W

i‘

 
  

I
w

M

‘UWHWIWJH

I

IIHHIHHIWIWHHH

!
("M373

T

lllllllfllfllllll

50 73

 

 

 

 

 

 

  
    
   

  

LIBRARY

Michigan State
Universlty

This is to certify that the

thesis entitled

PHYTOCHEMICAL BASIS FOR HOST PLANT SELECTION
BY GENERALIST AND SPECIALIST SWALLOWTAIL

BUTTERFLI S
presente by

Cheryl R. Frankfater

has been accepted towards fulfillment
of the requirements for

Masters, degree in inhomolugy

 

Major professor

/
Date Au ust 12 -’

c.7639 Msu-‘m- A;=---~'-- ‘ ' " '3”, ”,lmtimu'on

1

 

 

PLACE N RETURN BOX to romovo this checkout from your record.
TO AVOID FINES return on or before date duo.

DATE DUE DATE DUE DATE DUE ll

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU Is An Afflrmotivo AdlanlEqud Opportunity Institution

PHYTOCHEMICAL BASIS FOR HOST PLANT SELECTION BY GENERALIST
AND SPECIALIST SWALLOWTAIL BUTTERFLIES

By

Cheryl R. Frankfater

A THESIS

Submitted to
Michigan State University
in partial fiilfillment of the requirements
for the degree of

MASTER OF SCIENCE

Department of Entomology

1996

ABSTRACT

PHYTOCHEMICAL BASIS FOR HOST PLANT SELECTION BY GENERALIST
AND SPECIALIST SWALLOWTAIL BUTTERF LIES

By

Cheryl Frankfater

Swallowtail butterflies species show clear patterns of oviposition preference
among plants in the field and in the laboratory. Oligophagous butterflies, such as Papilio
troilus and P. palamedes, oviposit on a select group of lauraceaous plants. In contrast, P.
glaucus, a polyphagous swallowtail butterfly, will deposit eggs in a consistent preference
hierarchy on trees from several families. Oviposition preferences are maintained by the
presence of oviposition stimulants and deterrents in the foliage of host and non-host
plants, detected by chemoreceptors located on the front tarsi of butterflies. Polar extracts
and fractions from the preferred host plants of P. troilus and P. palamedes sprayed on to
various substrates have been shown to stimulate oviposition relative to the controls.
However, in addition to contact chemical stimuli, other sensory cues, such as color, odor
and texture may also be important in eliciting maximum oviposition. Rejection of non-
host plants also has a chemical basis. P. glaucus avoid ovipositing on host leaves sprayed
with polar extracts from a non-host tree species. The experimental results decisively
show that phytochemical stimuli detected through tarsal chemoreceptors governs host

plant selection by these swallowtail butterflies.

ACKNOWLEDGMENTS

7 I would like to thank Dr. Mark Scriber for introducing me to swallowtail
butterflies. (I had never even seen a swallowtail butterfly before I came to Michigan
State). His enthusiasm for their study inspired me and made me excited about my own
project. I would also like to recognize my committee members for their contributions to
my research. Many thanks go to Dr. Muralee Nair, for allowing me to work in his lab
and generously providing me with advice and resources. I am greatly appreciative of Dr.
James Nitao for his assistance with everything from chemical techniques to statistics. I
always knew that I could count on him for help. Dr. Jim Miller gave me a solid logical
foundation in his Nature and Practice of Science seminar, and worked with me to
improve my lectures when I was a teaching assistant in his Insect Physiology class. Dylan
Parry helped me enormously in editing and refining my thesis. I am very grateful to my
parents, for giving me an early interest in science, and for their endless supply of love and
support. They also helped me get organized (to put it mildly) during their visits to East
Lansing! I would like to extend a special thanks to Scott Alcock, who has stood by me
across distances large and small, for always listening and responding to my concerns
about my research and other matters. Patti Tidwell first taught me the ins and outs of
swallowtail butterfly rearing, and I sincerely enjoyed her friendship throughout the

iii

tedium of larval caretaking. Dr. Wayne Wehling also shared friendship and advice along
the way. I am also grateful to Renee Brenner, who conscientiously and efficiently carried
the lab through many rearing seasons. Thanks go to Ben Emery and my father for
assisting me in setting up late night-early morning bioassays and to all my other friends
who ever got unwittingly roped into taking care of larvae. I would like to extend a warm
appreciation to Amy Roda and Beth Dankert who are the greatest friends a person can
have. Ultimayhem and Throwing Muses can take credit for many of the great times I had

in Michigan, both on and off of the playing field.

iv

TABLE OF CONTENTS

List of Tables ....................................................................................................................... vi
List of Figures .................................................................................................................... vii
Introduction ............................................................................................................................ 1
Preface ....................................................................................................................... 1
Ancestral host-affiliations ......................................................................................... 2
Evolution of oviposition preference ......................................................................... S
Proximate factors ....................................................................................................... 8
Host finding mechanisms .......................................................................................... 9
The fundamental importance of phytochemistry .................................................... 13
Chapter 1
Chemical basis for host recognition by two oligOphagous swallowtail butterflies, Papilio
troilus and Papilio palamedes (Lepidoptera: Papilionidae) ................................................ 19
Abstract .................................................................................................................... 19
Introduction ............................................................................................................. 19
Materials and Methods ........................................................................................... 21
Results .................................................................................................................... 26
Discussion ............................................................................................................... 27
Chapter 2
Chemical extracts from a non-host deter oviposition by a polyphagous
swallowtail butterfly, Papilio glaucus (Lepidoptera: Papilionidae) ................................... 37
Abstract ................................................................................................................... 37
Introduction ................................ 38
Materials and Methods ........................................................................................... 40
Results .................................................................................................................... 44
Discussion .............................................................................................................. 46
Conclusion ........................... - ................................................................................................ 5 8
Appendix 1 Record of deposition of voucher specimens ................................................ 62
Appendix 1.1 Voucher specimen data ............................................................................... 63
List of references ................................................................................................................ 64

Table 1.

Table 2.

Table 3.

Table 4.

LIST OF TABLES

Average lifetime percentage of eggs distributed across a 7-choice array of
plant species for different populations of Papilio glaucus .............................. 15

Average lifetime percentage of eggs distributed across a 7-choice array of
plant species for different populations of Papilio troilus and P. palamedes...16

Average lifetime percentage of eggs distributed on three lauraceous host
plants by Papilio troilus and P. palamedes ..................................................... 17

Percentage of eggs laid by 5 Papilio glaucus females from Highlands County,
Florida in a 3-choice bioassay with sweetbay, redbay and white ash .............. 50

vi

Figure 1.
Figure 2.
Figure 3.

Figure 4.

Figure 5.
Figure 6.
Figure 7.

Figure 8.

Figure 9.

Figure 10.

Figure 11.

Figure 12.

LIST OF FIGURES

Consensus taxonomy for monophyletic section III Papilio ............................. 18
Papilio troilus 4-choice filter paper bioassay with spicebush extracts ............ 33
Papilio troilus 3-choice assay with spicebush extract on tulip tree ................. 34

Comparison between spicebush extracts in a 5-choice filter paper array with

a 5-choice tulip tree array with Papilio troilus ................................................ 35
Papilio palamedes 5-choice tulip tree leaf bioassay ........................................ 36
Papilio glaucus from FL in a 3-choice bioassay with redbay extracts ............ 51
Papilio glaucus from FL in a 5-choice bioassay with redbay extracts ............ 52

Papilio glaucus from FL in a 3-choice bioassay with the hexane extract
of redbay .......................................................................................................... 53

Papilio glaucus from FL in a 3-choice bioassay with the ethyl acetate
extract of redbay ............................................................................................... 54

Papilio glaucus from FL in a 3-choice bioassay with the methanol extract
of redbay .......................................................................................................... 55

Papilio glaucus from OH, PA and KY in a 3-choice bioassay with
the methanol extract of redbay ......................................................................... 56

Papilio glaucus from OH, PA and KY in a 4-choice bioassay with the
ethyl acetate and aqueous fractions of the active methanol extract ................. 57

vii

INTRODUCTION

Preface

The geographic range of papilionid species (Lepidoptera) is determined by their
successful adaptation to various biotic and abiotic conditions. Thermal constraints,
geographic barriers, predation, competition and host affiliation all affect current
distribution patterns of swallowtail species. The ability of different papilionid species to
recognize, accept and survive on various plant species, paramount to their successful
radiation across latitudes and ecological niches, hinges on physiological recognition,
behavioral selection and biochemical detoxification of the host plant allelochemicals.
While vision and olfaction play a role in host location by gravid female swallowtails, the
final decision whether or not to oviposit rests on chemical stimuli perceived through
gustatory sensilla on the foretarsi (Feeny et al. 1988, Roessingh et al. 1991 , Nishida
1995). The amount and ratio of oviposition stimulants and deterrents in foliage provide
the basis for the relative preferences of swallowtails for their host plants (Honda and
Hayashi 1995). The research presented here focuses on assessing the relative importance
of leaf chemistry for oviposition for several swallowtail species. Both oviposition
stimulants and deterrents will be studied.

In the lab, female swallowtail butterflies, whether generalist or specialist, show a

1

2

consistent hierarchy of oviposition preference among host plants. Given the choice
between leaves of seven species in an oviposition bioassay, Papilio glaucus, a highly
polyphagous sWallowtail, deposits eggs with greatest frequency on tulip tree, hop tree,
white ash and black cherry, relatively fewer eggs on spicebush, and fewest on Rhamnus
spp. and quaking aspen (Scriber unpubl., see Table l). Swallowtails specializing on
Lauraceous plants, such as P. troilus and P. palamedes, also show consistent oviposition
preferences, placing greater than 70% of their eggs on these hosts in a 7—choice array
(Table 2). In 3-choice bioassays, P. troilus females put the majority or their eggs on
spicebush and sassafras, and lay notably fewer eggs on redbay, whereas P. palamedes lay
eggs predominantly on redbay, with fewest on sassafras, and basically reject spicebush.
(Lederhouse et al. 1992, see also Table 3). Over evolutionary time, natural selection on
host choice and chemical constraints imposed by physiological adaptations provided the

basis for current oviposition acceptance patterns.

Ancestral host-affiliations
Ancestral host-affiliations have guided the formation of modern swallowtail-plant
associations. F eeny’s chemical facilitation hypothesis (1991) suggests that adaptations to
host plant chemistry may pave the way for host shifts to chemically similar taxa. It is
believed that the ancestor of section III Papilio predominantly fed on Rutaceae before the
disappearance of many of its more semi-tropical representatives from the middle latitudes
during global cooling (Scriber 1995). Rutaceae-feeding may have preadapted several

Papilio species (such as Papilio machaon) to utilization of the Umbelliferae and

3

Compositae, since these three families share several chemical constituents (Dethier
1941 ). Papilio glaucus provides an example of a species that may have "escaped"
chemical constraints and has broadened its host range to include families of many
temperate plants (Scriber et al. 1991 a).

Several pieces of evidence support the hypothesis that Rutaceae is the ancestral
host family of both the glaucus and troilus lineages (section 111), two closely related
monophyletic taxa (Figure 1). The survivorship of the glaucus group caterpillars on
hoptree (Ptelea trifoliata, Rutaceae) is notable. More than 60% of all first instar
caterpillars from 6 of the 8 glaucus group species survived to the second instar on
hoptree. Because troilus group larvae refuse to initiate feeding on plants outside the
Lauraceae, presumably due to a lack of feeding stimulants, Scriber et al. (1991a) could
not assess the detoxification ability of those caterpillars on rutaceous plants. The shared
use of Rutaceae by many species in sections 111, IV and II of the tribe Papilionini (Scriber
1984, 1996) as well as some species in the Graphiini and Troidini also suggests ancestral
origins of rutaceous feeding. Additionally, natural oviposition by P. glaucus has been
observed on Rutaceae and Lauraceae (Scriber et al. 1991a).

Similar kinds of evidence points to Lauraceae as a possible ancestral host family
of the two groups. 4 glaucus group species, plus 2 troilus group species show 60% or
more survival of first instars on sassafras (Sassafras albidum, Lauraceae). Section V
Papilio as well as other clades within the Papilionini and Leptocircini also utilize plants
in this family (Scriber et al. 1991a). Both the Rutaceae and Lauraceae have in common

the production of benzylisoquinoline alkaloids, essential oils and coumarins (Berenbaum

4

1995). Behavioral recognition and metabolic detoxification mechanisms may have
facilitated shifts to current hosts with underlying chemical similarities.

Ancestral adaptations to Rutaceae and Lauraceae chemistry conceivably placed
Papilio troilus on a path for utilization of modern, temperate-zone Lauraceae. Polyphagy
of the contemporary glaucus group is less easily explained. However, the Rutaceae,
Magnoliaceae (tulip tree is the preferred host plant of P. glaucus) and Lauraceae families
are conspicuously united by the production of essential oils, lignans, mono- and
sesquiterpenes, and to alesser extent, phenylpropanoids (Berenbaum 1995).

Butterfly species may still behaviorally recognize and oviposit on ancestral hosts.
Allozyme electrophoresis on the glaucus species group indicates that the tiger swallowtail
species P. canadensis and P. glaucus, previously considered subspecies (Hagen et al.
1991), have most recently diverged from each other in comparison with the other glaucus
group species. In lab bioassays, both P. glaucus and P. canadensis lay a substantial
number of eggs on tulip tree, a favorite host of P. glaucus, although essentially all P.
canadensis larvae fail to survive on the foliage. P. canadensis females laid a striking
average of 43% of the eggs on tulip tree in a 3 choice bioassay between tulip tree, black
cherry and quaking aspen despite the fact that P. canadensis larvae can survive and
develop successfully only on the latter two hosts (Scriber et al. 1991b, Bossart and
Scriber 1995). The majority of P. canadensi's never encounter tulip tree in their range,
except at their southern border. The common ancestor of P. glaucus and P. canadensis
likely utilized plants in the family Magnoliaceae before the two subspecies diverged

geographically and physiologically, and P. canadensis appears to have retained the ability

to recognize and accept this host.

Evolution of oviposition preference

In Papilio, oviposition preference at least generally reflects larval performance in
terms of growth rate, pupal weight, and survivorship. For the most part, females will.
avoid oviposition on plants toxic to their progeny. P. canadensis populations that
encounter tulip tree at its northernmost limit in Central Michigan tend to show a
decreased preference for oviposition on this plant compared to their counterparts from
northern Michigan, Canada and Alaska that do not naturally contact tulip tree (Bossart
and Scriber 1995). Singer et al. (1988) found that the degree of preference 3 female has
for one of two host plants is significantly positively correlated with the mean larval
weight of her offspring on that host after 10 days. Similarly, Bossart (1993) found that
the average relative growth rate of P. glaucus larvae on tulip tree positively correlated
with the mother’s preference for tulip tree in a 2-choice bioassay with tulip tree and
sweetbay. Under experimental conditions, Euphydryas chalcedona butterflies preferred
to oviposit on the more nutritious of their two host plants (Williams, 1983).

However, discrepancies exist between oviposition preference and larval
performance on potential host plants in the field. The general trend of P. canadensis
females to avoid tulip tree in the lower latitudes appears inconsistent with the total
inability of the larvae to detoxify the foliage. These findings suggest that the ability to
avoid tulip tree is evolving slowly in populations that encounter it (Bossart and Scriber

1995). In the field, E. chalcedona butterflies oviposit on both D. aurantiacus and S.

6

californica, regardless of their differing nutritional value to the larvae. S. californica, the
less abundant and less persistent host, often thrives in shadier habitats than D.
aurantiacus where E. chalcedona butterflies rarely fly. Other selection pressures in
addition to larval performance regulate host plant selection in the field.

A multitude of ecological factors may direct the evolution of oviposition
preference (Strong 1988 series, Bemays and Chapman 1994a). Parasitism, for example,
may take a heavy toll on larval survival and levels may vary with host plant (Damman
and Feeny 1988, Ohsaki and Sato 1994). Although all 3 Pieris species studied by Ohsaki
and Sato (1994) developed best on the cultivated cabbage species in the laboratory, two
species preferred to oviposit on other plants that allowed escape from parasitism although
larval performance suffered slightly. Smiley (1978) noted that Helicom'us melpomene
utilized only 1 out of the 5 of the plants that supported greatest caterpillar survivorship
and fitness in the lab, ovipositing exclusively on the species that offered protection from
parasitoids. Interspecific competition with the gypsy moth, Lymantria dispar, for
quaking aspen foliage as well as an increased parasitism rate by generalized tachinids in
gypsy moth infested quaking aspen stands may have facilitated the shift to gypsy moth-
free white ash (Scriber and Redman 1993).

The abundance and availability of potential host plants affect evolution of
oviposition strategies (Bossart and Scriber 1995, Wiklund 1984). Sweetbay (Magnolia
virginiana) is the only common host plant encountered by Papilio glaucus populations
throughout most of Florida (Scriber 1986). Florida P. glaucus females showed an

increased preference for sweetbay relative to females from populations in Georgia and

7

Ohio that rarely or never encounter sweetbay. The imposed “ecological monophagy”
(Scriber 1986) probably accounts. for the documented regional specialization in
oviposition preference (Bossart and Scriber 1995).

Thompson and Pellmyr (1991) suggest several other explanations for oviposition
choice seemingly at odds with larval performance in the laboratory. For example, the
preferred plant may be rare; a plant may be a recent addition to a habitat and there has not
been ample time to evolve an avoidance mechanism; the habitat may be unfavorable for
ovipositing females or larval growth; females may oviposit on more “toxic” plants that
allow for larval sequestration although slower growth ensues; and females may search for
clumped hosts in species that require consumption of several individuals to complete
development. Rausher (1995) hypothesized that tradeoffs in larval and adult fitness may
account for counterintuitive oviposition behavior of females Battus philenor females. In
northern Mexico, B. philenor mothers preferentially search in sunny, rather than shady
habitats, although the larval survival was much lower in sunny habitats and the rate of
pupal parasitism was much higher. Rausher (1995) suggests that the abundance of orb-
weaving spiders in shady habitats may increase adult mortality or that therrnoregulatory
constraints decrease oviposition efficiency in shady habitats although neither alternative
has been tested.

Oviposition preference evolves in the context of other dynamic life history traits
and phenology. For example, cold areas that offer a notably shorter developmental
window between frosts may select for delayed emergence, delayed mating or delayed

egg-laying that facilitate the host shift to phenologically late and nutritious white ash

8

noted in populations of Papilio canadensis from northern Michigan. In addition, a short
developmental window may also select for adaptations such as larger egg and first instar
size and pupation at lower mass to insure fast completion of development (Scriber 1994)
such as has been observed in Alaskan populations (Ayers and Scriber-1994).

The interaction of thermal degree day restrictions coupled with the number of
generations possible in a given area may further complicate evolution of oviposition
preference. Gravid females may show increased specificity for nutritious host plants in
locations that allow barely enough time to complete the maximal number of generations.
Host specificity may relax in latitudes where there is abundant time for completion of n
generations, but not enough time for completion of n+1 generations. The overall effect
may be alternating bands of greater and lesser host specificity (N ylin 1988, Scriber and
Lederhouse 1992).

Evolution may affect host selection on several levels, from interspecific
discrimination to selection of a particular oviposition site on a given plant individual.

However, many proximal factors also affect the oviposition decisions of females.

Proximate factors
Circumstantial factors impacting the life of the adult female influence her choice
of oviposition sites. For example, the proximity and location of nectar sources strongly
influences the oviposition patterns of butterflies (Murphy 1983, Grossmueller and
Lederhouse 1987). Despite the presence of sufficient tulip tree foliage during both

generations of P. glaucus, females only visited and oviposited in the area during the

9

second flight when the thistles were in bloom (Grossmueller and Lederhouse 1987).
Females of another butterfly species, Euphydryas chalcedona, may avoid certain sites due
to harassment by males attempting to mate (Williams 1983). The internal motivation of a
female also influences her oviposition selectivity. In the case of Battus philenor, the size
of the egg cluster is influenced by time since the last oviposition opportunity and host '
plant characteristics (Rausher 1995). In a no-choice situation, Pieris rapae will oviposit
on Iberis amara after an initial delay, a plant it normally avoids. Bossart (1993) similarly
found a negative correlation between host specificity and egg load, and a positive
correlation between age (reflecting a decline in egg load) and host specificity (see also
Scriber 1993). The time and energy budget of a female, coupled with her motivational

state will influence oviposition decisions (Underwood 1994).

Host finding mechanisms

The size and surrounding vegetation of a plant affects its apparency to ovipositing
females. Masumoto et a1. (1993) observed that Anthocharis scolymus females approach
taller host plants with greater frequency, particularly when the surrounding vegetation
was sparse. Numbers of eggs and larvae were greatest on taller hosts, but decreased as
the density of surrounding conspecific plants and other vegetation increased (Masumoto
et al. 1993). Monarch butterflies (Oyeyele and Zalucki 1990) and parsnip webworrns are
also inclined to oviposit on taller plants (Zangerl and Berenbaum 1992).

Ovipositing butterflies use visual and chemical cues to lead them to the

appropriate larval host plants. Rausher (1978) observed that female Battus philenor

10

preferentially alight on leaves of a particular shape, regardless of whether or not the plant
is a host plant, indicating that discrimination of leaf shape occurs at a distance. Papilio
demoleus shows a positive orientational response to glass-screened cut branches of a host
(Citrus limettioides) and a non-host (Gossypium hirsutum) indicating a visual attraction
to general foliar characteristics (Saxena and Goyal 1978).

Butterflies show intrinsic behavioral responses to various colors (Ilse 193 7,
Saxena and Goyal 1978, Kolb and Scherer 1982, Traynier 1984). For example, certain
colors or wavelengths may elicit either a feeding or drumming response (Ilse 1937, Kolb
and Scherer 1982). In an experiment where discs were illuminated with monochromatic
light at various wavelengths, wavelengths between 542-578 nm elicited the greatest
number of drumming events by Pieris brassicae females. When the discs were sprayed
with sinigrin to elicit oviposition, the greatest number of eggs were laid discs illuminated
under 522-554 nm. These wavelengths compose the yellow-green region of the spectrum
(Kolb and Scherer 1982). Gravid P. demoleus females were observed to fly toward
yellowish-green and green muslin patches (Saxena and Goyal 1978). In his experiments
on learning, Traynier (1984) noted inherent preferences of P. rapae for oviposition on
green-colored substrates, and even particular shades of green.

Olfactory cues both serve to attract and to arouse the ovipositional urge in
butterflies. In a wind tunnel, two polyphagous moths, Heliothis virescens and
T richoplusia ni, were attracted by odors from several plant species (Bemays and
Chapman 1994). Papilio demoleus showed a significant orientational preference to a

Citrus branch placed outside the nylon net wall of their cage relative to a glass-screened

ll

branch on the opposite side. Some abdominal curling and oviposition occurred on the net
adjacent to the unscreened plant without any contact with the foliage, indicating that odor
also serves in part to stimulate oviposition in the presence of visual cues (Saxena and
Goyal 1978). Containers of host-plant volatiles elicited more landings by Papilio
polyxenes on a contact-stimulant treated model plant relative to control cages lacking the
volatile stimuli. In these studies, the presence of volatiles increased the overall activity of
the butterflies, as measured by “flutter bouts” (F eeny et al. 1989). Manduca sexta were
observed to fly toward an odor source and cease flight upon its removal. Volatile
compounds may not only serve as an orientation cue, but as a physiological primer
stimulating flight and oviposition in the presence of additional'cues (Bemays and
Chapman 1994).

Upon alighting on a plant, a gravid female perceives contact chemical cues
through tarsal chemoreceptors that influence her final decision whether or not to oviposit
(Feeny et al. 1988, Nishida and Fukami 1989, Honda 1990, Nishida et al. 1990,
Roessingh et al. 1991, Sachdev-Gupta 1993). The presence of oviposition stimulants and
relative lack of deterrents is crucial in the stimulation of oviposition. However, at this
level, the perception of the nutritional state of a plant (Myers 1985), color (Myers 1985,
Robertson 1987) and even phonological cues (Rausher and Papaj 1983, Darnman and
Feeny 1988) also influence her final decision. Information gained from sampling many
conspecific plants allows the butterfly to discriminate between several individuals afier
alighting (Traynier 1986).

A sensory synergism of acceptable visual and chemical cues leads to the greatest

12

likelihood of oviposition. For example, although Pieris brassicae females will land and
drum on discs illuminated with yellow-green light, oviposition will only occur in the
presence of a sinigrin solution (Kolb and Scherer 1982). Odor, color and contact
chemical stimuli elicit the greatest number of orientation, abdominal curling and
oviposition reponses by Papilio demoleus, and the absence of one or more of these cues
reduces the frequencies of these behaviors (Saxena and Goyal 1978).

The ability of butterflies to learn may enable them to link chemical cues with
physical cues (Traynier 1986, van Loon et al.1992). Battus philenor forms a search
image that can be altered by experience. Early in the season, broad leaved Aristolochia
reticulata predominate. Frequent encounters with A. reticulata reinforces the broad leaf
search image, enabling the butterflies to efficiently locate other A. reticulata plants. Later
in the season, A. reticulata foliage increases in toughness and declines in nutritional value
while narrow leafed A. serpentaria remains nutritious and able to support larval
development. A corresponding increase in landings on narrow-leaved foliage can be
observed after the maturation of A. reticulata. Papaj and Rausher (1987) also
demonstrated the ability of naive B. philenor females to learn to land on large young
plants in preference to large old plants after given the opportunity to land on plants in an
experimental enclosure. Pieris rapae approach and lay more eggs on nitrogen fertilized
plants than controls (Myers 1985, Wolfson 1980). Myers (1985) found that leaf
greenness and egg count were correlated and that greener plants had a higher
concentration of nitrogen in their leaves. In all four cases (Battus philenor, Eurytides

marcellus, Papilio xuthus, Pieris rapae) the larvae benefited from the oviposition

l3

decisions of the female, implicating a nutritive or chemical link between plant phenotype

and differential oviposition.

The fundamental importance of phytochemistry

The balance of stimulants and deterrents in a leaf determine whether or not a
butterfly will oviposit. In the case of P. rapae, the acceptability of bean seedlings (a non-
host) depended on the ratio of the stimulant compound to deterrent compound introduced
systemically into the plant. However, at very high stimulant levels, increasing the
concentration of the deterrent did not impede oviposition (Renwick and Huang 1994).
Differential sensitivity to stimulants and deterrents by closely related species of
butterflies provides the basis for differential host ranges in several Pieris and Papilio
species. The deterrent phellamurin found in the rutaceous plant Phellodendron amurense
strongly inhibits oviposition by Papilio protenor, while is only partly inhibitory to
Papilio xuthus at natural concentrations (Honda and Hayashi 1994). The dissimilar
sensitivities of the two species to this compound provide the basis for marginal
acceptance of Phellodendron amurense by P. xuthus and its complete rejection by P.
protenor in the field (Honda and Hayashi, 1995). Honda and Hayashi (1995) note that in
several Papilio species, such as P. xuthus, P. protenor and P. machaon, oviposition
stimulants and deterrents are commonly flavonoids or their derivatives. They
hypothesize that flavonoids originally deterred oviposition by Papilio species, but that
sensorial adaptation to the compounds allowed the colonization of the foliage containing

them. Based on feeding trials with many insects, Bemays and Chapman (1994a) have

l4

concluded that Oligophagous insects are deterred by more compounds than polyphagous
insects, and that polyphagous feeders are less sensitive to deterrents. This scenario may
also explain host acceptance by generalist and specialist ovipositing swallowtail
butterflies as well.

Several compounds with very similar structures often have ovipositional activity,
lending support to the chemical facilitation hypothesis proposed by Dethier (1941) and
Feeny (1991). In the pierid system, for example, indolyl and aromatic glucosinolates
appear more stimulatory than their aliphatic counterparts (Renwick and Huang 1994).
Closely related Papilio species may share sensitivity to the same family of chemical
compounds and may have used such compounds as a bridge in their radiation onto
different plant taxa. Additionally, host plants of a particular species may have in
common several characteristic compounds and a lack of strong deterrents. Elucidation of
the oviposition stimulants and deterrents in the various Papilio host plants will clarify the
mechanisms for host acceptance and host shifts, thereby providing clues to the
evolutionary relationships between different swallowtail species. I have chosen to make
the characterization of oviposition stimulants and deterrents for several species of
swallowtails the focus of my research.

In my thesis, I investigate the role of host chemistry in stimulating oviposition in
two Oligophagous butterflies, P. troilus and P. palamedes. Also, I present evidence for
the presence of oviposition-deterring compounds in a non-host of the polyphagous
butterfly Papilio glaucus. My thesis reinforces the fundamental importance of leaf

chemistry to host plant selection by specialized and generalized Papilio.

15

 

 

 

 

 

3 a HS 3 a as: M: a am pm a in mm a me 3 a 92 am a «fl do 3%sz
sets...—
3. 3 a a: 3 « EN 2 s 3. E a 3m 2 a e... 2 a 2. 2 a me 339 do v.56
an 3 a 3m 3 a ”.8 no a 3 em a 3m to a 3 2 a 5 2 a a.» 3:3 .8 x56
. £336
3 a est. 5. a we we a SN 2: « Ba 2 a 3 _.E a 3m 2 a as do BEES
S 2 a as on a 9mm ed a .2 me a gm 2 a .E o.~ a 3: o.~ « 3 do £30
3.5
we a fl: 3 a 4.: do a a... 3 a .3. 3 a a." mm a N.” S a ms .8 seem
m as a S: 3 a a: am a em we a menu 3. a as am a 3: am a «.2 .8 $55
, .3322
W M... m M M W“ M W. m H W O W m W W SEER Smack
w a 1. r d v o H n m. s w m
N v s v n m N d a . v x m s
a D . D a m u 0 n V u D 0 a.
s 3 m a a V m 1 a 3 8 3 u. 3 m.
W u. W D s W. 9 3 m M a W n.
w.

 

 

 

 

 

 

 

 

60:39:93 .3 B “opium 88m .33 .«o 5883 was wqtam 05 82m 829:8 Saw =< 6333M

Smack mo moosflaom Hoocowmv .8.“ 36on ESQ we .35 385$ a $88 “683166 ammo mo $8583 unava— owfio>< ._ 2an

16

 

 

 

 

 

 

 

 

 

 

 

a am e e e e E as .8 Wessex
«3.8—..—
W I uohoESem Snack
9 o.m H mam Wm H fie mm H flu md H m... c ad H a... 5.0 H edw .oU gauged
m a Wm H flu in H 1N c 9o H e... N.m H N.m H6 H mdw do «:30
, :56
i
d m m m m w W m m w. o m m w .m 33s 33$
9 ad 1... d U D .l. n S O n 0
w v a v n m w Nd 3 Mr. V x. w W
W. m w m MW m 0 m V8 m o n
9 V S 1 D U. 3 S
s v u. w 3 s w. a a m v a a u.
) an 3 d a I” V
(ax V d 3 W M 3
a W. u
3

 

 

 

 

 

 

 

 

Beaming": .3 3 custom 88m .33 mo SEES one woman 05 89a BEER. 8% 3.. .aohoESQQ N has Shot
223$ mo mcouflaoa Epcot? com 86on Ema mo .35 36:95 «.388 wousnwnmmv wwwo mo owficoocom carom: owfio>< .N 2an

16

 

 

 

 

 

 

 

 

_ «H a c a e E 95 .8 2.5sz
«3.83
mmhmESum Smack
e o.m H QM Wm H —.v fin H fin md H m6 c ad H a... >6 H cdw .oU 3:226..—
m e Wm H m.~ Wm H ed 6 ed H 3. N6 H ~.m HS H mg .00 «ENG
own—O
{
.. mm mm m w E .8 m m. 31......
9 a 1.. d U D .l. n S 0 n 0
m V 9 V n W w N d 3 m. V V4. W m-
m. m m m a m 0 m V 8 m o m
9 V S 1 a D U. 3
s w. u. w. a s w a m w. a w. u
w a w a w w m a
w w u
H

 

 

 

 

 

 

 

 

.855:an .3 8 “Bram 88m 4&2 mo 3883 can marge. 05 82m 3388 Sac =< .amhmESaQ N RS» .322“
Swank .«o 323389 Eobbfi 8m 863m Ema mo $5 0223;. «.388 35356 mmwo Mo 09358qu 2583— owfio>< .m 2an

17

Table 3. Average lifetime percentage of eggs distributed on three Lauraceous host plants
by Papilio troilus and Papilio palamedes. Data compiled between the years 1991-1994.
After Scriber et al. unpublished.

Papilio troilus

 

 

 

 

 

 

 

 

 

 

 

 

 

LOCATION REDBAY SASSAFRAS SPICEBUSH FEMALES (n)
Michigan 15.8 i 5.8 44.4 d: 7.3 39.8 :1: 7.4 17
Ohio

Gallia Co. 13.3 :L- 6.3 47.5 i 13.5 39.2 d: 16.6 19
Lawrence Co. 20.6 i 2.3 57.2 2}: 8.6 22.2 :t 6.4 9
Georgia

Echols Co. 30.6 at 11.5 23.6 i 12.2 45.7 :t 11.6 7
Florida

Highlands Co.* 45.9 :h 16.6 29.9 at 5.2 24.3 i 21.2 6
Papilio palamedes .

Georgia

Echols Co. 67.7 i 32.3 28.3 i 28.3 4.0 :l: 4.0 3
Florida

Highlands Co.* 74.7 i 4.8 13.8 :h 3.5 11.5 :t 3.2 10

 

 

 

 

 

 

* Redbay is the only Lauraceae species in this group of hosts that occurs in the southern

2/3 of Florida.

 

 

P. canadensis

P. gla ucus
P. alexiares P. rutulus
P. eurymedon
P. troilus
P. palamedes glaucus group
P. pilumnus

troilus group

secfioan
Papilio species

Figure 1. Consensus taxonomy for monophyletic section HI Papilio. From
Scriber et al. 1991.

CHAPTER 1
Chemical basis for host recognition by two Oligophagous swallowtail butterflies, Papilio
troilus and Papilio palamedes (Lepidoptera: Papilionidae)
Abstract
Females of Papilio troilus and P. palamedes, two closely related swallowtail

butterflies, oviposit almost exclusively on a few woody plant species in the family
Lauraceae. Dependence on chemical oviposition elicitors found only in the foliage of the
host plants may explain the behavioral fidelity shown by the two Oligophagous
butterflies. The role of host plant chemistry in stimulating oviposition was assessed by
extracting the preferred host foliage of the respective butterfly species, spraying the
extracts and fractions on various substrates, and assessing oviposition relative to controls.
P. troilus and P. palamedes were stimulated to oviposit on filter paper or non-host leaves
sprayed with polar extracts of their primary host plants. However, oviposition was
enhanced on non-host leaves relative to filter paper, perhaps due to a combination of

visual, tactile and olfactory cues that are characteristic of leaf surfaces.

Introduction
Host plant affiliations primarily determine the geographic range of lepidopterans.
The swallowtail butterflies P. palamedes and P. troilus recognize a very restricted range
of plant taxa as hosts, ovipositing exclusively on a few species within the Lauraceae. P.

19

20

palamedes inhabits the eastern coastal states from southern Virginia south through
Florida and west toward Texas and Mexico, closely corresponding to the geographic
distribution of redbay (Persea borbom’a), their predominant host in the field. The range
of P. troilus includes that of spicebush (Lindera benzoin) and sassafras (Sassafras
albidum), ranging from Canada to Florida east of the Mississippi River and occurring in
sympatry with P. palamedes in southern Florida (Lederhouse et al. 1992). P. troilus
utilizes redbay only in southern Florida, outside the habitat of spicebush and sassafras
(Nitao et al. 1991’).

Results of laboratory oviposition bioassays conform to patterns of host use in the
field. The two specialist species accurately place their eggs on lauraceous foliage when
placed in an array with 7 other non-host leaves in small bioassay containers (see Table 2,
introduction). P. palamedes and P. troilus also exhibit the relative preferences among
lauraceous hOSt species. In bioassays with leaves from 3 lauraceous species, P.
palamedes females deposit the majority of their eggs on redbay, and oviposit to a lesser
extent on spicebush and sassafras. P. troilus females oviposit mainly on spicebush and
sassafras, and place fewer eggs on redbay (Lederhouse et al. 1992, see Table 3, chapter
1).

I Behavioral responses to compounds distributed throughout a plant community
account for the patterns of host use observed in the field and laboratory. The ratio of
chemical stimulants and deterrents found in foliage determine the range of plant species
accepted for oviposition by the two specialist swallowtail butterflies (Honda and Hayashi

1995). Commonly, mixtures of stimulant compounds found within host plants can act in

21

combination to synergistically elicit oviposition in swallowtail butterflies (Ohsugi et al.
1985, Feeny et al. 1988, Nishida and F ukami 1989, Honda 1990, Sachdev-Gupta et al.
1993). The stimulant mixture generally contains compounds ubiquitous to many plant
families as well as those unique to a particular species (Nishida 1995).

Behavioral reliance on compounds characteristic of a particular plant taxa may
account, in part, for the ovipositional specificity of Oligophagous swallowtail species
(Ohsugi et al 1985, Feeny et al 1988, Nishida and Fukami 1989, Honda 1990, Sachdev-
Gupta et a1 1993). For example, the oviposition stimulants of the swallowtails
Atrophaneura alcinous and Battus philenor were characterized as a mixture of
aristolochic acids and various sugars. Aristolochic acids are found predominantly in
plants within the Aristolochia (Aristolochiaceae), and their indispensability in eliciting
oviposition probably accounts for the sole use of plants in that genus by the two species.
A physiological dependence on a few compounds exclusive to the family Lauraceae for
oviposition stimulus may underlie the strong host recognition of P. troilus and P.
palamedes both in the field and in multi-choice bioassay arenas in the lab. With that
rationale, I focused on the chemical extraction and fractionation of oviposition stimulants
found in spicebush and redbay, the respective favorite host plants of P. troilus and P.

palamedes.
Materials and Methods
I ° ro

Field-caught or hand paired, lab-reared females were stored at 18°C until their use

22

in the bioassay. Storage at that temperature increased the lifespan of the females relative
to butterflies maintained at room temperature. Each bioassay arena consisted of 1 round,
transparent plastic container (26 cm in diameter x 9 cm) lined with a paper towel at the
bottom. Oviposition substrates, consisting of either filter paper or non-host leaves treated
with host extracts, were placed along the sides of the container equally spaced from each
other. The petioles of the leaves were inserted into waterpics to maintain leaf turgor. A
single female occupied each bioassay arena, where she encountered all treatments and the
control substrate simultaneously. Bioassay containers were stacked on turntables that
rotated once every six minutes and were illuminated on one side with incandescent lights
(see Scriber 1993). Once a day, egg counts on the oviposition substrates were tallied and
fresh treatments and controls were supplied. Eggs laid greater than 2.5 cm away from a
substrate (the approximate distance between a female's foretarsi and curled ovipositor), or
equal distances between two treatments, were scored as "container." Females were fed a
20% honey solution once daily. Females were allowed to oviposit for several days, and
analysis was performed on the percentage of total eggs laid on each substrate over the

course of the bioassay.

Mm:
E ' i ' ' h

In 1994, fresh, intact spicebush leaves (130.7 grams) collected from Kalamazoo
county, Michigan were extracted for 1 minute with chloroform (3x), then with methanol

(3x) in an extraction column. The procedure yielded 0.8 g of chloroform extract and 8.3

23

g of methanol extract.

A modified extraction procedure was used in 1995. Freeze-dried, ground
spicebush leaves (120.5 g) collected from Gallia County, Ohio were extracted
sequentially with 3 solvents. The leaf powder was extracted first with 700 ml of hexane
for 5 minutes, then soaked (> 6 hours, 2x) in 700 ml of hexane. The procedure was
repeated with ethyl acetate and methanol. The residues were dried in vacuo at 40°C.
Subsequent procedures focused on the methanol extract, which yielded 20.3 grams of
dried powder. An aliquot (10.2 g) of the methanol extract was redissolved in 200 ml of
20% methanol in a flask during agitation with a stir bar. The solution was extracted (3x)
with 50 ml of ethyl acetate, distributing 8.0 g of material into the aqueous phase and 2.1 g
into the ethyl acetate phase. The ethyl acetate—soluble material was dried under a vacuum
at 40°C, and the aqueous phase was stored frozen at -20°C until use in the bioassays.

Ex ' i ti b

In 1995, 72.9 g of freeze-dried, milled redbay leaves collected from Highland
County, Florida were extracted with ca. 400 ml of each solvent, following the protocol
for the 1995 extraction of spicebush. The procedure yielded 16.4 g of vacuum-dried
methanol extract. The crude methanol extract (5.7 g) was partitioned between water and
ethyl acetate as above, distributing 1.1 and 4.6 g into the organic-phase and aqueous-
phase, respectively.

An additional 56.3 g of the redbay leaf powder stored at -80°C was extracted later
in the summer to give 13.7 g of methanol extract. A portion of this material (6.9 g) was

partitioned as above between ethyl acetate and water, yielding 1.6 and 5.3 g of material,

24

respectively.

Bigassaxs:
41119ng filter paper biggssgy (figure 22

A leaf area meter was used to measure the average surface area of 65 freshly
picked spicebush leaves weighing a total of 22.4 grams. Aliquots of each extract from
1994 (average g extract/cmz) were redissolved in their original extraction solvent and
pipetted onto filter paper triangles of known surface area. The solvent was allowed to
evaporate off of the paper strips before their inclusion in the bioassay. Four filter paper
triangles, treated with either the methanol extract, the chloroform extract, both extracts,
and solvent only were placed in each bioassay arena.
5- ' t r er i w' h i ' r

Using spicebush extracts and fractions of leaf samples from 1995, gram leaf
equivalents (average gram extract/leaf, hereafter abbreviated GLE) of the crude methanol
extract and the water fraction were sprayed evenly onto both sides of filter paper triangles
using a chromatographic sprayer. Approximately four times the OLE of the ethyl acetate
fraction was misted onto the filter paper triangles due to a calculation error. However,
this mistake did not significantly alter the results of the bioassay (see Figure 4A). Within
one oviposition chamber, each female could contact all 5 oviposition substrates: 3 filter
paper triangles (coated respectively with the crude methanol extract, the ethyl acetate
fraction and the water fraction) an untreated filter paper triangle, and a spicebush leaf

with its petiole inserted into an aquapic to maintain leaf turgor.

25
-c'n-hia'ss 're 5
GLE of the crude extracts and fractions were sprayed separately onto both sides of
tulip tree leaves using a chromatographic sprayer. Tulip tree leaves were chosen as a
substrate because it is not a host plant of P. troilus or P. palamedes and it was thought
that recognition of a leaf surface would enhance the accuracy of oviposition on the
treatment choices. The 3-choice leaf bioassay with P. troilus consisted of a tulip tree
treated with the 1994 methanol extract, an untreated tulip leaf, and a spicebush leaf. In
the 5-choice bioassay, oviposition preference of P. troilus was tested on the 1995 crude
methanol extract, the ethyl acetate and water fractions and untreated tulip tree and
spicebush leaves. In the case of P. palamedes, redbay extracts and fractions were used,
and 1 or 2 redbay leaves collected from plants in the greenhouse served as the positive

control instead of spicebush.

Statistical analysis:

The percentage of eggs laid by individual females on each treatment throughout
the duration of the bioassay was calculated. The percentages were arcsine transformed
prior to their analysis to normalize the variances. A standard two-way analysis of
variance (ANOVA) for unreplicated data was performed, with treatments and butterflies
as the main effects (Sokal and Rohlf 1981). The term "unreplicated" refers to the fact that
only 1 butterfly occupied each block, or bioassay arena. If treatment was found to be
significant, a one-way ANOVA was performed on the arcsine transformed percentages,

excluding the “container” category to maintain independence between treatments.

26

Specific significant differences between treatments could then be characterized with the

Student-Newman-Keuls test (SuperANOVA 1993).

Results

In the 4-choice bioassay with P. troilus (Figure 2), two-way ANOVA showed
significant differences between treatments (F4 = 6.60, p = 0.0015), but not butterflies (F9
= 0.01 , p = 1.0). Filter-paper triangles coated with the combined chloroform + methanol
extracts had high ovipositional activity, receiving 46% of all eggs laid, significantly more
than the control and the other two treatments according to a one-way ANOVA (p < 0.01,
Figure 1). Filter paper triangles coated with either the methanol or chloroform extract
alone were not found to be significantly different from the control. Twenty-four percent
of the eggs were scattered randomly around the oviposition container, and were included
in the “container” category.

The 3-choice leaf bioassay (Figure 3) demonstrated the effect of the cues from the
oviposition substrate on ovipositional selectivity. Only 5% of the eggs were laid off of
the intended substrates on the sides and bottoms of the assay chamber. In the 2-way
ANOVA, just the treatment category was found to be significant (F2 = 14.051, p <
0.0001). A 1-way ANOVA revealed that P. troilus females laid significantly more eggs
on tulip tree leaves treated with the methanol extract (48%) and intact spicebush leaves
(36%) than tulip tree leaves sprayed with methanol solvent only (p < 0.01).

In the presence of an intact spicebush leaf, P. troilus female laid negligible

amounts of eggs on filter paper triangles sprayed with spicebush extracts in the 5-choice

27

assay (Figure 4A). 89% of the eggs were laid on the spicebush leaf, while significantly
less (9%) were laid on the other filter paper substrates together (p < 0.01, 1-way
ANOVA). In contrast, tulip tree leaves sprayed with spicebush extracts proved more
stimulatory than both the tulip tree leaf control and the spicebush leaf itself (Figure 4B).
Only 10% of the eggs were placed on spicebush compared to 30% on leaves containing
the crude methanol extract, 30% on leaves treated with the ethyl acetate fraction, 18% on
leaves treated with the aqueous fraction and 12% on the untreated tulip tree leaves. A
one-way AN OVA revealed that the crude methanol extract and the ethyl acetate fraction
received significantly more eggs than did the untreated spicebush and tulip tree leaves (p
< 0.05).

In the case of P. palamedes as well, tulip tree leaves sprayed with the crude
redbay extracts received more eggs than redbay leaves (Figure 5). Only the crude
methanol extract was found to besignificantly different from the untreated redbay and
tulip tree controls by a one-way ANOVA (p < 0.01). In both P. troilus and P. palamedes
tulip tree assays, the butterflies laid comparable percentages of eggs on the untreated tulip
tree leaves and their preferred host leaf, and laid eggs with greatest frequency on tulip
tree leaves coated with host extracts. The results of all the bioassays demonstrate that

host plant chemicals stimulate oviposition by P. troilus and P. palamedes.

Discussion
Leaf chemistry clearly plays a role in host recognition by P. troilus and P.

palamedes. In the P. troilus four-choice filter paper bioassay (Figure 2), the

28

chloroform and methanol extract-treated filter paper triangles received more eggs than
the untreated filter paper. The filter paper treated with both extracts received 46% of
the eggs laid, significantly more than any other substrate, due to either a synergism or
increased dosage of the oviposition stimulants. Results from the leaf bioassays
(Figures 3, 4B and 5) also strongly implicate a role for chemical stimulants in
oviposition. Tulip tree leaves sprayed with the methanolic extract of the preferred host
and its fractions received significantly more eggs from the respective Papilio species
than the untreated tulip tree leaf and even the host leaf itself.

However, chemical cues are not solely responsible for host recognition. In the
P. troilus four-choice filter paper bioassay, a significant percentage of eggs was
distributed around the bioassay arena, away from the filter paper substrates. A lack of
'cues vital to the recognition of a host surface may account for this seemingly
indiscrirninant dispersion of eggs. Off-substrate oviposition declined substantially in
the 3-choice and 5-choice all-leaf assays, most likely due to the perception of general
leaf cues such as color, texture and odor that characterizes host tissue (Figures 43 and
5). A comparison of the two 5—choice bioassays with P. troilus (Figure 4) dramatically
demonstrates the importance of physical cues in host recognition. In the presence of an
intact spicebush leaf, P. troilus females laid negligible amounts of eggs on extract-
treated filter paper. However, when sprayed onto the surface of tulip tree leaves, P.
troilus females preferred the extracts to the spicebush leaf itself.

The role of physical cues in host recognition was demonstrated by Stadler

(1974) who found that the eastern spruce budworm, Choristoneura fitmiferana,

29

preferred to oviposit on paper in the form of coniferous needles to flat sheets of the
same material. Grant and Langevin (1994) also have shown that although conifer-
feeding Choristoneura Species laid significantly more eggs on the hexane extract-treated
filter paper compared to the control, they preferred to oviposit on fake conifer foliage
to extract-treated filter paper. Pieris brassicae demonstrates “wavelength specific
behavior,” by flying down and drumming on discs illuminated by light only in the 522-
554 nm (yellow—green to green) range (Kolb and Scherer 1982). Similarly, Saxena and
Goyal (1978) found that Papilio demoleus butterflies showed a positive orientational
response to the glass-screened leaves of both lime (host plant) and cotton (non host)
relative to the blank wall on the opposite side of their cage. Since the leaves of the two
plants differ in shape, the butterflies may be attracted to general plant architecture or
leaf color. Using muslin patches colored with various dyes, Saxena and Goyal ( 1978)
then demonstrated a maximum attraction for green or yellow-green colors. In my
experiment, filter paper triangles treated with methanol or ethyl acetate extracts had a
vibrant green color, perhaps cuing oviposition in the four-choice filter paper assay.
Since both P. troilus and P. palamedes laid approximately equal amounts of eggs on
the untreated tulip tree leaf and their true host leaf, shape does not appear to be a
distinguishing cue at close proximity.

A gestalt of visual and chemical cues greatly boosts the oviposition response in
comparison with either cue by itself (Saxena and Goyal 1978, Traynier 1986). Ground
disks illuminated with light of yellow-green wavelengths could elicit drumming

behavior of Pieris brassicae, but never oviposition. Disks sprayed with sinigrin, an

3O

oviposition stimulant of P. brassicae, and illuminated with white light, were
completely ignored. However, when both stimuli were present together, significant
oviposition occurred (Kolb and Scherer 1982). Similarly, moistened yellow-green
muslin treated with an ethanol extract from the Citrus host plants elicited more visits,
abdominal curling, and oviposition by Papilio demoleus than either stimulus alone.
The addition of host odor in the form of an ether extract further increased all three
behavioral responses (Saxena and Goyal 1978). .

Tactile cues also play a role in host acceptance, particularly in moths
(Ramaswamy 1994). In my experiments, the chloroform extract—coated and chloroform
+ methanol extract coated-filter paper may have had a more waxy texture than either
the untreated or methanol extract-treated filter paper, as leaf surface waxes are soluble
in chloroform (Figure 1). A smooth texture in combination with chemical stimulants
found in either extract may have caused more eggs to be distributed on the triangles
treated with both extracts. However, Shorey (1964) found that T richoplusia ni did not
lay significantly more eggs on wax paper than blotting paper, although the proportion
of eggs deposited on each substrate changed with the relative humidity. Oviposition
stimulants dissolved in the wax layer (Bemays and Chapman 1994, Renwick and
Huang 1994) could also account for oviposition on the chloroform extract-treated filter
paper.

It is intriguing that both P. troilus and P. palamedes laid significantly more
eggs on tulip tree leaves sprayed with certain host extracts than on their actual host

leaf. It is possible that extracts applied to the surface of the non-host leaf have an

31

increased functional dosage due to their accessibility to the tarsal chemoreceptors,
unmasked by a surface layer of wax. In the confines of the plastic container, saturation
of the air by leaf volatiles would also make it difficult for the butterflies to rely on
olfactory cues to guide them to their true host leaf. Alternatively, a synergism between
tulip tree leaf chemistry and the host plant extracts may have occurred.

An idiosyncracy of both of the 5-choice leaf bioassays, but not the 3—choice
bioassay, was that the tulip tree control leaves received as many eggs as did the
lauraceous host leaves. The redbay leaves taken from potted trees in the greenhouse
were yellowish and of poor quality and may explain this unusual trend in the bioassay
with P. palamedes. Altemately, the phenomenon may be due to “spillover”
oviposition from neighboring extract coated leaves. The resinous extracts applied to
the surface of the tulip tree leaves may have adhered to the tarsi of the butterflies. The
large, continuous leaf surface area of the tulip tree leaf coupled with the closer
proximity of all the leaves in the 5-choice assays may have exacerbated the problem of
"spillover" oviposition in these arenas. Additional bioassay errors may have resulted
from chemical changes in the tulip tree leaves due to their removal from trees, the
synergism of tulip tree chemistry with that of the host plant extracts and altered
permeability of the tulip tree leaves due to solvent effects. However, in other
bioassays that used both a solvent-treated and untreated control leaf (see Chapter 2), no
significant differences between the two were detected.

Although several factors are responsible for the recognition of a substrate as a

host, as evidenced by the results here, leaf chemistry does dictate inter- and intra-

32

specific oviposition decisions (Myers 1985, Nishida and Fukami 1989, Feeny 1991,
Roessingh et al. 1991, Sachdev-Gupta et al. 1993, Honda and Hayashi 1995).
Reliance on compounds unique to particular plants may keep related butterfly taxa
faithful to chemically similar groups of plants. Elucidation of oviposition stimulants
can permit parsimonious deductions about taxonomic relationships within the genus

Papilio and their host affiliations past and future.

Mean % eggs

33

60

 

 

 

     

 

b
50 .-
4o -_
23.6 :t 5.9
30
ill
20 list? 1
10 -t
a
o
a 5 ‘L
S a a g 2
9 £6 9.’ £3
2 E r:
5 E = 8
Treatment

Figure 2. Percentage of eggs (mean :1: SE.) laid by Papilio troilus females
(n = 6) on each of 4 choices: filter paper triangles coated with methanol
extract (methanol), chloroform extract (chloroform), a combination of both
extracts (both) and an untreated filter paper triangle (untreated). Extracts
were made from fresh spicebush leaves. Significant differences between
means (p < 0.01) are shown for arcsine transformed percentages.

*Container refers to the percentage of eggs (mean :1: S. E.) laid off the filter
paper substrate onto the sides and bottoms of the bioassay arena. It was
excluded from the l-way ANOVA and SNK test that determined the

significance differences shown here.

34

60

 

Mean % eggs

 

 

 

Spicebush MeOH MeOH Container‘
extract , solvent
Treatments

Figure 3. Percentage of eggs (mean :t S. E.) laid by Papilio
troilus females (n=10) on each of three choices: an untreated
spicebush leaf, a tulip tree leaf coated with the methanol extract
of spicebush and a tulilp tree leaf sprayed with methanol only.
Significant differences between means (p < 0.01) are shown for
arcsine transformed percentages.

35
100

 

90 ~—
80 +

60 —~
50 -.
4o ——

Mean % eggs

30 ~~
20 +

10 4-
b

 

 

Spicebush MeOH EtOAc Aqueous

b 2.3 :1: 1.0
t L
solvent container

 

 

40
B
35 «— b

Mean % eggs

 

      

Spicebush MeOH EtOAc Aqueous
Treatments

0.6 :l: 0.4

 

   

Tulip tree Container

Figure 4. Percentage of eggs (mean S. E.) laid by Papilio troilus females on each of
5 choices: a spicebush leaf, substrate sprayed with crude methanol extract (MeOH),
substrate sprayed with the ethyl acetate fraction (EtOAc), substrate sprayed with the
aqueous fraction (Aqueous) and an untreated substrate. Significant differences
between means (p < 0.05) are shown for arcsine transformed percentages. A. The
substrates are filter paper triangles (n = 32 females). B. The substrates are tulip tree

leaves (n = 16 females).

40

Mean % eggs

36

 

 

 

10.3 :t 4.7

     

> I o a) 4L
‘3 o g 8 E 2
'O 0 ._, cu ._
m E LIJ 3 9 9
tr 2' '5 g
'— 0

Treatment

Figure 5. Percentage of eggs (mean :I: S. E.) laid by Papilio palamedes
females (n = 9) on each of 5 choices: an untreated redbay leaf, a tulip tree
leaf sprayed with methanol extract, a tulip tree leaf sprayed with the ethyl
acetate fraction, a tulip tree leaf sprayed with the aqueous fraction, and an
untreated tulip tree leaf. Redbay leaves were taken from potted greenhouse
trees and were yellow and of poor quality. Significant differences between
means (p < 0.05) are shown for arcsine transformed percentages.

 

CHAPTER 2
Chemical extracts from a non-host, Persea borbonia, deter oviposition by Papilio
glaucus, a polyphagous swallowtail
Abstract
The ability to perceive and respond to phytochemicals that reliably indicate the

suitability of a potential host plant confers a selective advantage to ovipositing female
swallowtail butterflies. Papilio glaucus females do not oviposit on redbay (Persea
borbonia: Lauraceae), a non-host that does not support larval development and is
commonly found in habitats alongside their'principal host plant, Magnolia virginiana, in
central Florida. I tested the hypothesis that deterrent compounds present in the leaves of
redbay mediate its rejection by P. glaucus. Florida populations of P. glaucus did not
oviposit on host leaves sprayed with the methanol extract of redbay foliage, although they
accepted solvent treated and untreated tulip tree leaves in 3 choice bioassays.
Additionally, tulip tree leaves sprayed with methanolic extracts of redbay also deterred
oviposition by P. glaucus females from Ohio, Kentucky and Pennsylvania, although these
populations do not naturally encounter redbay. Clearly, deterrent compounds found
within this non-host are the basis of its rejection by P. glaucus. The evolutionary

scenarios may that may account for this behavior are addressed.

37

38

Introduction

Papilio glaucus, Papilio palamedes and Papilio troilus occupy the same habitat
patches in central and southern Florida. Although P. glaucus will oviposit and feed on
plants from over a dozen plant families (Scriber et al. 1991), sweetbay (Magnolia
virginiana) is the only host plant found in the region (Scriber 1986, see Table 4). P.
palamedes and P. troilus oviposit and feed exclusively on redbay (Persea borbonia), as
the other potential lauraceous hosts, sassafras and spicebush, do not extend that far south
(N itao et a1. 1991). Redbay and sweetbay grow in intermixed stands in the hammock
habitat, often with interdigitating branches, and all three swallowtail species are common
in those areas when nectar sources are in bloom. P. glaucus does not oviposit on redbay
(Scriber unpubl.), and the foliage is toxic to the larvae (Scriber et al. 1991).

Several hypotheses have been suggested to account for the vast host breadth of
polyphagous Lepidoptera such as P. glaucus. Both Oligophagous and polyphagous
Lepidoptera can detect and process a variety of chemical stimuli through their peripheral
nervous system via chemoreceptive sensilla (Blaney and Simmonds 1990, Roessingh et
al. 1991, Schoonhoven et al. 1992). The central nervous system of the polyphagous
butterfly may simply recognize a broader range of chemical stimuli as “host” than
Oligophagous butterflies. Altemately, the avoidance of non-hosts may rest on the
presence of deterrents. Both hypotheses are not mutually exclusive, and a combination of
these mechanisms may increase ovipositional accuracy. Often, the relative ratio of
compounds interpreted as stimulatory and deterrent determines the host range of a

butterfly species (Renwick and Huang, 1994; Honda and Hayashi, 1994). I predict that P.

39

glaucus rejects redbay due to the presence of deterrent compounds in the leaves.

Several studies have shown that volatiles and non-volatile compounds found in
non-host foliage deter landing and oviposition (F eeny et al. 1989). For example, volatiles
from cabbage, a non-host, inhibited Papilio polyxenes from landing on cellulose sponges
treated with host plant extract (F eeny et al. 1989). Electrophysiological studies with P.
polyxenes have shown that a single neuron in the tarsal chemoreceptors responds in a
characteristic deterrency pattern to polar cabbage extracts. The action potentials
generated by the deterrent neuron differ in amplitude and time period of response from
those generated by the 2 neurons that respond to the stimulatory host plant extract
(Roessingh et al. 1991). The differential coding of deterrent and stimulatory sensory
input may provide the basis for the distinct behavioral responses.

The interpretation of certain compounds as deterrent may have evolved as a
means of avoiding plants toxic to the larvae. The deterrent compound itself may be toxic,
or reliably indicate the presence of other toxic compounds (Dethier 1980, Bemays and
Cornelius 1992). Neural recognition of such compounds from the foliage of non-hosts
may confer a selective advantage in habitats where hosts and non-hosts occur
sympatrically. The ability of populations of P. glaucus from southern Florida to
recognize deterrent compounds in redbay, a non-host found commonly in the same
habitat as sweetbay, was tested. Northern populations of P. glaucus, outside the range of
redbay, were also tested to see if they have the ability to recognize deterrents from a non-

host they never encounter.

40

Materials and Methods
E ' r db rse r oni .'

In 1995, 72.9 g of freeze-dried, milled redbay leaves collected from Highlands
County, Florida were extracted once with ca. 400 ml of hexane for 1 minute, and then
extracted with 400 ml of hexane for two 6 h periods in an extraction column with
stopcock drain. The procedure was repeated with ethyl acetate and methanol to yield 1.1,
1.2 and 16.4 g of hexane, ethyl acetate and methanol crude extracts, respectively.

A second batch of intact redbay leaves stored at -20°C from the spring of 1995
were extracted in 1996 according to the above protocol with 1200ml of each solvent. The
procedure yielded 0.3, 1.7 and 5.7 g of hexane, ethyl acetate and methanol extract,
respectively. An aliquot of the methanol extract (1.9 g) was redissolved in 20 ml of
methanol and 130 m1 of water and extracted 3x with 50 ml of ethyl acetate. The ethyl
acetate and water fractions contained 0.3 g and 1.6 g of material, respectively. The ethyl
acetate fraction was dried in vacuo, and the aqueous extract was stored at -15°C until use

in the bioassay.

r i e i
Each gravid female was placed separately in a round, plastic container (26 cm
diam. x 9 cm) lined on the bottom with a paper towel. Redbay extracts were dissolved in
their parent solvent and sprayed with a chromatographic Sprayer onto both sides of tulip
tree leaves. The leaf petioles were placed in a water-filled aquapic to maintain leaf

turgor. Within each bioassay container, the female could contact all treatment and

41

control leaves. The tulip tree leaves were harvested the day of their placement in the
bioassay from either potted tulip trees or a tulip tree growing on the confines of Michigan
State University. The leaves were rinsed with distilled water and dried before application
of extracts and placement in the bioassay. Throughout the bioassay, the butterflies were
fed a 20% honey solution once daily while eggs were tallied and the old leaves replaced.
Bioassay containers were stacked on turntables that rotated once every six minutes and
were illuminated one side with incandescent lights (see Scriber 1993). Eggs laid greater
than an inch away from a substrate (the approximate distance between a female’s foretarsi
and curled ovipositor), or equal distances between two treatments, were scored as
“container.” The bioassay continued over several days, and analysis was performed on
the percentage of total eggs laid by each female on each substrate over the course of the

bioassay.

B ' it u r'

First generation offspring of P. glaucus females collected from Florida in March
1995 were reared on sweetbay (M. virginiana). After eclosion, the female butterflies
were stored at 18°C until they were hand-paired and placed in the bioassay 24 hours after
mating.

Three bioassays were employed. The first bioassay tested oviposition preferences
between three choices. The "all" treatment consisted of gram leaf equivalents of all three
extracts (hexane, ethyl acetate and methanol) dissolved together in a mixture of the three

solvents and sprayed onto a tulip tree leaf. A mixture of equal proportions of hexane,

IS

42

ethyl acetate and methanol was sprayed onto a leaf to form the "solvent" treatment. The
control category consisted of a completely untreated tulip tree leaf. Therefore, each
female could choose simultaneously between the "all," "solvent" and "control" leaves.

The second bioassay design was composed of 5 choices. The hexane, ethyl
acetate and methanol extracts were dissolved individually in their respective solvents and
sprayed on separate tulip tree leaves. The three treatments were placed in each arena
along with a "solvent" leaf (treated as above) and the "control" leaf.

The third bioassay consisted of each extract (hexane, ethyl acetate and methanol)
assayed individually with a “solvent” leaf (the same solvent used to redissolve the

extract) and "control" leaf for a total of 3 choices.

't r h r ti 1

P. glaucus pupae from Ohio, Kentucky and Pennsylvania, reared from females
collected in 1995, were removed from cold-storage (4°C) and allowed to eclose. Adult
female P. glaucus were stored at 18°C until they were hand-paired to males (field
captured Papilio canadensis or lab-reared P. glaucus).

Results from the previous studies (1995) indicated that only the methanol extract
inhibited oviposition by Florida P. glaucus. Therefore, only the methanol extract was
tested with northern populations.

In the first bioassay, 5 females were each placed in a separate bioassay arena
containing an untreated tulip tree leaf, a tulip tree leaf sprayed with methanol, and a tulip

tree leaf sprayed with the methanol extract dissolved in methanol. The first day of the

43

bioassay, approximately 0.04 g of extract were sprayed onto each treatment leaf. The
next 2 days the dose was reduced to 0.02 g-leaf“, and for the duration of the bioassay

0.01 g-leaf“ was used. Although I cannot account for antagonistic interactions at
chemosensory receptor sites, I wanted to be sure that the extract layer did not serve as an
impermeable physical barrier to the perception of tulip tree chemistry. By reducing the
dose, I effectively reduced thickness of the redbay extract layer on the surface of the tulip
tree leaf. Reduction of the dose did not affect the results of the bioassay. Butterflies
were allowed to oviposit until their death, and the lifetime percentage of eggs laid on each
leaf was calculated.

The second bioassay consisted of 4 choices. The ethyl acetate extract was
redissolved in ethyl acetate and sprayed onto tulip tree leaves. The aqueous fraction was
removed from the freezer, thawed, and sprayed directly onto tulip tree leaves. The
amounts of both extracts used was their gram yield proportional to the amount of crude
methanol extract partitioned between the two solvents (1.9 g), multiplied by 0.01 (the
dose cf the crude methanol extract). Controls consisted of a leaf sprayed with ethyl

acetate and an untreated leaf. Nine butterflies were allowed to oviposit until their death.

The percentage of eggs laid by individual females on each treatment throughout
the duration of the bioassay was calculated. The percentages were arcsine transformed
prior to their analysis to normalize the variances. A standard two-way analysis of

variance (ANOIA) for unreplicated data was performed, with treatments and butterflies as

44

the main effects (Sokal and Rohlf 1981). The term "unreplicated" refers to the fact that
only 1 butterfly occupied each block, or bioassay arena. If treatment was found to be
significant, a one-way ANOIA was performed on the arcsine transformed percentages,
excluding the “container” category to maintain independence between treatments.
Specific significant differences betweentreatments could then be characterized with the

Student-Newman-Keuls test (SuperANOVA 1993).

Results
Fl r' l ' n P ili lauc 5
I In a three-choice bioassay, P. glaucus females avoided tulip trees sprayed with the

combination of all three extracts (Figure 6). A two-way AN OIA on arcsine transformed
percentages revealed a significant treatment effect (F. = 19.44, p < 0.0001), but not a
significant butterfly effect (FB = 0.003, p = 1). A one-way ANOIA confirmed the
significant treatment effect (F2 = 19.09, p < 0.0001). The mean percentage of eggs laid
on “all” was significantly lower than that of “solvent” (p < 0.01) and “control” (p <
0.01). “Solvent” and “control” were not significantly different from each other.

Two-way AN OIA performed on the arcsine transformed percentages from the 5-
choice oviposition trials with P. glaucus from Florida (Figure 7) revealed treatment (F5 =
4.69, p = 0.0014), but not butterflies (F,() = 0.11, p = 0.9997), to be a significant main
effect. One-way ANOIA similarly found a significant difference between treatment
means (F4 = 3.97, p = 0.0072). Significantly more eggs were laid on leaves treated with

the hexane extract than with the methanol extracts (p < 0.01). Although not detected to

45

be significantly different, fewer eggs were laid on methanol extract-coated leaves than
“solvent” and “control” leaves.

The percentages of eggs laid did not differ between treatments or butterflies in the
three choice bioassays with the hexane extract (Figure 8) according to two-way and one-
way ANOIA. The butterflies tended to lay eggs equally on all three choices. A
significant difference between treatments was found by two-way ANOIA (F3 = 5.72, p =
0.018), but not one-way AN OIA, in the three choice bioassay with ethyl acetate extract
(Figure 9) with Florida P. glaucus. Variation between treatments may be attributed to an
increase in the percentage of eggs laid on leaves sprayed with ethyl acetate extract
relative to the “solvent” and “control” treatments. However, due to the small sample size
(n = 4) the results should be interpreted with caution.

Significant differences between treatments, but not butterflies, was revealed by
two-way ANOIA in three-choice bioassays (Figure 10) between leaves sprayed with
methanol extract, “solvent” leaves and “control”leaves (F2 = 10.48, p = 0.0084). One-
way AN OIA similarly detected a significant difference between treatments (F2 = 10.74, p
= 0.014). The butterflies laid significantly fewer eggs on leaves sprayed with methanol
extracts of redbay than on “solvent” or “control” leaves (p < 0.05).

The percentages of eggs laid on “solvent” leaves was not significantly different
from the percentages laid on “control” leaves in any of the trials.

or. e 'n o '41'1'032 . Ill 0 ' .1 a In! " n vue ,"0

Like the populations of P. glaucus from Florida, P. glaucus from Ohio, Kentucky

46

and Pennsylvania deposited fewer eggs on tulip tree leaves coated with methanol extract
than on untreated tulip tree leaves and tulip tree leaves sprayed with solvent (Figure 11).
A two-way AN OIA revealed a significant treatment effect (F2 = 112, p < 0.0001), but not
a significant butterfly effect (F4 = 0.092, p = 0.982). One-way ANOIA similarly revealed
a significant treatment effect (F2 = 161, p < 0.0001). A significantly lower percentage of
eggs were laid on leaves coated with methanol extract than on “solvent” or “control”
leaves (p < 0.01).

In 4-choice bioassays with the aqueous and ethyl acetate fractions of the active
methanol extract (Figure 12), two-way AN OIA revealed only a significant treatment
effect (F3 = 79.94, p < 0.0001), as did the one-way ANOIA (F, = 106.09, p < 0.0001).
Leaves sprayed with the aqueous and ethyl acetate fractions received significantly fewer

eggs than the “solvent” and “control” leaves (p < 0.01).

Discussion

The presence of deterrents in the leaves of non-hosts is a common phenomena. A
significantly larger number of compounds inhibit feeding and oviposition than stimulate
those responses (Bemays and Chapman 1994). Plants that possess deterrent compounds
may actually prove adequate for development when consumed by larvae whose gustatory
sense has been removed. Therefore, the evolution and maintenance of deterrent
compounds may enable a plant to repel unwanted herbivores (Dethier 1980). Perception
of deterrent compounds that truly signal toxicity may also benefit the herbivores who

avoid oviposition and consumption of poisonous foliage (Dethier 1980).

47

Females of the polyphagous P. glaucus do oviposit on and the larvae consume
certain species of Lauraceae, such as sassafras and spicebush, albeit with lower resulting
survivorships than with more favored host plants such as tulip tree and black cherry
(Scriber et al. 1991). However, P. glaucus caterpillars fail to survive on redbay (Scriber
et al. 1991), and adults do not oviposit on this tree (see Table 4). This suggests that
deterrent compounds in redbay accurately reflect the presence of toxins, and the ability of
P. glaucus to detect the deterrents benefits this species. These results also suggest that
redbay is perceived by P. glaucus to be more chemically different from spicebush and '
sassafras than they are from each other, at least in terms of ovipositional activity.
Perhaps redbay foliage contains different, more potent toxins, or the same compOunds are
present in higher quantities in redbay than in spicebush and sassafras.

It is interesting that northern populations of P. glaucus that never encounter
redbay still respond to deterrents present in foliar extracts. It is possible that redbay
compounds have inherent deterrent properties to P. glaucus and the ancestor of the
glaucus-troilus lineages (see Figure 1). P. palamedes may have adapted to efficiently
utilize redbay foliage during its divergence from the glaucus group. Altemately, P.
glaucus may have evolved the ability to perceive and respond to compounds found in
redbay. P. borbonia may have been among those species inhabiting the Midwestern
United States that were pushed southward toward southern Florida and Texas by global
cooling (Scriber 1988). The divergence of the glaucus and troilus groups may have
occurred before this period of global cooling, and the detection and interpretation of

redbay chemistry as deterrent may have resulted from frequent encounters by P. glaucus

48

unable to detoxify the foliage. The relative reluctance of northern populations to oviposit
on sassafras and spicebush may be attributed to the legacy of ovipositional inhibition by
Persea chemistry mediated by compounds of similar or identical structure.

The detection of deterrents enables P. glaucus to avoid ovipositing on non-host
foliage. Swallowtail butterflies normally fly away immediately after alighting on non-
host leaves, in contrast to their continued sampling and drumming of host leaves
(Darnman and Feeny 1988, Rausher 1995). Olfactory and gustatory perception of
deterrent compounds may provide butterflies with an efficient mechanism for evading
non-host foliage, especially in areas where the growth forms of host and non-host trees
intermingle (Feeny et al. 1989, Roessingh et al. 1991) as is the case in the Hammock
habitat of central Florida. Sensitivity to strongly deterrent compounds in toxic plants
may play an important in host location by polyphagous species, who are neurally wired to
accept a greater number and taxonomic variety of species (Bemays and Chapman 1994).

The acceptance or rejection of a plant by ovipositing polyphagous butterfly
species, such as P. glaucus, is probably not dictated solely by the presence or absence of
deterrent chemistry (Dethier 1980). Rather, acceptance of a subset of compounds more
ubiquitous among plant taxa may account for use of plants from many families.
Additionally, evidence exists that the balance of cues from a variety of senses, integrated
by the central nervous system, determine the acceptability of a substrate for oviposition
(Miller and Strickler 1984, Ramaswamy 1994). For example, Traynier (1986) discovered
that when Pieris rapae females are presented with 2 discs of a different shade of green

and a blue disc, each bearing a different concentration of oviposition stimulant, the

49

females invaryingly placed more eggs on one of the green discs, although the blue disc
may have received the higher dose of the stimulant. The green color plus the
intermediate concentration of the stimulant may have more closely approximated the
"average" sensory input required to stimulate oviposition (Traynier 1986). Nevertheless,
in a scenario where many sources of external and internal stimuli determine the
acceptability of a host, oviposition deterrents may serve to tip the balance of sensory cues

toward unacceptability, causing females to reject a potentially unsuitable host.

50

Table 4. Percentage of eggs laid by 5 Papilio glaucus females from Highlands County,
Florida in a 3-choice bioassay on leaves from trees found in their habitat. Sweetbay, the
primary host plant, and redbay, a non-host, are abundant, while ash is rarely encountered
(Scriber, unpubl.).

 

 

 

 

 

 

 

Papilio glaucus Sweetbay Redbay Ash eggs (n)
9 2 from Florida Magnolia virginiana Persea borbonia Fraxinus trifoliata

1 74.7 1.3 24.1 79

2 70.1 9.3 20.6 107

3 57.1 0 42.9 42

4 66.7 . 0 33.3 12

5 57.1 0 42.9 42

 

 

 

 

Mean :b S. E. 65.4 :t 3.5 2.7 :t 1.7 32.8 :t 4.6 282

 

 

Mean % eggs

60

50 ..

4o .»

30 .-

20 ~~

1o ——

51

 

6.2 :1: 2.2

 

 

     

 

All Solvent Control Container

Treatments

Figure 6. Percentage of eggs (mean :t S. E.) laid by Papilio glaucus females
from Florida (n=14) on each of three choices: tulip tree leaves sprayed with a
combination of the hexane, ethyl acetate and methanol extracts (all), tulip
tree leaves sprayed with a 1:1:1 ratio of each solvent (solvent), and an
untreated tulip tree leaf (control). Significant differences at p < 0.01 are
shown for arcsine transformed percentages.

*Container refers to the mean percentage of eggs laid off the treatment and
control substrates, and is excluded fiom analysis by l-way AN OVA and the
SNK test. The mean :1: S. E. are shown above. '

Mean % eggs

40

35

30 a.

25 —~

20 -_

15-

1o -.

52

 

 

 

 

Hexane EtOAc Methanol Solvent Control Container
Treatments

Figure 7. Percentage of eggs (mean i S. E.) laid by Papilio glaucus females
from Florida (n=11) on each of 5 choices: tulip trees leaves sprayed with
hexane, ethyl acetate and methanol extracts, solvent or untreated.

Significant differences between treatments at p < 0.01 for arcsine
transformed percentages are shown.

Mean “lo eggs

60

50 —»

4o 4

30 +

20 ~~

1o 4

53

 

2.6: 1.6

 

 

 

 

Hexane Solvent Control Container

Treatments

Figure 8. Percentage of eggs (mean i S. E.) laid by Papilio glaucus females
from Florida (n=5) on each of 3 choices: a tulip tree leaf sprayed with hexane
extract (hexane), a tulip tree leaf sprayed with hexane (solvent) and an
untreated tulip tree leaf (control). No siginificant differences between means
were detected by l-way and 2-way AN OVA on arcsine transformed

percentages.

Mean % eggs

60

50 ~-

54

 

 

 

3.0: 1.8

 

   

Ethyl acetate Solvent Control Container

Treatments

Figure 9. Percentage of eggs (mean i S. E.) laid by Papilio glaucus females
from Florida (n = 4) on each of 3 choices: an untreated tulip tree leaf
(control), a tulip tree leaf sprayed with ethyl acetate (solvent) and a tulip tree
leaf sprayed with ethyl acetate extract (ethyl acetate). No significant
differences were detected by 1-way AN OVA on arcsine transformed

percentages.

Mean % eggs

55

 

 

 

 

 

 

Methanol Solvent Control Container

Treatments

Figure 10. Percentage of eggs (mean at S. E.) laid by Papilio glaucus
females from Florida (n=3) on each of 3 choices: a tulip tree leaf sprayed
with methanol extract, a tulip tree leaf sprayed with methanol (solvent) and
an untreated tulip tree leaf (control). Significant difierences between means
(p < 0.05) for arcsine transformed percentages are indicated.

56

 

Mean % eggs

 

 

Methanol Solvent Control Container

Treatments

Figure 11. Percentage of eggs (mean :t S. E.) laid by Papilio glaucus
females (11 = 5) from OH, PA and KY on each of 3 choices: a tulip tree leaf
sprayed with methanol extract (methanol), a tulip tree leaf sprayed with
methanol (solvent) and an untreated tulip tree leaf (control). SignifiCant
differences between means (p < 0.01) are for arcsine transformed

percentages are shown.

 

Mean % eggs

50

45 4

4o-

35

30 «~

25—

20 ~»

15 —~

10 4»

57

 

0.7 :1: 0.5

 

 

 

     

 

   

 

Ethyl Aqueous Solvent Control Container
acetate

Treatments

Figure 12. Percentage of eggs (mean :1: S. E.) laid by Papilio glaucus
females (n=9) from OH, PA and KY on each of 4 choices: an untreated
tulip tree leaf (control), a tulip tree leaf sprayed with ethyl acetate (solvent),
a tulip tree leaf sprayed with the ethyl acetate fraction (ethyl acetate), and a
tulip tree leaf sprayed with the aqueous fraction (aqueous). The ethyl
acetate and aqueous fractions are subdivisions of the original crude
methanol extract (see text p 40). Significant differences between means (p <
0.01) for arcsine transformed percentages are shown.

W

More than 560 species of swallowtail butterflies (Lepidoptera: Papilionidae),
inhabit the world, with over 200 species in the genus Papilio. The majority of the species
inhabit tropical zones, and decline in number over the temperate and neartic latitudes
(Scriber 1995). However, it is thought that the origin of the Papilionidae is in eastern
(Hancock 1983) or western North America (Scriber et al 1991). The evolutionary history
that has shaped the geographic distribution and speciation of these butterflies remains an
uncertainty, although clues from paleontology, biogeography and botanical affiliation
have generated hypothesis concerning many swallowtail species.

The study of the chemical mechanisms governing host plant acceptance can
provide clues to the taxonomic relationships between swallowtail butterfly species as well
as provide explanations of how current patterns of host use may have diverged fi'om the
ancestral. The chemical facilitation hypothesis, generated by Dethier (1941) and Jermy
(1976, 1984) and expanded by Feeny (1991) states that the probability of colonization of
a novel host plant increases with increasing biochemical similarity of the new hosts to the
current hosts. Specifically, if a plant contains compounds that a butterfly already
recognizes as a oviposition or feeding stimulants, lacks chemical deterrents and can be
detoxified by the larvae, the plant has increased likelihood of becoming a host. I have

chosen to focus on the role that oviposition stimulants and deterrents play in host

58

59
acceptance by polyphagous (Papilio glaucus) and Oligophagous (Papilio troilus and
Papilio palamedes) species of swallowtail butterflies.

My research has demonstrated a chemical basis for host acceptance by P. troilus
and P. palamedes. In an all-filter paper bioassay, filter paper triangles coated with
spicebush extracts received more eggs by P. troilus than did the untreated control.
Additionally, P. troilus and P. palamedes laid more eggs on non-host leaf surfaces
sprayed with the respective host plant extracts than on the untreated leaves.

Although chemical cues are responsible for maintaining ovipositional fidelity on
the correct host species, other cues play a major role in eliciting oviposition. Relative to
the filter paper, extracts on non-host leaf surfaces received a greater percentage of eggs.
The relative unacceptability of the filter paper substrate manifest itself in the large
number of eggs laid off the triangles, scattered randomly about the oviposition arena
when no leaves were present. In contrast, in the all-leaf bioassays, the vast majority of
oviposition was confined to the tulip tree leaf surfaces treated with hoSt plant extract.
Although color plays a large role in orienting and arousing oviposition, sensory inputs
other than color may have been responsible in the recognition of the leaves as suitable
host substrate in my bioassays, since the green of the methanol extract was not enough to
stimulate significant oviposition. The relative role of texture, odor, and the reflectance of
the host surface in eliciting oviposition remains to be assessed for these Papilio species.

The results of my experiments decisively show that polar fractions of non-host
plant extract (Persea borbonia) misted onto host leaves deter oviposition by populations

of P. glaucus from southern Florida, Ohio, Kentucky and Pennsylvania. The detection of

60

deterrents by the adult female is advantageous when such compounds accurately reflect
the toxicity of a plant to her progeny, as is the case here. Only populations of P. glaucus
from Florida encounter P. borbonia, which grows alongside their host plant, Magnolia
virginiana, in dense swamp habitat. Sensitivity to contact deterrents in redbay may
prevent P. glaucus from making fatal oviposition mistakes in areas of dense growth of
both tree species.

Populations of P. glaucus from Ohio, Kentucky and Pennsylvania are deterred
from ovipoSition by extracts of a host plant that they never encounter. Gene flow could
in part explain that phenomenon, but if the rejection of redbay were a recent adaptation to
life in central and southern Florida, a gradient of deterrency decreasing in northern
populations would be expected. An alternate explanation may be that redbay compounds
have properties that activate an intrinsic deterrent response by P. glaucus. For example,
certain compounds can generate characteristic irregular or “deterrent” temporal patterns
of neural action potentials due to their interaction with the receptor cell membrane
(Schoonhoven et al. 1992). P. palamedes may have adapted to interpret redbay
compounds as oviposition stimulants, facilitated by historical use of lauraceous hosts by
troilus group species. On the other hand, P. glaucus may have evolved to recognize
redbay chemistry as deterrent. P. borbonia may have been among those species
inhabiting the midwestem United States that were pushed southward toward southern
Florida and Texas by global cooling. After the divergence of the troilus and glaucus
groups, frequent and unsuccessful colonization attempts of redbay may have facilitated

the evolution of its rejection by P. glaucus. The legacy of ovipositional inhibition by

61

compounds in redbay may explain a relative reluctance of northern populations to
oviposit on the predominant lauraceous plants (sassafras and spicebush) compared to
more favored host plants such as tulip tree and white ash (see Table 1, chapter 1). The
elucidation of the deterrent present in redbay foliage may likely explain the relative
preference of P. glaucus for Magnoliaceae relative to Lauraceae and perhaps even
Rutaceae.

Deterrent compounds tend to have high activity even when isolated (Honda and
Hayashi 1995), which may make their identification easier than that for oviposition
stimulants, which have high activity only when mixed together in synergy (Ohsugi et al
1985, Honda 1986, Feeny et al 1988, Nishida and Fukami 1989, Honda 1990, Sachdev-
Gupta et al 1993, Nishida 1995). Elucidation of the compound(s) in redbay deterrent to
P. glaucus is worthy of future study.

My research has demonstrated that host selection by P. troilus, P. palamedes and
P. glaucus is mediated, in part, by detection of chemical stimulants and deterrents in host
and non-hosts foliage. Biotic and abiotic factors determine the evolution of the direction
and range of gustatory receptor sensitivity and its neural interpretation. Current
behavioral responses to ecologically relevant compounds provide the background against

which past and future host affiliations can be extrapolated.

APPENDICES

APPENDIX 1

APPENDIX 1

Record of Deposition of Voucher Specimens*

The specimens listed on the following sheet(s) have been deposited in
the named museum(s) as samples of those species or other taxa which were
used in this research. Voucher recognition labels bearing the Voucher
No. have been attached or included in fluid-preserved specimens.

Voucher No.: 1996-7

Title of thesis or dissertation (or other research projects):

PHYTOCHEMICAL BASIS FOR HOST PLANT SELECTION BY GENERALIST
AND SPECIALIST SWALLOWTAIL BUTTERFLIES

Museum(s) where deposited and abbreviations for table on following sheets:

Entomology Museum, Michigan State University (MSU)

Other Museums:

Investigator's Name (8) (typed)
Cheryl Frankfater.

 

 

Date Segtemba: 19, 1996

*Reference: Yoshimoto, C. M. 1978. Voucher Specimens for Entomology in
North America. Bull. Entomol. Soc. Amer. 24:141-42.

Deposit as follows:

Original: Include as Appendix 1 in ribbon copy of thesis or
dissertation.

Copies: Included as Appendix 1 in copies of thesis or dissertation.
Museum(s) files.
Research project files.

This form is available from and the Voucher No. is assigned by the Curator,
Michigan State University Entomology Museum.

62

APPENDIX 1.1

APPENDIX 1.1
Vbucher Specimen Data

Pages

_1_.

1 of

Page

aufimuo>aaa ouwum damaged: onu ca uumoaov

oumn

Moumusu

 

mcxmamm mama

 

 

.Ebomdz xwoaosouam

 

you maoEHooam woumwa o>onm ecu wo>woumm

 

nmpmmxcmum axnmzo

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mnwmmfi .oz Honoao> Avoamuv Amvmamz o.uoumw«umo>cH
Ahumwmoooa ma mucosa Hmcoauwvvw away
meanmmo .o xumz
xn umpomaaoo mcmEHumam Hams
mm\am\m panama
a mvfinoam ..ou mncmazmflz mmumsmamq oaaflamd
mama .ooo unmamo
fl mufinoam .maEmh mmumemama oaaflama
mm\m\m pzmamo
a owns ..oo oocmnzmn mnawonp afiafimmm
mm\m\a pzmamo
H axunpcmx ..oo Emcuao mnaaohu oawmmma
om\mm\m pcmamu
a cane ..oo mucouzmn. Axumuv maoammm oaaflqmd
mm\fim\m pcmamo
a case ..oo mucopzmn mauamam oaaflamd
mm\mfi\m ucmgmo
a >xoapcmx ..oo Hamzod Azoaao>v mauamam oaaflamd
moor woufimomov use com: no vouooaaou aoxmu Honuo no moauoam
m e r r m m e .m m mcoefiuomm How mums Hanna
erOdellapWS
newsmaawmaa
“M w A.4. no .A .A P. n“ 1“ an

 

 

"mo noneaz

 

 

63

LIST OF REFERENCES

LITERATURE CITED

Ayers, M. P. and J. M. Scriber. 1994. Local adaptation to regional climates in Papilio
canadensis (Lepidoptera: Papilionidae). Ecological Monographs 64: 465-482.

Berenbaum, M. R. 1995. Chemistry and oligophagy in the Papilionidae. In: J. M. Scriber,
Y. Tsubaki and R. C. Lederhouse (eds.). Swallowtail butterflies: their ecology and
evolutionary biology. Walsworth Publishing Co., Missouri. pp. 27-38.

Bemays, E. A. and M. Cornelius. 1992. Relationship between deterrence and toxicity of
plant secondary compounds for the alfalfa weevil Hypera brunneipennis.
Entomologia Expermentalis et Applicata 64: 289-292.

Bemays, E. A. and R. F. Chapman. 1994. Behavior: the process of host-plant selection.
pp 95-165. In Miller, T. A. and H. S. van Emden (eds). Host-plant selection by
phytophagous insects. Chapman & Hall, New York. Chapter 4.

Blaney, W. M. and M. S. J. Simmonds. 1990. A behavioural and electrophysiological
study of the role of tarsal chemoreceptors in feeding by adults of Spodoptera,
Heliothis virescens and Helicoverpa armigera. Journal of Insect Physiology 36: 743-
756.

Bossart, J. L. 1993. Differential selection and adaptation in different host environments:
genotypic and phenotypic variation in host use traits in the tiger swallowtail butterfly,
Papilio glaucus L. Dissertation. Michigan State University, East Lansing, Michigan,
USA.

Bossart, J. L. and J. M. Scriber. 1995. Genetic variation in oviposition preference in tiger
swallowtail butterflies: interspecific, interpopulation and interindividual comparisons.
In: J. M. Scriber, Y. Tsubaki and R. C. Lederhouse (eds.). Swallowtail butterflies:
their ecology and evolutionary biology. Walsworth Publishing Co., Missouri. pp.
1 83 -1 93.

Darnman, H. and P. Feeny. 1988. Mechanisms and consequences of selective oviposition
by the zebra swallowtail butterfly. Animal Behavior 36:563-573.

64

65

Dethier, V. G. 1941. Chemical factors determining the choice of food plants by Papilio
larvae. American Naturalist 75: 61-73.

Dethier, V. G. 1980. Evolution of receptor sensitivity to secondary plant substances with
special references to deterrents. The American Naturalist 115: 45-66.

F eeny, P. 1991. Chemical constraints on the evolution of swallowtail butterflies. In: P.
Price, T. M. Lewinsohn, G. W. F emandes and W. W. Benson (eds). Plant-animal
interactions: evolutionary ecology in tropical and temperate regions. John Wiley
& Sons, Inc. pp. 315-340. Chapter 15.

Feeny, P., K. Sachdev-Gupta, L. Rosenberry and M. Carter. 1988. Luteolin 7-0-(6"-
malonyl)-B-D-glucoside and trans-chlorogenic acid: oviposition stimulants for the
black swallowtail butterfly. Phytochemistry 27: 3439-3448.

Feeny, P., E. Stadler, I. Ahman and M. Carter. 1989. Effects of plant odor on
oviposition by the black swallowtail butterfly, Papilio polyxenes (Lepidoptera:
Papilionidae). Journal of Insect Behavior 2: 803-827.

Gossard, T. W. and R. E. Jones. 1977. The effects of age and weather on egg-laying in
Pieris rapae L. Journal of Applied Ecology 14: 65-71.

Grant, G. C. and D. Langevin. 1994. Oviposition responses of four Choristoneura
(LepidopterazTortricidae) species to chemical and physical stimuli associated with
host and nonhost foliage. Environmental Entomology 23: 447-456.

Grossmueller, D. W. and R. C. Lederhouse. 1985. Oviposition site selection: an aid to
rapid growth and development in the tiger swallowtail butterfly, Papilio glaucus.
Oecologia 66:68-73.

Grossmueller, D. W. and R. C. Lederhouse. 1987. The role of nectar source distribution
in habitat use and oviposition by the tiger swallowtail butterfly. Journal of the
Lepidopterists’ Society 41: 159-164.

Hagen, R. H., R. C. Lederhouse, J. Bossart and J. M. Scriber. 1991. Papilio canadensis
and P. glaucus (Papilionidae) are distinct species. Journal of the Lepidopterists’
Society 45: 245-258.

Hancock, D. L. 1983. Classification of the Papilionidae (Lepidoptera): a phylogenetic
approach. Smithersia 2: 1-48. '

Honda, K. 1986. Flavone glycosides as ovipostion stimulants in a papilionid butterfly,
Papilio protenor. Journal of Chemical Ecology 12: 1999-2010.

 

66

Honda, K. 1990. Identification of host-plant chemicals stimulating oviposition by
swallowtail butterfly, Papilio protenor. Journal of Chemical Ecology 16: 325-337.

Honda, K. and N. Hayashi. 1995. A flavonoid glucoside, phellamurin, regulates
differential oviposition on a rutaceous plant, Phellodendron amurense, by two
sympatric swallowtail species, Papilio protenor and P. xuthus: the front line of a
coevolutionary arms race? Journal of Chemical Ecology 21: 1531-1539.

Ilse, D. 1937. New observations on responses to colours in egg-laying butterflies.
Nature 140: 544-545.

J ermy, T. 1976. Insect—host-plant relationship—Coevolution or sequential evolution?
In: T. Jenny (Ed.). The host-plant in relation to insect behavior and reproduction.
Plenum, New York. pp. 109-113.

Jenny, T. 1984. Evolution of insect/host plant relationships. American Naturalist 124:
609-630.

Kolb, G. and C. Scherer. 1982. Experiments on wavelenth specific behavior of Pieris
brassicae L. during drumming and egg-laying. Journal of Comparative Physiology
149: 325-332.

Lederhouse, R. C., M. P. Ayers, J. K. Nitao and J. M. Scriber. 1992. Differential use of
lauraceous hosts by swallowtail butterflies, Papilio troilus and Papilio palamedes
(Papilionidae). Oikos 63: 244-252.

Masumoto, T., S. Nomakuchi and K. Sawada. 1993. Host plant conspicuousness and the
distribution of eggs and larvae in the butterfly, Anthocharis scolymus (Lepidoptera:
Pieridae). Researches in Population Ecology 35: 241-250.

Miller, J. R. and K. L. Strickler. 1984. Finding and accepting host plants. In: W. J. Bell
and R. T. Cardé (eds). Chemical ecology of insects. Sinauer Associates, Inc.,
Sunderland, Massachusetts. pp. 127-157. Chapter 6.

Murphy, D. D. 1983. Nectar sources as constraints on the distribution of egg masses by
the checkerspot butterfly, Euphydryas chalcedona (Lepidoptera: Nymphalidae).
Environmental Entomology 12: 463-466.

Myers, J. H. 1985. Effect of physiological condition of the host plant on the ovipositional
choice of the cabbage white butterfly, Pieris rapae. Journal of Animal Ecology
54: 1 93-204.

 

67

Nishida, R. 1995. Oviposition stimulants of swallowtail butterflies. In: J. M. Scriber, Y.
Tsubaki and R. C. Lederhouse (eds). Swallowtail butterflies: their ecology and
evolutionary biology. Walsworth Publishing Co., Missouri. pp.l7-26.

Nishida, R. and H. F ukami. 1989. Oviposition stimulants of an Aristolochiaceae-feeding
swallowtail butterfly, Atrophaneura alcinous. Journal of Chemical Ecology 15:
2565-2575.

Nishida, R., T. Ohsugi, H. Fukami and S. Nakajima. 1990. Oviposition deterrent of a
- Rutaceae-feeding swallowtail butterfly, Papilio xuthus, from a non-host rutaceous
plant, Orixajaponica. Agric. Biol. Chem. 54: 1256-70.

Nitao, J. K., M. P. Ayers, R. C. Lederhouse and J. M. Scriber. 1991. Larval adaptation
to lauraceous hosts: geographic divergence in the spicebush swallowtail butterfly.
Ecology. 72: 1428-1435.

Nylin, S. 1988. Host plant specialization and seasonality in a polyphagous butterfly,
Polygonia c-album (Nymphalidae). Oikos 53: 381-386.

Ohsaki, N. and Y. Sato. 1994. Food plant choice of Pieris butterflies as a trade-off
between parasitoid avoidance and quality of plants. Ecology 75:59-68.

Ohsugi, T., R. Nishida and H. Fukami. 1985. Oviposition stimulant of Papilio xuthus, a
Citrus-feeding swallowtail butterfly. Agric. Biol. Chem. 49: 1897-1900.

Oyeyele, S. O. and M. P. Zalucki. 1990. Cardiac glycosides: and oviposition by Danaus
plexippus on Asclepiasfruticosa in south-east Queensland (Australia), with notes on
the effect of plant nitrogen content. Ecological Entomology 15: 177-1 85.

Papaj, D. and M. D. Rausher. 1987. Components of conspecific host-discrimination
behavior in the butterfly, Battus philenor. Ecology 68: 245-53.

Ramaswamy, S. B. 1994. Physiological basis of feeding and oviposition behavior in
moths. In: T. N. Ananthakrishnan (ed). Functional dynamics of phytophagous
insects. Science Publishers, Lebanon, N. H. pp. 55-78.

Rausher, M. D. 1978. Search image for leaf shape in a butterfly. Science 200:1071-1073.

Rausher, M. D. 1995. Behavioral ecology of oviposition in the pipevine swallowtail,
Battus philenor. In: J. M. Scriber, Y. Tsubaki and R. C. Lederhouse (eds).
Swallowtail butterflies: their ecology and evolutionary biology. Walsworth
Publishing Co., Missouri. pp. 53-62.

68

Rausher, M. D. and D. R. Papaj. 1983. Demographic consequences of discrimination
among conspecific host plants by Battus philenor butterflies. Ecology 64:1402-1410.

Renwick, J. A. A. and F. S. Chew. 1994. Oviposition behavior in lepidoptera. Annual
Review of Entomology 39:377-400.

Renwick, J. A. A. and Huang. 1994. Interacting chemical stimuli mediating oviposition
by Lepidoptera. In: T. N. Ananthakrishnan (ed). Functional dynamics of
phytophagous insects. Science Publishers, Lebanon, N. H. pp 79-94.

Robertson, H. G. 1987. Oviposition site selection in Cactoblastis cactorum
(Lepidoptera): constraints and compromises. Oecologia 73: 601- 608.

Roessingh, P., E. Stadler, R. Schtini and P. F eeny. 1991. Tarsal contact chemoreceptors
of the black swallowtail butterfly Papilio polyxenes: responses to phytochemicals
from host- and non-host plants. Physiological Entomology 16: 485-495.

Sachdev-Gupta, K., P. Feeny and M. Carter. 1993. Contact oviposition stimulants for
the pipevine swallowtail butterfly, Battus philenor, from an Aristolochia host plant:
synergism between inositols, aristolochic acids and a monogalactosyl diglyceride.
Chemoecology 4: 19-28.

Saxena, K. N. and S. Goyal. 1978. Host-plant relations of the citrus butterfly Papilio
demoleus L.: orientational and ovipositional responses. Entomologica Experimentalis
et Applicata 24: 1-10.

Schoonhoven, L. M., W. M. Blaney and M. S. J. Simmonds. 1992. Sensory coding of
feeding deterrents in phytophagous insects. In: E. Bemays (ed.). Insect-plant
interactions. Volume IV. pp. 59-79. Chapter 3.

Scriber, J. M. 1984. Larval foodplant utilization by the world Papilionidae (Lep.):
latitudinal gradients reappraised. Tokurana (Acta Rhaloceralogica) 2: 1-50.

Scriber, J. M. 1986. Origins of the regional feeding abilities in the tiger swallowtail
butterfly: ecological monophagy and the Papilio glaucus australis subspecies in
Florida. Oecologia 71 :94-103.

Scriber, J. M. 1988. Tale of the tiger: beringial biogeography, binomial classification,
and breakfast choices in the Papilio glaucus complex of butterflies. In: K. C. Spencer
(ed.). Chemical mediation of coevolution. Academic Press, New York. pp. 240-301.

Scriber, J. M. 1993. Absence of behavioral induction in oviposition preference of
Papilio glaucus (Lepidoptera: Papilionidae). The Great Lakes Entomologist 26:81-95.

69

Scriber, J. M. 1994. Climatic legacies and sex chromosomes: Latitudinal patterns of
voltinism, diapause, size, and host-plant selection in two species of swallowtail
butterflies at their hybrid zone. In: H. V. Danks (ed.). Insect Life-cycle
Polymorphism. Kluwer Academic Publishers, the Netherlands. pp. 133-171

Scriber, J. M. 1995. Overview of swallowtail butterflies: taxonomic and distributional
latitude. In: J. M. Scriber, Y. Tsubaki and R. C. Lederhouse (eds). Swallowtail
butterflies: their ecology and evolutionary biology. Walsworth Publishing Co.,
Missouri. pp 3-8. Chapter 1.

Scriber, J. M. 1996. Tiger tales: natural history of native North American swallowtails.
American Entomologist 42: 19-32.

Scriber, J. M. and R. C. Lederhouse. 1992. The thermal environment as a resource
dictating geographic patterns of feeding specialization of insect herbivores. In: M. R.
Hunter, T. Ohgushi and P. W. Price (eds). Effects of resource distribution on animal-
plant interactions. Academic Press, San Diego, CA. pp. 429-466.

Scriber, J. M., R. C. Lederhouse, and R. C. Hagan. 1991a. Foodplants and evolution
within the Papilio glaucus and P. troilus species groups (Lepidoptera: Papilionidae).
In: P. W. Price, T. M. Lewinsohn, G. W. F emandes, and W. W. Benson (eds). Plant-
animal interactions: evolutionary ecology in tropical and temperate regions. Wiley
and Sons, Inc. pp. 341-373.

Scriber, J. M., B. L. Giebink and D. Snider. 1991b. Reciprocal latitudinal clines in
oviposition behavior of Papilio glaucus and P. canadensis across the Great Lakes
hybrid zone: possible sex-linkage of oviposition preferences. Oecologia 87: 360-368.

Scriber, J. M. and A. Redman. 1993. Interspecific competition as a cause of hostplant
shifts in Michigan Lepidoptera. Bulletin of the Entomological Society of America
74: 432-433.

Shorey, H. H. 1964. The biology of T richoplusia ni (Lepidoptera: Noctuidae). 111.
Response to the oviposition substrate. Annals of the Entomological Society of
America 57: 165-170.

Singer, M. C., D. Ng, and C. D. Thomas. 1988. Heritability of oviposition preference
1 and its relationship to offspring performance within a single insect population.
Evolution 42: 977-985.

Smiley, J. T. 1978. Plant chemistry and the evolution of host specificity: new evidence
from Heliconius and Passiflora. Science 201: 745-747.

70

Smiley, J. T. 1985. Are chemical barriers necessary for evolution of butterfly-plant
associations? Oecologia 65: 580-583.

Sokal, R. R. and F. J. Rohlf. 1981. Biometry. W. H. Freeman & Co. San Fransisco. 859
PP-

Stéidler, E. 1974. Host plant stimuli affecting oviposition of the eastern spruce budworrn.
Entomologica Experimentalis et Applicata 17: 176-188.

Strong, D. R. 1988. Special feature: insect host range (series of articles). Ecology 69:
886-915.

Thompson, J. N. 1988. Evolutionary ecology of the relationship between oviposition
preference and performance of offspring in phytophagous insects. Entomologica
Experimentalis et Applicata 47: 3-14.

Thompson, J. N. and O. Pellmyr. 1991. Evolution of oviposition behavior and host
preference in lepidoptera. Annual Review of Entomology 36:65-89.

Traynier, R. M. M. 1984. Associative learning in the ovipositional behaviour of the
cabbage butterfly, Pieris rapae. Physiological entomology 9: 465-472.

Traynier, R. M. M. 1986. Visual learning in assays of sinigrin solution as an oviposition
releaser for the cabbage butterfly, Pieris rapae. Entomologica Experimentalis et
Applicata 40: 25-33.

Underwood, D. L. A. 1994. Intraspecific variability in host plant quality and
ovipositional preferences in Eucheira socialis (Lepidoptera: Pieridae). Ecological
Entomology 19: 245-256.

van Loon, J. J. A., T. C. Everaarts and R. C. Smallegange. 1992. Associative learning in
host-finding by female Pieris brassicae butterflies: releaming preferences. In: S. B. J.
Menken, J. H. Visser and P. Harrewijn (eds). Proc. 8th Int. Symp. Insect-Plant
Relationships. Dordrecht: Kluwer Acad. Publ.

Watanabe, M. 1995. Population dynamics of Papilio xuthus larvae in relation to the life
history of the host tree. In: J. M. Scriber, Y. Tsubaki and R. C. Lederhouse (eds).
Swallowtail butterflies: their ecology and evolutionary biology. Walsworth
Publishing Co., Missouri. pp. 101-106.

Wiklund, C. 1984. Egg-laying patterns in butterflies in relation to their phenology and
the visual apparency and abundance of their host plants. Oecologia 63: 23-29.

71

Williams, E. H. 1981. Thermal influences on oviposition in the montane butterfly
Euphydryas gillettii. Oecologia 50:342-346.

Williams, K. S. 1983. The coevolution of Euphydryas chalcedona butterflies and their
larval host plants. Oecologia 56: 336-340.

Wolfson, J. L. 1980. Oviposition response of Pieris rapae to environmentally induced
variation in Brassica nigra. Entomologica Experimentalis et Applicata 27:223-232

Zangerl, A. R. and M. R. Berenbaum. 1992. Oviposition patterns and hostplant
suitability: parsnip webworrns and wild pamsip. The American Midland Naturalist
128: 292—298.

"I11111111111111.1711