Insect Life
Tiger Tales: Natural History of
Native North American Swallowtails
J.
MARK
SCRIBER
% Neonate (lot instar) Larval Survival
T
IlE FIRST DRAWING OF A NORTH
AMERICAN
butterfly was of a male tiger swallowtail, produced more than 400 years
ago by John White in an area now called
North Carolina. White was the commander
of Sir Walter Raleigh's third expedition to the
North American continent to an area generally referred to as Virginia. The styljzed water
color drawing was done by White in 1587, apparently after a Native American Indian boy
presented him the specimen of this "manaanggwas" (=butterfly), and was later reproduced as a wood-cut in the very rare book
RHAMNACI!AE
TILIACI!AE
RATANAC!J\E
o
,
,
"
,
,
",
-~""\
RUTACEAE
_ _
OLEA-CEAE
-
_
--==--
ROSACEAE
-
-
-
'\,
'\
~''\
PaplUo
species
"
"
',...p ',,,
~
""
"'~
c~.~
A~IFI\ICAN ENTOMOI.O(;IST
•
Spring 1996
01!:!!~.~.1!:!!.~01!:!!1!:!!~0
~Iaucus
,
O~?
"nlexiares
canadensis
0
scamandcr
?
gammas
•
palamedes
~
troilus
pilumnus
~
?
A. _____
c.------
? • B.~ _?_?
?
?
~.
00000
••
00000000000
D.???????????
D.-----OOO?
? ? ?
000000
?
?
0
?
V V
PHYLOGENY
Based upon
mtDNA & allozyme cladograms & phenograms
r?dJ0~1!:!!.0
0.00
••••
?????????????
"-V
CONSENSUS
•••
••••
ril'J~0
••
~.ril'J1!:!!0?~
E.-~.00.ril'J~~.ril'J.r?dJ0
~
multicnudotus
.0
0~00.ril'J
0.00
0
.., :::::don
lnsectorium sive Minimorum Animalium
Theatrum (Holland 1931, Tyler et al. 1994).
Even today, as one of the most notable
North American natives, the tiger swallowtail
butterfly is abundant and widespread. From
the male specimen, Linnaeus named the tiger
swallowtailPapilio turnus, which he assumed
to be a different species than the dimorphic
dark female morph called Papilio glaucus L.
It is now known that P. glaucus, which has
both dark and yellow form females, occurs
throughout
the eastern half of the United
States. Other smaller taxa were described
north of the Great Lakes as P. g. canadensis
Rothschild & Jordan and in Alaska as P. g.
arcticus Skinner. Both of these smaller species
(currently =canadensis) lack the dimorphic
dark form female. In the south, a putative
Florida and Gulf states subspecies, P. g. australis Maynard, and the Mexican swallowtail
species, P. alexiares Hoppfer, have been described, both with dimorphic (dark and yellow) females. Three largely sympatric species
of tiger swallowtails occur in the western half
of the United States (P. rutulus Lucas, P. eurymedon Lucas, and P. multicaudatus Kirby).
All three apparently
lack the dark morph
form of females and have been variously reported to have local varieties, subspecies, and
aberrations of form.
Our research group has been investigating
the comprehensive
natural history, ecology,
14~;'
=70-100%
,
BEI'UIACEAE" "
SALlCACEAE _ ••..•.
:::: ~
LAURACEAE
:;:0·
£2a = 15· 69~o
•
~
~
:5
~
~
(J
0
"
~
~
~ ~
~ ~ ~
~
~
~
a
0
~ ••
(J
~
&l
; iI I
Consensus Papi/iD phylogeny on the left has been constructed using a comparison of allozymes (Hagen & Scriber
1991) and analysis of mtDNA (Sperling 1993). Different
plant utilization abilities of neonate larvae (i.e.,survival
throught the first instar) have been determined for various
populations of PapiliD and their hybrids on ten food plant
families (Scriber et a!. 1991). Best survival is shown in dark
boxes. Presumed origin of detoxication abilities are shown
for each plant family with the letters A-E and the associated
dotted lines from the left (note that the Rhamnaceae
connects only to P. eurymedon). Papilio species seem to
feed on all host families up the phylogenetic lines above a
particular letter. For example, P. troi/us, P. palamedes, and
P. pilumnus can feed on the Lauraceae (having the
detoxication abilities derived at A), but not on any other of
the nine families. P. multicaudatus has host use abilities for
Rosaceae, Oleaceae, and Rutaceae (derived at C) but not
host use ability for Salicaceae, Betulaceae, Platanaceae, or
Tiliaceae (derived at D) or Rhamnaceae (derived at E).
Limited information is currently available on P. pi/umnus and
Central American and South American P. garamas and P.
scamander; however, we have analyzed their hybrids with
the P. glaucus and P. palamedes group. Total number of
families is approximately 1,000, and total number of their
offspring bioassayed is in excess of 25,000 individual
larvae. Recent placement of the South American scamander
and Central American garamas closer to glaucus than troi/us
may suggest that the ability to detoxify certain Magnoliaceae may have had a single origin instead of multiple origins in
different continents (see Table 1).
19
and systematics of these North American P.
glaucus and P. trailus swallowtail butterfly
species groups and their closest Central and
South American relatives for 15-20 years.
The compulsion to study this tiger swallowtail complex and their sister group (theP. troiIus complex) was catalyzed by an undergraduate writing assignment in 1969, which
required a geographical and ecological analysis of the distributions and abundances of the
species in the two complexes with a discussion
of the biotic and abiotic factors (historic and
current) that appeared to limit the range of
species involved. This article describes recent
progress toward completion of that project
with continued experimental investigation of
causes that delineate the range of this primitive, polyphagous,
and indigenously North
American glaucus complex in comparison to
the related sister P. troilus complex.
Basic Natural History of Life Stages
(opposite page) Geographic
distribution of the P. glaucus
species group (Section III,
Papilionidae). Final (fifth) instars
are shown here, including P.
glaucus with its os mete rial
glands everted because of
disturbance. Also shown in the
upper left of the glaucus group is
a five-species puddling group of
males near the Blue Mountains
of Washington.
20
Unlike most of their more tropical relatives
(approximately
500 species), the tiger swallowtails of the temperate/boreal
forests of
northern North America spend from four to
eight months as diapausing pupae, often under snow. This is possible because of physiological adaptations that enhance acclimation
and survival under very cold temperatures
(possibly because of the presence of ethylene
glycol or related cryoprotectants
[Kukal et al.
1991]). The escape oftiger swallowtails from
the specialization on tropical host plant families (e.g., Rutaceae, Lauraceae, and Magnoliaceae) that, with the Aristolochiaceae,
Apiaceae, and Annonaceae, comprise the diets of 90-95% of all swallowtail butterfly
species, and their persistence at high latitudes,
has made them doubly unique among the
world's Papilionidae.
In the spring, adult males eclose, begin to
fly, and search for emerging females with
which to mate. These males may fly along
canyons of the Rocky Mountains, streams of
the boreal forests across Canada, or hedgerows and islands of trees from Florida to
Maine and northward.
They may live 1-2
months and may mate several times. We have
observed with spermatophore counts that females also mate several times (up to five or
six) during the course of their 3-6 week naturallife span in the field. A series of compressed
spermatophores
can easily be counted upon
dissection of multiply-mated
female abdomens. We have shown that multiple matings
are important for restoring female fertility,
which usually declines rapidly after about one
week of oviposition (i.e., 7-8 days postmating). A successful mating consists of the transfer of the male spermatophore,
which can be
of variable size, into the female's bursa copulatrix. It is believed that sperm from the most
recent spermataphore
fertilize the eggs laid
subsequently (i.e., sperm precedence).
Eggs are fertilized as they are oviposited
singly onto host plant leaves. The P. glaucus
group generally prefers to oviposit on the top
of leaves and the three Lauraceae-specialized
members of the P. troilus group (P. troilus L.,
P. palamedes Drury, and P. pilunmus Boisduval) on leaf bottoms. Eggs develop and change
to a yellowish or darker brown color before
eclosion of the neonate larva. Eggs in the P.
glaucus group are initially green, whereas
eggs in the trailus group species are yellowish.
Normal-colored
P.glaucus adults, which have
been reared as larvae on artificial diet, produce eggs that are basically blue or turquoise.
The eggs are much the same color as the blue
larvae that result whenP. glaucus are fed diets
that lack carotenoids found in normal host
plant leaves.
Egg development to larval eclosion is thermally sensitive and generally requires 5-10
days under normal field conditions. The first
instar often eats the egg shell and proceeds to
Jay down a silk mat on the leaf surface, thus
providing a firm attachment.
This behavior
of weaving a silken foothold persists through
the remaining stadia (usually five). In P. troiIus, this behavior results in a leaf roll in which
the larva hides. However, P. glaucus larvae
generally are exposed and do' not form complete rolls of the leaves they rest upon.
Larvae of tree-feeding swallowtails spend
most of their time resting on leaves of their
host plant, but may change location several
times throughout their 3-5 week lives. For P.
glaucus, neonate and second instars usually
feed at the edge of the leaf where their mother
placed the egg. Location changes usually occur immediately premolt or postmolt. With
patience, one can observe feeding bouts of
1-2 minutes, which are spaced several hours
apart in mid- to late instars. In contrast,
movement to a new leaf for feeding, with a
return to the same resting pad, is common for
third to fifth instars. The distance traveled
from resting to feeding site can be several feet,
and it is common for mid-late instars of P.
glaucus (and probably other species of the
group) to gnaw through the petioles of the
leaves they have just fed on until the partially
consumed leaves fall to the ground. This leafAMEI\ICAN
ENTOMOLOGIST
•
Spril/g 1996
22
gnawing or leaf-clipping behavior may remove evidence of caterpillar presence to natural enemies that locate hosts by olfaction (e.g.,
hymenopteran
parasites) or, perhaps, those
that use leaf feeding damage as a visual clue
(possibly birds as well as human researchers).
It is also feasible that this behavior prevents
chemical induction in leaves that are adjacent
to damaged leaves and now may be of lower
quality or are toxic for subsequent larval feeding. Although allelochemical induction has
been a hot topic for the past two decades, it is
not known what selection pressure may have
shaped the leaf-clipping behavior inP. glaucus
nor how extensive it is on various host plant
species. Perhaps swallowtail larvae do not
conditions are more similar to central Alaska
than to geographic regions elsewhere in the
continental United States (Scriber 1994). Butterflies in the thermal depressions of Michigan also have short growing seasons and
experience more severe seasonal constraints
than those in surrounding counties. Although
there is a gradual latitudinal decline in forewing length from Florida (62-66mm) to Alaska
(42-44 mm), there are exceptions where locally smaller adult sizes are found in cold
pockets of Michigan (40-43 mm) than otherwise observed in other populations at similar
latitudes (45-50 mm) (see Scriber 1994). At
the same growth rate, a Papilio caterpillar
that pupates at a smaller size will complete its
clip leaves on plants that do not express chem-
growth before another
ical induction.
Whereas larval growth rates of all insects
are temperature-determined
to some extent,
our group has learned that nutritional quality,
or suitability, of host leaves is the major determinant of larval development rates. However,
temperature is also significant in the development of life history traits (i.e., strategies) and
interacts with leaf quality at different latitudes. For example, seasonal constraints determining the number of generations per year,
and the need for reproductive synchrony of
these populations, have resulted in basically
predictable annual flight periods at such locations. The first flight ofP. glaucus in the United States is usually in early March when new
southern Florida butterflies appear. The first
flight of Papilio elsewhere occurs as spring
moves northward
across the continent. By
late May, P. glaucus and P. canadensis may be
simultaneously on the wing from the central
United States all the way to Alaska. A second
generation of tiger swallowtails is usually observed from the Great Lakes southward to the
Gulf States from Texas to Florida, where a
third or even fourth flight may occur.
Altitudinal variation in seasonality and
voltinism is also important. Should the wrong
host choice be made by ovipositing P. glaucus
or P. canadensis females at certain locations
(e.g., at high elevations or high latitudes), larvae may not successfully complete development to the pupal (overwintering)
stage.
From the Great Lakes region northward to
Alaska, only a single generation is possible,
even on the best host plants (Ayres & Scriber
1994). The northernmost
populations
are
photoperiodically
insensitive obligate diapausers. Cold pockets, or phenological depressions, are known locally in sections of
northern Michigan and Wisconsin, where
to a larger size. Such early pupation behavior
for smaller caterpillars may allow survival
that would otherwise be impossible in such
thermal depressions. Although many other
natural selection pressures and ecological
variables may affect adult size, such size reductions greatly enhance the odds for rapid
pupation and -thus seem exceedingly important by assuring that they may reach this diapause stage before winter sets in.
The combination
of cold tolerance and
generalized feeding abilities on nontropical
host plant families (e.g., Rosaceae, Oleaceae,
Salicaceae, Betulaceae, Tiliaceae, Platanaceae, and Rhamnaceae) have doubtlessly been
major factors permitting the radiation of six
species of the P. glaucus group of swallowtail
butterflies into the higher global latitudes.
Both adaptive suites of traits (winter survival
and polyphagy) are considered rare in swallowtails and both may be a partial explanation for the decline in species richness of the
Papilionidae from tropical to boreal habitats.
that continues
to grow
Molecular and Biochemical
Phylogenetics of the Group
The phylogenetic distinctions suggested in
1959 by Lincoln Brower (based on various
morphological,
color, and genitalic traits for
the eight North American species of Munroe's
section III of the Papilionidae [1961)), basically agrees with our analysis of 26 electrophoretic allozyme loci for these same species.
Our laboratory has found sufficient differentiation among P. glaucus, P. rutulus, P. eurymedon, P. multicaudatus, and P. alexiares
for each to be considered a distinct species,
confirming earlier suggestions by Brower and
Munroe. We have recently (Hagen et al. 1991)
added P. canadensis, which was previously
A\1ERICAN ENTO,\101.0(;IST
•
Spring 1996
The P. troilus
Species Group
P.
regarded as a subspecies or seasonal variant
of P. glalicus for 200 years. There was no evidence of differentiation
between the Florida
subspecies P. glauClts australis Maynard and
P. g. glaucus from other regions. There was
also insufficient allozyme differentiation
to
distinguish P. canadensis from the putative
Alaskan subspecies P. canadensis arcticus.
Also we recognized three distinct species in
the P. troillis group in section III: two subspeA~lFRICAN ENTmIO\ ()(;IST
•
5prillg
1996
cies of P. troilus L. (P. t. troilus on the coast
and P. t. ilioneus inland), P. palamedes
Drury, and P. pilumnus Boisduval. All of
these are more closely related to each other
than to any of the six P. glaucus species.
Within the glaucus group, the most surprising phylogenetic relationships were that
P. glaucus and P. canadensis are different
species and that the Mexican P. alexiares
appears more closely related to P. glaucus
Geographic distribution of the P.
froilus species group.Originally
believed to be part of the P.
glaucus species group (based
largely on its resemblance to the
tigerlike wing pattern), P. pilumnus
was moved into its current
taxonomic position in the troi/us
group by Brower (1959).
23
(left) Assorted mortality causes
and natural enemies of Papilio in
North America.
(A) Leaf-cutter
ants, Arta texana (Buckley)
carrying off pieces of sassafras
leaf, some of which had P. troi/us
eggs, in the Big Thicket of East
Texas. (B) Forest tent caterpillar,
Malacosoma disstria Hubner, at
the edge of severe defoliation
beginning
consumption
of a small
choke cherry seedling with a P
canadensis egg (green, lower
right) in northern Wisconsin. (C)
Tachinid fly pupae from P.
glaucus fifth instar larva. (0) Mite
eating neonate (first instar) P
troi/us on sassafras tree in
Ingham County, Michigan.
(E)
Parasitoid (Trogus sp.) searching
for additional larvae immediately
after stinging fifth instar P.
g/aucus on tulip tree in Ingham
County, Michigan. (F) Common
scene for both P. g/aucus and P.
canadensis (here) is this
(Podisus sp.) feeding
pentatomid
on hemolymph
from impaled
larva. (G) Formica sp. ants killing
P canadensis larva on black
cherry in northern Michigan.
24
AMERICAN
ENTOMOI.O(;IST
•
Spring 1996
than is P. canadensis. Secondly, the sympatric
western species P. rutu/us and P. eurymedon
were found to be sister species with extremely
little divergence in biochemically detected genetic traits. Sperling's (1993) restriction site
analysis of mitochondrial
DNA produced
phylogenies that were basically concordant
with the allozyme phylogenies for the g/aucus, canadensis, and a/exiares relationships,
and the rutu/t/s and eurymedon relationship.
Although virtually identical in mtDNA
and allozymes,rutulus
andet/rymedon clearly
are different species. They express different
adult wing color patterns, larval host use
traits, and oviposition preferences, in addition to different wing morphology, genitalia,
and mating
preferences
(Brower
1959).
However, we have found (Hagen & Scriber
1995) that hybrids between these two species
are unusual for Papilio hybrids because they
do not show the Haldane effect (sterility or
deficiency of the heterogametic
sex among
offspring in F, hybrids; in birds and Lepidoptera, it is generally believed that females
are XY, or heterogametic, and males are XX,
or homogametic). The lack of Haldane effect
implies a close genetic relationship between
rutt/lus and eurymedon and a lack of major
adaptive differences controlled by factors on
the sex chromosomes. This contrasts with the
P.glaucus and P. canadensis sister taxa, which
differ quite significantly for traits on the sex
chromosomes
(e.g., diapause
regulation,
adult wing polymorphism,
suppressors
of
dark morph female color, and differential oviposition preferences for certain species of the
Salicaceae and the Magnoliaceae host plants
as well as Haldane effect; Scriber 1994).
There is doubtlessly a fascinating role involving various sex-linked factors, including the
Haldane effect, in speciation processes of
these butterflies that needs to be clarified.
and V) (see Table 1) of the family. In contrast
to conventional wisdom of the past 4-5 decades, our interspecific hybridization studies
between P. g/aucus and P. scamander, and
pairings of each with the Lauraceae-specialized North American P. troi/us and P. palamedes, suggest a closer genetic relationship
between the glaucus and scamander groups
(both extremely polyphagous) than between
the glaucus and troilus groups (both North
America). Based on mtDNA analyses, Sperling (1991) added evidence that the polyphagc)Us P. scamander and its relative in Central
America, P. garamas Hubner (or P. abderus
Hopffer) of the P. homerus group, may be
more closely related to the North AmericanP.
g/aucus group than are the North American
Lauraceae-specialized
P. troilus group.
These findings
raise questions
about
whether a generalist ability to feed on the
Magnoliaceae,
the Lauraceae, and several
additional plant families may have originated
only once, in a common ancestor of the geographically widely separatedglaucus
and scamander species. With a probable phylogeny
based on our findings, this and other questions may be more rigorously addressed,
What are the specific detoxication
mechanisms and biologically active phytochemicals
A:vmm:AN
ENTmlOl.m;)ST
•
Spring 1996
P. glaucus from various museum
records and personal collections.
Within peripheral extremes it is
likely that dark morphs exist in
most, if not all, counties without
dots.
'0'
Host Plants and Range Limitations for
Tree-feeding Swallowtails: Evolution of
Polyphagy in Papilionidae
Of the 560 or more species of Papilionidae
in the world, about 65% feed on one of five
tropical plant families (Rutaceae, Apiaceae,
Aristolochiaceae,
Anonaceae, or Lauraceae).
The two most polyphagous
species are the
North American P. glaucus and the South
American P. scamander Boisduval. Both species feed on dozens of host species from five to
nine plant families. Traditionally, these butterflies have been taxonomically
considered
representatives of different sister sections (III
Range of dark morph females of
,.,
2::)---
.ILlS
IOQ
I
,
200
I
..
I
'00
1,1.
*00
I
,
25
involved with differential host use abilities?
Are the toxic chemicals also the cues for adult
oviposition avoidance? Have these behaviors
and abilities evolved many times, independently, or only once? How do adult oviposition preferences correlate with larval detoxication capabilities on the hosts? With regard to physiological detoxication abilities or
behavioral oviposition preferences, which are
evolutionarily
more stable or labile? What
role do host plants and differential detoxication abilities play in the speciation process,
and are close host affiliations the major explanations for the currellt geographic distribution of North American species of the P.
glaucus group? What are the dynamics and
causes of the narrow zone of overlap/tension/
hybridization
or genetic introgression
between P. glaucus and P. canadensis that has
been maintained for at least 100 years in the
Great Lakes and Appalachian mountain plant
ecotones? What are the key adaptive traits
and environmental limitations that determine
the thermal ecology, diapause physiology,
.'
I
-
genodynamics
(local dispersal, introgressive
hybridization, sex-linkage, genetic hitch-hiking, etc.), and current distributions of North
American swallowtail butterflies?
Although P. glaucus larvae have the capability to develop on woody plants in at least
10 families, other species and subspecies within the P. glaucus group differ significantly in
their ability to survive and grow, particularly
on Magnoliaceae, Lauraceae, Rutaceae, Salicaceae, Betulaceae, Platanaceae, and Rhamnaceae. TheP. trai/us group species are phytochemical specialists with larvae feeding only
on particular species of Lauraceae at any given location. The combination of phytochemical affiliations and glacial phytogeography
have doubtlessly played major roles in determining the distributions
of key host plants
and butterfly species in these two species
groups. For example, the Magnoliaceae and
Lauraceae are primarily found in the southeastern part of North America, as are those
affiliated butterfly species. P. g/aucus is found
on sweetbay or tulip tree of the Magnoliaceae
..
\'
\
Degree
~ Days (C)
~'-~1390~ •• __ 1500
",'
Range of the model species,
Bartus phi/enor, and the
geographic position (solid black
band) of the average seasonal
thermal unit accumulations
corresponding
to 1,390-1 ,500°C
degree-days above a base
threshold of 1Q°C. These data
were assembled
from various
state publications and are based
upon 20-30 year averages
generally between 1950 and
1980 (see text for discussion).
26
100
...I
• I
.ILI:S
100
I'
'0'
I
300
I.
-00
I
I
J
..',.'
AMERICAN
ENTOMOI.O(;lST
•
Sprillg 1996
Table 1. Species within the Papilionidae
family (modified
from Munroe
1961)
Section
II
9
No. species"
Primary host-plant
families"
Lauraceae
]43
Rutaceae, and less commonly
on Apiaceae (=Umbelliferae)
machaon group also on
Compositae
III
IV
V
9
35
20
Rutaceae,
Piperaceae
scamalzder'f group,
and homems' group
g/a/lCUs( complex on
Magnoliaceae,
Lauraceae
Rutaceae, Rosaceae
Salicaceae, Betulaceae,
Oleaceae, Platanaceae,
Rhamnaceae
on Lauraceae,
Hernandiaceae,
Magnoliaceae,
others
troi/us(
complex on Lauraceae
(;cographic ranges were as follows: Section I, Africa, Asia, and IndoAustralia; Section III, worldwide; Section II, North America;
Section IV, Central America; Section V, Central or South America .
•r See Brown
1994 (in Scriber et al. 1995c).
" See also Hagen and Scriber 1991, Scriber et a!. 1991, Sperling 1993.
'"Includes the six P. K/allCl/s group species (g/a/lcus, rt/tu/IIS. eurymedon, mll/ticaudatus. a/exiares, and the recently added canadensis;
see fig. 2).
d Scamalzder
group species consists of P. he/lanichus Hewitson, scamander Boisduval bircha/li Hewitson, and xalzthop/eura Godman
and Salvin.
(' Homert/s group species consists of P victorinus Doubleday, cepha/lls Godman and Salvin, c1eotas Gray, aristeus Cramer, jlldicael
Oberthiir, IJaramas (Hubner), homems Fabricus, warscewiczi Hopffer, cacius Lucas, euterpinus Godman and Salvin, diazi Racheli and
Sbordoni (see Tyler 1975).
ITroilus group consists of P. troilus, P. pa/amedes, and P. pi/llmnus (see Fig. 3).
and P.1Jalamedes on red bay of the Lauraceae
(whereas troilus feeds on redbay in Florida, it
prefers spicebush or sassafrass elsewhere).
Similarly, the western and northern distribution of key Salicaceae and Betulaceae species
across North America basically corresponds
with that of P. canadensis or P. YUtulus (see
Scriber et al. J 991). Because these polyphagOlls glallClls group species transcend the
range of any single host species, other biotic
and abiotic factors must contribute to geographic range limits (and the dynamics of
gene flow).
Although genetically based interspecific
and intraspecific differences in host plant acceptance, detoxication,
and larval growth
abilities exist among North American swallowtails, we have observed a wide range of
reported and potential plants (more than 70
species from 12 families) that are capable of
supporting successful development. Sufficient
phenotypic and genotypic variation in ovipositional behavior has been observed in all P.
glauCtis group species tested, thus allowing
ecological mistakes and possibilities for evolutionary host shifts. Although heritable and
behaviorally
fixed preferences
for natural
hosts are generally observed in multichoice
laboratory studies, an average of 2-8% of the
AMfl\lCAN EN1()~101()(;IH
•
Spring 1996
plants selected for oviposition in these studies
are nonhosts or toxic for larvae (Scriber
1993). Although some of the patterns of ovipositional preference have been shown (via
reciprocal, hand-paired
hybrids) to be sexlinked in P. glaucus and P. canadensis (for
quaking aspen or tulip tree [Scriber 1994]),
the differential detoxication abilities in larvae
appear to be based on autosomal traits. Recent work with P. glaucus (Bossart & Scriber
1995) suggests that in some cases, there is a
correlation of adult oviposition preference
and larval growth performance.
However,
this may be the exception rather than the rule
for phytophagous insects.
Our studies of neonate larval survival on,
and adult ovipositional behavior toward, variOlls potential hosts suggest some rather stable
phylogenetic constraints on host range in certain Papilio. However, the variation that is
present is sufficient to permit natural selection
to effect some major host shifts. Based on
these contrasting observations, we conclude
that the role of past and current host plant
distributions in determining geographic range
limits for these native North American butterflies is preeminent but not exclusive (see Fig.
1; Suggested Readings).
27
•••
•••
•••
•••
•••
••• -..........-..
•••
•••
•••
...
...
•••
• ••
•• .\t
••••
•••
••••
••
.-- .
"
.Q•
Q ••
e>C)~
000
PROPORTION
(!)O~
OF
ALLELES:
CANADENSIS
000
TYPE
•••
Ldh Pgd
Proportion
of canadensis
alleles
for Ldh, Pgd, and Hk loci in
samples of butterflies
from
Michigan,
and
Minnesota,
Wisconsin counties. Hk was not
scored for samples from all
counties; only 2 circles are
shown in those cases. Filled
circle, 100% canadensis; open
circle, 0% canadensis (=100%
glaucus); intermediate frequencies of canadensis
alleles are
indicated by the proportion of
each circle filled. Shaded region
indicates the position of
1,400-1 ,500°C seasonal
isotherm
represents
in Michigan; this
the predicted northern
limit for a multivoltine
28
life cycle.
0C)C9
Hk
Biotic Factors that Affect the
Geographic Range Limits of Species
In addition to host plant distribution, other
biotic factors such as competing gene pools
(i.e, inter- and intraspecific competition), parasites, and predators may play major roles in
determining whether a particular swallowtail
butterfly is absent or rare in a particular area
of North America. For example, extensive
gypsy moth defoliations (750,000-1 million
acres annually) have imposed a serious impact on twoPapilio species in Michigan. Furthermore, there are associated natural enemies that increase with gypsy moth populations (e.g., the parasitic and polyphagous tachinid fly Compsilura concinnata Meigen),
that also cause serious mortality in some populations ofP. canadensis andP. glaucus at distances more than 50 miles ahead of the gypsy
moth defoliation front. More important to
the local population dynamics of native swallowtails is the effect of extensive aerial application of pesticides for gypsy moth control.
We have observed that the microbial pesticide
from Bacillus thuringiensis kurstaki Berliner
can kill Papilio species for 20,30, and 40 days
after spraying on tree leaves sprayed at 16 BIU
(billion international units)/acre aerial application rates used for gypsy moths. Such interactions of microbial pesticides with nontarget
arthropods (including natural enemies) warrant considerable research because the available information is so sketchy at this time.
Diseases also may be devastating in some geographic or topographic areas depending on
habitat, microclimate, and even the species of
host plant that larvae feed on. The role of
various generalist and specialist natural enemies (see Fig. 4) in Pap ilia population dynamics is currently under investigation. We hope
to determine the relative importance of these
native and introduced agents in the evolution
of diet breadth in Pap ilia relative to other
causes such as competition, chemical coevolutionary constraints, and the voltinism/suitability hypothesis (see Scriber & Lederhouse
1992).
Perhaps the most familiar example of the
potential significance of natural enemies on
the distribution of P. glaucus is related to the
adult dark morph (mimetic) color polymorphism found in some portion of P. glaucus feAMERICAN ENTOMOI.()(;IST
•
S/Jrillg
1996
males from the Great Lakes into Mexico. For
decades, it has been presumed that the dark
Illorph females mimic the distasteful pipevine
swallowtail butterfly, Battus philel10r L., and
that the frequency of dark form females is
higher where the selective advantage is greater because of large populations of the model
species. This argument is basically believable;
however, the recent increased frequency of
dark females in the southern half of Florida
and the lack of any dark morphs in the southwestern United States, despite the occurrence
of the model, imply the need for alternative
explanations.
Furthermore, the northern limits of the sex-linked (Y chromosome)
dark
morph gene seems to be more precisely delineated by abiotic factors (thermal unit accumulations of 1,400°C above a base 10°C
developmental
threshold) than by the geographic distribution of the distasteful model,
B. philel1or. A seasonal accumulation of degree-days fewer than 1,400 selects intensively
against completion of a second generation for
P. glaucus and other Lepidoptera (e.g., European corn borer) at this Great Lakes transition zone where a sex-linked obligate diapause regulation occurs. This sex-linkage of
diapause regulation and dark color expression and suppression are described in more
detail elsewhere (Scriber et a!. 1995a).
The reason for the lack of complete fixation of the dark morph in females in areas of
high B. philenor populations,
such as the
Smoky Mountains and Gulf states, has been
suggested by various authors to result from a
balanced polymorphism exerted by male mating preferences for yellow morph females.
Robert Lederhouse has led our research group
in evaluation of this hypothesis by observing
male responses to tethered, size-matched females of different color morphs. Based on
geographically diverse field studies over several years, we now conclude that male mating
preferences will vary with the relative frequency of the two morphs at various locations
and it seems doubtful that male intersexual
selection could prevent the genetic fixation of
the dark (mimetic) allele. For example, in
Florida, where dark morph females have increased rapidly (from 5 to 35% of the population in 30 years), males are still much more
likely to court and copulate with yellow morph females than with dark morph females. In
southern Ohio, where dark morphs comprised 80% (or more) of the population,
males were equally or less likely to copulate
with yellow than dark morph females. The
intensity of the selective advantage of this preA~lFRf('AN
ENTOMlll
O(;IST
•
Sprillg 1996
Final instar larvae of P glaucus
(top), P scamander (center), P
palamedes (bottom), and their F1
hybrids (g x s) and (p x s) which
are respectively between the
parental types. Female parent is
listed first for hybrids.
sumed mimicry remains unclear; however, we
have observed that in the northern edge of the
range, a sharp and predictable limit in the
occurrence of mimetic dark morphs is basically independent of the model species. This dark
morph range limit appears to result from several interrelated ecological factors (sex-linkage, voltinism, plant ecotones, hybrid zones,
global warming, and Pleistocene glaciations).
Because host plant distributions are clearly
related to temperature (extremes and seasonal degree-day accumulations),
it is not surprising that plant ecotones (or transition
zones) between life zones (e.g., Merriam's)
29
correlate with abiotic (thermal) factors. We
have found that the center of the transition
life zone of Merriam, which runs along Canada and across the Great Lakes region down
into the Appalachian Mountains, is closely
approximated
by the seasonal thermal unit
accumulation of 1,400 degree-days (above a
base lO°C). Also, the center of this plant ecotone (Fig. 6) is the precise northern limit of
dark morph tiger swallowtail females, the
northern limit of tulip tree detoxication abilities (which corresponds roughly to the northern range of tulip tree), and the northern limit
of bivoltine (two generation)
capabilities.
The southern limits of aspen (Salicaceae)
detoxication, suppression of the dark morph
female, and presence of obligate (photoperiodically insensitive) diapause are known also
to occur in this same narrow zone. We now
know that this unique zone delineates the
parapatric juncture of the Canadian swallowtail species, P. canadensis, and the tiger swallowtail. Limited hybridization at this zone is
evident in subtle morphological intermediates
and in genetic introgression detected using diagnostic allozymes. In the Great Lakes region,
we see evidence of asymmetrical gene flow
with southward introgression in Salicaceae
detoxication abilities and X-linked color suppressor activity, but limited northward movement of Magnoliaceae detoxication abilities
or of Y-]inked dark morph genes. Additional
research with allozymes and mitochondrial
DNA may help clarify this picture.
Sex Chromosomes and Hybrid Zone
Dynamics
Experimental
investigations
of various
ecological determinants of range limits in P.
glaucus have not only clarified the species status of P. canadensis, but have elucidated what
may be a major evolutionary mechanism of
speciation: sex-linkage of key life history
traits (or coadapted gene complexes). With 30
sets of chromosomes, it seems especially provocative that we have located genes on the Xchromosomes
of two Papilio species that
control obligate versus facultative diapause,
mimetic versus nonmimetic wing color, and
differential oviposition preferences for aspen
versus tulip trees. These genes have been
mapped in relation tQ loci for five electrophoretic allozymes (Scriber 1994). The deleterious interspecific
hybridization
X-(Haldane) effect is also apparently controlled by
loci somewhere on this chromosome, and it is
]ikely that these losses in viability, fertility, or
30
survival of heterogametic hybrids (females in
Lepidoptera) would be a major selective force
in hybrid zones where limited movement of
sex chromosomes
(compared to autosomes)
has been observed for insects.
It is uncertain to what degree such accumulations of genes on the X-chromosome
of P.
glaucus and P. canadensis results from suppressed crossovers between the X and Y chromosomes or simply from faster rates of
evolution of traits on the X-chromosome
(e.g., recessive mutations)
compared
with
those on the autosomes. In any case, this fascinating comparative
autecological
natural
history and biogeographical survey of Fa/Jilio
species groups has permitted a reductionist
progression from field ecology into biochemical and molecular mechanisms that may help
explain the genetic basis of adaptations
for
the various populations
and species. These
mechanisms, in turn, have helped to define the
natural distribution limits of the taxa and to
provide insight into the speciation process.
Our ability to hand-pair and obtain viable
hybrids of almost every combination of species pairs among and between the North
American P. glaucus and P. troilus groups (as
well as with South American species) has facilitated our progress tremendously
(Scriber
et al. 1995b). It seems likely that only through
a combination
of approaches (e.g., descriptive and experimental,
laboratory and field,
molecular and ecological) will we be able to
understand the temporal and spatial distributions and interactions of individuals, popu]ations, and species. The phyto- chemical bases
and genetic mechanisms of differential larval
detoxication abilities have already been elucidated using hand-paired
hybrids and backcrosses with bioassays
on plants of the
Salicaceae and Magnoliaceae families (Scriber et al. 1989). Although the geographic borders betweenP. canadensis andP. glaucus may
be largely host-plant influenced and these separate rather distinctly at the plant transition
between boreal and temperate forests (Hoffman & Blows 1994), we know little about
intraspecific geographic differences in plant
phytochemistry
(Johnson & Scriber 1994)
and even less about interspecific (or hybrid)
phytochemistry
and its influence on herbivores (Strauss 1994).
Life history traits cannot be studied effectively one at a time because of the interactive
matrix of biotic and abiotic selective pressures
that can simultaneously assault any particular
phenotypic combination. The complexity of
interactions among life history traits and the
A~IERICAN
ENTOMOI.OCIST
•
Sprillg 1996
dynamics
of environmental
forces
may pre-
clude any rigorous
determination
of the relative fitness
of individuals
with any given
combination
of adaptations
because
fitness
must be evaluated
in relation
environment(s}
encountered.
studies
groups,
vergence
to the particular
Comparative
of related species such as these Pa!Jilio
with differing
degrees of genetic diand
phylogenetic
relatedness,
vide a powerful
approach
to these
fascinating
but complex
questions.
pro-
and other
Acknowledgments
I am particularly
grateful for the dedicated
work of the following students and close cooperators who, through their own hard work over the
past decade, helped clarify the biology of these
Papilio species: Matthew Ayres, Janice Bossart,
Keith Brown, Robert Dowell, Mark Evans, Cheryl
frankfater,
Bruce Giebink, Robert Hagen, John
Hainze, Kelly Johnson, Olga Kukal, Robert Lederhouse, Rick Lindroth, Heidi Luebke, Jim Maudsley, Ana Beatriz Barros de Morais, James Nitao,
Ahnya Redman,
Sally Rockey, frank Siansky,
Doozie Snider, felix Sperling, Patti Tidwell, and
Wayne Wehling. Research was partially supported
from the USDA, NSt~ and Michigan State University. I would also like to thank Matt Ayres, Robert
Lederhouse, and James Nitao for several of the
larval photos and Wayne Wehling for the five-species puddling adult photo. Janice Bossart, Catherinc Bristow, and Wayne Wehling provided helpful
reviews of the manuscript.
References Cited
Ayres, M. P. & J. M. Scriber. 1994. Local adaptations to regional climates in1'apilio calladellsis
(Lepidoptera: Papilionidae).
Ecol. Monogr. 64:
465-4R2.
Bossart,J. L. & J. M. Scriber. 1995. Maintenance
of ecologically significant genetic variation in
the tiger swallowtail butterfly through differential selection and gene flow. Evolution
(in
press).
Brower, L. P. 1959. Speciation in butterflies of the
1'. glallclIs group. 1. Morphological
relationships and hybridization.
Evolution] 3: 40-63.
Hagen, R. H. & J. M. Scriber. 1991. Systematics
of the Papilio glallclIs and r. troillls groups
(l.epidoptera:
Papilionidae):
Inferences from
allozymes.
Annals Entomol.
Soc. Am. 84:
3RO-395.
] 995.
Sex chromosomes
and speciation
in the
J. M.
Scriber, Y. Tsubaki & R. C. Lederhouse [eds.],
The swallowtail butterflies: their ecology and
evolutionary
biology. Scientific, Gainesville,
1'alJilio glallclIs group. pp. 2] 1-227.
III
fL.
Hagen, R. H., R. C. Lederhouse,J.
A.\II'HICAN
ENT().\H
>I.O(;I"T
•
L. Bossart &J.
Sprillg 1996
M. Scriber. 1991. 1'apilio canadensis and P.
(Papilionidae) are distinct species. J.
Lepid. Soc. 45: 245-258.
Hoffman, A. A. & M. W. Blows. ]994. Species
borders: Ecological and evolutionary perspectives. Trends Ecol. Evol. 9: 223-227.
Holland, W. J. 1931. The butterfly book. Doubleday, Page, New York.
Johnson, K. & J. M. Scriber. 1994. Geographic
variation in plant allelochemicals
of significance to insect herbivores. pp. 7-31. In T. N.
Ananthakrishnan
[ed.], Functional dynamics
of phytophagous
insects. Oxford and IBH,
New Delhi, India.
Kukal, 0., M. P. Ayres & J. M. Scriber. 1991.
Cold tolerance of pupae in relation to the distribution of tiger swallowtails.
Can. J. Zool. 69:
3028-3037.
Munroe, E. 1961. The generic classification of the
Papilionidae.
Can. Entomol. Suppl. 17: 1-51.
Scriber,J. M. 1993. Absence of behavioral induction in ovipositon preference of PalJilio glaucus
(Lepidoptera: Papilionidae).
Great Lakes Entomol. 26: 81-95.
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, pp.
133-177. In H. V Danks [ed.], Insect life-cycle
polymorphism:
theory, evolution, and ecological consequences for seasonality and diapause
control. Kluwer Academic, Dordrecht,
The
Netherlands.
Scriber, J. M. & R. C. Lederhouse.
1992. The
thermal environment
as a resource dictating
geographic patterns of feeding specialization of
insect herbivores, pp. 429-466. II1M. R. Hunter, T. Ohgushi & P. W. Price [eds.], Effects of
resource distribution on animal-plant
interactions. Academic, New York.
Scriber, J. M., R. L. Lindroth & J. Nitao. 1989.
Differential toxicity of a phenolic glycoside
from quaking aspen leaves by Pap ilia glaucus
subspecies,
their hybrids, and backcrosses.
Oecologia 81: 186-191.
Scriber, J. M., R. C. Lederhouse & R. Hagen.
1991. Foodplants
and evolution within the
Pap ilia glaucus and PalJilio troilus species
groups
(Lepidoptera:
Papilionidac),
pp.
341-373. III P. W Price, T. M. Lewinsohn, G.
W. Fernandes & W. W. Benson [eds.], Plant-Animal interactions: evolutionary ecology in tropical and temperate regions. Wiley, New York.
Scriber, J. M., R. H. Hagen & R. C. Lederhouse.
1995a. Genetics of mimicry in the tiger swallowtail butterflies, Papilio glaucus and P. canadens;s (Lepidoptera:
Papilionidae).
Evolution (in press).
Scriber,J. M., R. C. Lederhouse, and R. V. Dowell.
1995b.
Hybridization
studies with North
American swallowtails pp. 269-281. III Swallowtail butterflies: their ecology and evolutionary biology. Scientific, Gainesville, FL.
glallws
31
NEW
FRO
M
A NEW GENERAL
CATALOGUEOJ'mE
ANTS OJ'mE WORLD
Barry Bolton
"Thi book wi II have
permanent helf life.
All entomological
librarie and entomologi ts with even
a marginal interest
in ant mu t have it,
b cause it meliculou Iy crafted synonymi cover all of
ant lassifi alion
from Linnaeu to lhe
pre nt."
-E. O. Wil on
Here are all the
names, antique and
mod rn, of all the
ant that are or ever were-from the
ar tic to the tropical, the fo ilized to
the living, the mislabeled to the newly
chri tened member of the family
Formicidae. Along with Bolton's
Identifi ation Guide to the Ant Genera
of the World, it will be the e ential ref-
eren e for anyone, expert or amateur,
with an intere t in ants.
$12500
HARVARD
JOURNEY
TO THE ANTS
A Story of Scientific
Exploration
Bert Holldobler and
Edward o. Wilson
short, fascinating sequel to ... The
whi h won a Pulitzer Prize ... It' a
page-tu rner."
-Helen Fi her,
Scriber, J. M., Y. Tsubaki & R. C. Lederhouse
reds]. 1995c. Swallowtail butterflies: their
ecology and evolutionary biology. Scientific,
Gainesville, FL.
Sperling, F.A.H. 1991. Mitochondrial DNA phylogeny, speciation, and hostplant coevolution
of Papilio butterflies. Ph.D. dissertation, Cornell University, Ithaca, NY.
1993. Mitochondrial
DNA variation
and
Haldane's rule in the Papilio glaHcHs and P.
tmillls species groups. Heredity 71: 227-233.
Strauss, S. Y. 1994. Levels of herbivory and parasitism in host hybrid zones. Trends Ecol. Evol.
9: 209-2 L3.
Tyler, H. 1975. The swallowtail butterflies of
North America. Naturegraph, Healdsburg, CA
1994 revision by H. Tyler, K.S.B. Brown & K.
Wilson [eds.], Swallowtail butterflies of the
Americas. Scientific, Gainesville, FL.
"A
Ants,
ew York Times Book Review
"A marvelou tal -in word and brilliant color photos."
-Thea Singer, Boston Sunday Globe
"Beautifully writt n and iliustrated ... A
bustling but well-organized ant heap,
full of wonders natural and intellectual."
-Philip Morrison, cientific American
Suggested Reading
Scriber, ]. M., Y. Tsubaki & R. C. Lederhouse,
[eds.]. 1995. Swallowtail butterflies: Their
ecology and evolutionary biology. Scientific,
Gainesville, FL.
•
Belknap Press
95 color Illus. 36 halftones. 42 line Illus.
$16.95 paper
KRAKATAU
cloth
The Destruction
and Rea s mbly
of an Island Ecosystem
Ian Thornton
THE WISDOM OF
mEHIVE
The Social Physiology
Honey Bee Colonie
of
Thomas D. Seeley
The re ult of fifteen year of re earch
into the honey bee colony, this book
offer the clearest, most detailed picture
available of how a highly integrated
animal society works. Seeley draw on
the literature of molecular biology, cell
biology, animal and human ociology,
economic , and operations re earch to
compare the honey bee colony to other
functionally organized group. By comparing sy tems in different pecies,
Seeley argues, we can deepen our
under tanding of the me hani ms that
make clo e cooperation a reality.
19 halftones. 118 b/w & 6 color IlIus.. 6 tables
$49.95 cloth
On Augu t 27, 1883, the i land of
Krakatau near Java erupted with a force
nearly ten thousand times that of the
hydrogen bomb dropped on
Hiroshima, obliterating all plant and
animal life. Krakatau is a comprehensive account of the rea sembly of a
tropical fore t eco y tem on Krakatau
over the pa t entury.
32 halftones. 20 line IlIus..·$39.9s cloth
At book tores or from
Harvard
University
Press
Phone: 800-448-2242
httpJ/www.hup.harvard.edu
32
Jon Mark Scriber is a professor and chairperson of the Entomology Department at Michigan
State University in East Lansing, MI 48824. His
interests center on geographical insect ecology,
and range from biochemistry, behavior, and
physiology to genetics and evolutionary biology.
These broad interests and research foci are best
reflected in the recent book on swallowtail butterflies (Scriber et al. 1995) in the references. Current projects include the study of nontarget and
ecosystem impacts of gypsy moth management
decisions (Bacillus thuringiensis sprays and outbreak defoliations). He has also written two Science Citation Classics (Institute of Scientific
Information) on insect nutritional ecology.
AMERICAN ENTO.\101.O(;fST
•
Sprillg 1996