5 50 18.1 9
T1 2 0 died
T2 2 0 died
°° ° ,
Volume 62, Number 1 25
Table 1. (concluded)
4Days 8Days 12Days
LarvalFamily (n) %surv. meanwt. instar %surv. meanwt. instar %surv. meanwt. instar
WINTERACEAE
(cont.)
Tasmannialanceolata
G1 2 0 died
G2 3 0 died
G3 2 0 died
G4 2 0 died
T1 2 0 died
T2 2 0 died
ANNONACEAE
Annona muricata°
G1 2 0 died
G2 3 0 died
G3 2 0 died
G4 2 0 died
T1 2 0 died
T2 3 0 died
Annona reticulata
G1 2 0 died
G2 3 0 died
G3 2 0 died
G4 2 0 died
T1 2 0 died
T2 3 0 died
SAPOTACEAE
Pouteriaaustralis
G1 2 0 died
G2 3 0 died
G3 2 0 died
G4 2 0 died
T1 2 0 died
T2 2 0 died
°TherewasnonibblingorfecesinanyofthedishesofGeijerasalicifolia (Rutaceae);Beilschmiediaobtusifolia,Cryptocanjaglaucescens
Encliandradiscolor orNeolitseadealbata (Lauraceae);Dorypliorasassafras (Monimiaeeae);AnnonamuricataorA. reticulata (Annonaceae);
,
Tasmnannialanceolata(Winteraceae);orPouteriaaustralis(Sapotaceae).
were unsuitable for survival and growth (e.g. A(MS permit). In addition, neonates from all 4 families
Beilschmiedia abtusifolia, Cn/ptocanja glaucescens of P. glaucus could feed and survive on Tasmannia
,
Endiandra discolor and Neolitsea dealbata), insipida (Winteraceae). However, the congeneric T.
Cinnamomumcamphora C. virens and Litsea leefeana lanceolatawasunsuitableforanyoftheP.glaucus larvae
,
supported feeding (producing 327, 398, and 192 fecal and there were no feces. Some attempts to feed on the
pellets, respectively) and some growth oflarvae (Table Monimiaeeae and Rutaceae were observed for certain
1). However, evenwith some feedingstimulants inthese P. glaucus families, but this was not successful since all
3 hosts, all P. troilus larvae died before the third instar larvae died before day 8. Excellent survival and growth
andday 12 (Table 1). wereobservedonthe Magnoliaceae (Table 1). Magnolia
One family of the generalist P. glaucus also fed virginiana (sweet bay) is a favorite of P. glaucus in
(producing335, 220, and 187 fecalpellets, respectively) Florida (Scriber 1986, Scriber et al. 2001) and larvae
and grewon the same 3 Lauraceae species as P. troilus. grew well on the leaves ofthis large tree species from
OnlyC. camphora supportedgrowth to the third instar the Brisbane Botanical Gardens. All 4 families of P.
and up to day 12 (when killed in accordance with the glaucus also grew very well on leaves of Michelia
)
26 Journalofthe Lepidopterists’ Society
champaca andM. doltsopa despite theirgeographically Lauraceaespecialist, P. troilus. Itwasevidentthatsome
,
distant Asian origins. Phytochemical common of the Lauraceae assayed here Beilsclimiedia
( ,
denominators among Magnolia species might largely Cryptocarya Endiandra, and Neolit.sea species) were
,
explain this high suitability of such allopatric plant unsuitable for neonate growth and survival, although
species forP. glaucus. they did feed on one Litsea and two Cinnarnomum
The phylogenetic closeness of Winteraceae and species (Table 1). Differential utilization abilities of
Magnoliaceae may reflect some phytochemical plant species within the Lauraceae has been
similarities, as is suggested by the high survival and documented for P. troilus (LederhousC et al. 1992) and
successful growth of P. glaucus on Tasmannia insipida amongits geographicalpopulations inthe USA(Nitaoet
aswell as theMichelia andMagnolia species. However, al. 1991). The introduced southeast Asian
no survival (or feeding) on T. lanceolata (= Drimys Cinnarnomum camphora has elicited opposition and
aromatica was observed, suggesting different larvalfeedingbyP. troilus on an ornamentalplantingof
suitabilities (ortoxicities)within this plant genus. this tree in the USA (Morris 1989). It is known that the
furanocoumarin-metabolizing cytochrome P450
Discussion
enzymes found in many Rutaceae feeders (including P.
The evolutionary constraints that have restricted P. glaucus and P. canadensis Li et al. 2001) are lackingin
troilus to only Lauraceae, and the ecological P. troilus (Cohen et a;l. 1992). Behavioral cues
opportunities that were taken by P. glaucus on 9 (stimulants) to P. troilus adults and larvae also seem to
families ofplants in North America (see Fig.2; Scriber be missing in plants other than Lauraceae (Carter &
1988; Scriber et al. 1991b) were confirmed with our Feenv 1999; Carteret al. 1999; Frankfater and Scriber
neonate larval assays here using22 species ofAustralian
1999; Scriberetal. 2001).
plants. With a veiy narrow host range, the Spicebush While P. glaucus can and does use spicebush and
Swallowtail, P. troilus, grows with 2-4 times the sassafras naturally, they are not favored hosts. Survival
efficiency and rate of the generalist P. glaucus on the of2042 individuals from 44 different populations from
same plant, (Scriber & Feeny 1979; Scriber 1984b). In 17 different States (and Canada and Mexico) was only
fact there have been no other species of insects ever 14% overall, compared to 68% for P. troilus (6 States,
reported with significantly higher growth rates and 28 families, 621 larvae: Scriber 2005). While the
efficiencies in various instars than P. troilus on generalist P. glaucus does naturally feed on sassafras
sapcciecpetbusahn(dScrdiebteorxi2f0y05)c.loPsoetleyntiraellaltoesds offamaibliliietsies(toor baonrdbosnpiiaceablussohof(StchreibLearureatceaal.e1)97is5)t,oxircetdobalalyne(oPneartseeas
Rutaceae; Scriber et al. 2008a) is suggested by the tested, ki,lling 228 larvae ofthe Florida population and
unwillingness and/or inability of neonate P. troilus to 432 larvae of the northern P. glaucus populations
feedandgrowon anyplants inthe 6plantfamiliesother (Scriber et al. 1995, Scriber 2005, Table 2). Although
than Lauraceae in these bioassays. Despite the close unknown regarding specific toxins for Papilio, insect
phylogenetic relationships of the ancient Australian toxins have been identified from Persea (Maetal. 1988;
Monimiaceae and Winteraceae with the Lauraceae, Gonzalez-Colomaetal. 1990).
their leaves are unsuitable (repellent or toxic) for the
Table2. NeonatelarvalsurvivalofP troilusandP. glaucusonplantsofNorthAmericanLauraceae,andAustralianMonimiaceae,
andWinteraceae. Dataarepresentedas %survival, and(n=totallarvae).
Lauraceae Monimiaceae Winteraceae
RB SP SA CT(US) CT(A) Dsas Amos Tins Tlan
P. troilus 55% 86% 77% 50% 75% 0% 0% 0% 0%
(143) (156) (404) (82) (4) (5) (5) (4) (4)
P.glaucus 0% 24% 60% 62% 33% 0% 11% 67% 0%
(432) (579) (306) (134) (9) (9) (9) (9) (9)
AmRoBs==rseodutbhaeyr;nSsPa=ssasfpriacse,bAunsth;heSrAos=peSrasmsaafmroasscahlabtiudmum;T;iCnTs== TCaasmmpahnoniratrienesi(piindaU,SaAnd&TilnanAu=stTr.alliaan)c;eoDlastaas.=Donjpliorasassafras;
NorthAmericandata(4columnsattheleft) arefrom Scriberetal (1991, 1995)
Volume 62, Number ] 27
P. troilus
larva
on
Spicebush
P. glaucus
P. glaucus male with dimorphic Lauraceae specialist
larva Yellowand dark females
on sweet bay
Magnoliaceae favored, hut very
polyphagous (multi-family generalist)
Fig. 2. NorthAmericanP. glaucusandP. troilusontheirfavoredhosts.
Itis apparentthatP.glaucus andP. troilus attemptto North American P. troilus. Both ofthese plant species
feedonLitsea leefeana leaves fromAustraliaaswellas 2 are hosts of Graphium macleayanum Leach (Braby
species ofCinnamomum (C. campliora and C. oliveri 2000; Scriber et al. 2006). However, despite the use of
;
Table 1). However, these Laurcaeae species are not both species of Tasmannia (T. insipida and T.
suitable hosts for either butterfly species. Recent lanceolata) by Graphium macleayanum in Australia,
experimental feeding studies with camphor tree (C. these plants were totally unsuitable for P. troilus.
camphora) have shown this invasive tree species is However, the North American P. glaucus grew
acceptablefortheAntherospermamoschatum specialist successfully on T. insipida (but died on T. lanceolata
;
Graphium macleayanummoggana inTasmania(Scriber Tables I & 2). Thephytochemicalbasis andgenetically-
et al. 2006), as well as for the Rutaceae specialist, based differences in feeding behavior and larval
Papilio aegeus in Queensland (Scriber et al. 2007). A detoxification abilities deserve furtherstudy. The leafoil
fundamental common phytochemical arrayof nutrients cells of Tasmannia lanceolata are known to contain a
and allelochemieals in camphor tree apparently serves sesquiterpene chemical called polygodial, which has
thebasicnutritionalneeds forlarvaeofphylogenetically been shown to have antimicrobial activity (Kubo &
divergent Australian and North American taxonomic Taniguchi 1988) andpiscicidalproperties (Ciminoetal.
groups of Papilionidae. However, it remains unknown 1982). It has also been shown to have antifeedant
whether camphor tree has been an ancestral plant for properties forsome insects (Powelletal. 1995).
anyofthe Papilionidae. Species of Magnoliaceae, while toxic to P. troilus
Despitethe samecommon names, closephylogenetic (Scriber et al. 1991), can serve as a host for several
origins, and a similar aromatic smell between the Australian Papilionidae even though the plants are not
Monimiaeeae (sassafras= Doryphorasassafras southern found there naturally. The Rutaceae specialist, Papilio
;
sassafras = Antherosperma moschatum) and the aegeus was reared to pupation on Magnoliaceae
Lauraceae (sassafras= Sassafras albidum the including Magnolia virginiana (sweet bay; from North
),
Australian Monimiaeeae were not at all suitable for the America), Michelia champaca (yellow magnolia; from