602
E t o l u r i o t r . 5 2 ( 2 ) .1 9 9 1 {p. p . . 1 1 3 5
BUTTERFLIES AND PLANTS: A PHYLOGENETIC STUDY
N r x l , q s J , q N z rR N n S O n E NN y r - r N
Departntentrtf Zoologt',Unit,ersityoJ Stockholn,106 91 Stockholm,Swedert
I E - n t a i l : n i k l a s . j a n z . @z . o o l og i . s u .s e
Abstract.-A database on host plant records frorn 437 ingroup taxa has been used to test a number of hypotheses on
the interaction between butterflies and their host plants using phylogenetic methods (sirnple character optinrization.
concentrated changes test, and independent contraststest). The butterfly phylogeny was assembled from various s()urces
i r n d h o s t p l a n t c l a d e s w e r e i d e n t i f i e d a c c o r d i n g t o C h a s e e t a l . ' s r b c J - - b a s e dp h y l o g e n y . T h e a n c e s t r a l h o s t p l a n t a p p e a r s
to be associated within a highly derived rosid clade, including the family Fabaceae. As fossil data suggest that this
clade is older than the butterflies, they must have colonized already diversilied plants. Previous studies also suggest
that the patterns of association in most insect-plant interirctions are more shaped by host shifts, through colonization
and specialization. than by cospeciation. Consequently, we have focused explicitly on the mechanisms behind host
shilis. Our results confirm, in the light of new phylogenetic evidence, the pattern reported by Ehrlich and Raven that
related butterflies feed on related plants. We show that host shifts have generally been rnore comrnon between closely
related plants than between nore distantly related plants. This finding. together with the possibility ofa highertendency
of recolonizing ancestral hosts, helps to explain the apparent large-scale conservation in the patterns of association
between insects and their host plants, patterns which at the same tinle are more flexible on a more detailed level.
Plant growth form '*'as an even more conservative aspect of the interaction between butterflies and their host plants
than plant phylogeny. However, this is largely explained by a higher probability of colonizations and host shifts while
1 ' e e d i n go n t r e e s t h a n o n o t h e r g r o w t h f o n n s .
Ke\, words.-Coevolution,
host shifts, insect-host plant interactions, Lepidoptera. Papilionoidea. specialization.
Received March 22, 1996.
Accepted November 2.6, 1991
Few systems have played such an important role in our
understanding of how species interactions evolve as butterflies and their host plants. To a large extent this is the result
of a single influential paper by Ehrlich and Raven (196.1).
Their essay inspired a flood of paperi on difl'erent aspects of
this association, and a number of related hypotheses on the
evolution of insect-plant interactions have emerged. However, there have been few attempts to exploit the large database on butterfly-host plant affiliations to test such hypotheses using phylogenetic methods (Mitter and Brooks
1983; Miller 1987a). A major reason for this is that wellsupported phylogenies. fbr both butterflies and seed plants,
h a v e b e e n u n a v a i l a b l e .H o w e v e r , t h i s i s s l o w l y c h a n g i n g , a n d
today it is possible to put together reasonably robust phylogeniesfor both groups. Chase et al. (1993) have recently
published a molecular phylogeny for all seed plants, which
is probably the best estimate of large-scale angiosperm phylogeny to date. Butterfly phylogenies are also emerging and
we have constructed a plausible phylogeny across the butterfliesby combining these published estimates.
Ehrlich and Raven ( I 964) argued that the patterns of host
plant association that we observe today have been shaped by
a stepwise coevolutionary process in which plants evolve
d e f e n s e sa g a i n s t n a t u r a l e n e m i e s , a n d t h e s e e n e m i e s i n t u r n
e v o l v e n e w c a p a c i t i e st o c o p e w i t h t h e s ed e f e n s e s .P l a n t s t h a t
e s c a p ef i o m h e r b i v o r e s c a n d i v e r s i f y i n t h e a b s e n c eo l ' e n e m i e s . I n s e c t st h a t e v e n t u a l l y m a n a g et o c o l o n i z e o n e o l ' t h e s e
p l a n t s w i l l e n t e r a n e w a d a p t i v ez o n e a n d c a n i n t u r n d i v e r s i f y
o n t o t h e r e l a t i v e s o l ' t h i s p l a n t , b e c a u s et h e y w i l l b e c h e m i c a l l y s i m i l a r . E h r l i c h a n d R a v e n a r g u e d t h a t t h e s ep r o c e s s e s
h a v e l e d t o t h e m a i n p a t t e r n t h e y h a d o b s e r v e d ,n a m e l y t h a t
related butterflies tend to f'eed on related groups of plants.
M o s t o r a l l p l a n t d i v e r s i f i c a t i r ) nu p t o t h e l e v e l o f r e s o l u t i o n
used in our analysis had probribly already taken place at the
time the butterflies started to diversify. The oldest known
butterfly fossil dates back to 48 M.Y.B.P and the diversification of the butterfly families probably took place at the end
o f t h e C r e t a c e o u s ,a b o u t 6 6 M . Y . B . P ( E m m e l e t a l . 1 9 9 2 ) .
At least some families even in the most recently derived of
the plant clades used in this analysisdate to this time, such
as, Urticaceae(a member of Rosid I in Chase et al., 1993):
9 0 M . Y . B . P . , R u t a c e a e( R o s i d 2 ) : 5 2 M . Y . B . P . , A p i a c e a e ( A s t e r i d 2 ) : 5 2 M . Y . B . P . , A p o c y n a c e a e( A s t e r i d l ) : 6 0 M . Y . B . P .
(dates fiom Eriksson and Bremer 1992). Therefore, the clades
themselves must be even older. It is reasonable to regard the
e v o l u t i o n o f c u r r e n t a s s o c i a t i o n sa s a r i s i n g g e n e r a l l y t h r o u g h
butterfly colonization of already-diversified hosts, and that
is the approach we shall take. This is not to say that coevolution is an unimportant process in the interaction between
butterflies and their host plants. only that evidence fbr it
should be sought at other levels of resolution.
There are two fundamentally different approaches to comparative analyses using phylogenetic data. One approach
seeks to find and explain general ecological or evolutionary
c o r r e l a t i o n s( F e l s e n s t e i n1 9 8 5 ; G r a f e n 1 9 8 9 ; H a r v e y a n d P a gel 1991: Pagel 1992), while the other seeksto reconstruct
a n d e x p l a i n p a r t i c u l a r h i s t o r i c a l e v e n t so r s e q u e n c e so f e v e n t s
along branches in a phylogeny (Mitter and Brooks 1983:
Coddington l98U; Sill6n-Tullberg 1988; Maddison 1990,
B r o o k s a n d M c l e n n a n l 9 9 l ) . T h e s e a p p r o a c h e sa r e c o m plernentary (Coddington 1994; Nylin and Wedell 199,1;Pagel
1 9 9 4 ) a n d w e h a v e i n t h e p r e s e n tp a p e r u s e d b o t h , d e p e n d i n g
on the prcblem.
The question of ancestral host associationsis a clearly
historical problem. On what plant did the first butterfly feed'/
This question is interesting in its own right, but answering
it also provides necessary information for any phylogenetic
tests regarding direction of evolution of host plant associat i o n s . E h r l i c h a n d R a v e n ( 1 9 6 4 ) r e g a r d e di t m o s t l i k e l y t h a t
the ancestral host plant family was Aristolochiaceae. ln con-
486
1 - i l 9 9 l J T h c -S o c i c t l ,l i r r t h c S t r " r d o
v l ' E , r o l u t i o n .A l l r i g h t s r c s c r v c d
487
BUTTERFLIES AND PLANTS
trast, Scott (1986) noted that Fabaceae were eaten by the
most basal branches of several butterfly families and suggested that the ancestral host probably was a legume. More
recently, Ackery (1991) suggestedthat Malvales may instead
b e l h e a n c e s l r a lh o s t a s s o c i a t i o n .
Ehrlich and Raven (1964) noted several factors influencing
the association between butterflies and their host plants, but
particularly stressed the importance of the plant's secondary
metabolic substances. They noted that many higher taxa of
plants are characterized by distinctive secondary chemistry,
and cited a number of examples in which related butterflies
feed on related plants. They also cited examples where related
butterflies are t'eeding on unrelated plants with chemical similarities. These observations are consistent with, though not
sufficient to demonstrate, the importance of plant chemistry.
However, nobody has tried to determine whether and at what
taxonomic scales the tendency to feed on related plants is
statistically demonstrable for butterflies as a whole, as opposed to selected examples.
Although plant chemistry has been viewed as the prime
factor governing the evolution of butterfly-host plant associations (e.g. Feeny 1915, 1976, l99l; Jermy 1916, 1984
Scriber and Slansky 1981: Berenbaum 1983; Zangerl and
Berenbaum 1993; Fiedler 1995b), other aspects of the host,
not necessarily well correlated with phylogeny, might also
have a large eff'ect (Benson et al. 1975; Smiley 1978; Price
et al. 1980; Courtney 1984; Bernays and Graham 1988; Anderson 1993). One example is host growth form. Different
growth forms can dominate in different habitat types, which
have distinct combinations of microclimate, enemies, etc.
These require specialized adaptations, making it more difficult for a butterfly to colonize a new growth form and habitat
than to colonize a new host plant with a different chemical
composition in the same habitat. Herbaceous host plants may
also require a different search behavior by ovipositing females than do arboreal hosts. Thus, host growth fbrm might
be more conserved than host clade membership, when both
are traced on the butterfly phylogeny.
We might also expect that the propensity to shift between
host taxa would differ between butterflies feeding on diff'erent
growth forms. For example, Feeny (1916) hypothesized that
herbaceous plants are def'endedby diverse "qualitative" toxins that require corresponding diverse physiological and behavioral adaptations in the attacking insect, while trees are
characterized by "quantitative" def'enseconsisting of a limited number of digestion-reducing agents such as tannins,
which do not require specialized detoxification tactics. Under
this hypothesis, trees make up a chemically more homogenous group. Host shifts should thus be easier and more common between trees than between herbs. This would influence
the relationship between phylogenetic and growth form conservatism, especially if tree feeding is common among butterflies. Similar arguments have been advanced to explain an
apparent association between tree feeding and polyphagy
(Futuyma 1976; Fiedler 1995a), but these hypotheses has
never been tested using phylogenetic methods.
In this paper, we have performed phylogenetic analyses of
the interaction between butterflies and their host plants to
address the lbllowing questions regarding the patterns and
causes of host shifts, colonizations and specialization: (1)
Are patternsof butterfly host plant utilization nonrandomso
that relatedbutterfliesfeed on relatedplants,as suggestedby
Ehrlich and Raven? (2) What was the ancestralhost plant
association,and has this associationconstrainedhost plant
utilizationin butterflies?(3) Are host shifts involving closely
relatedplant speciesmore common than shifts to more distantly related plants'?(4) Are there idenrifiablegroups of
unrelatedplants that often occur together as hosts? (5) Is
plant phylogenya more conservativeaspectofbutterfly-plant
associationsthan plant growth fbrm, or vice versa?(6) Are
major host shifts more common in woody-plant-feeding
than
in herb-f'eeding
lineages?(7) Are tree-feedingbutterfly taxa
associatedwith a larger numberof host plant cladesthan are
herb-feeding taxa?
MErnoos
Unlessotherwisestated,all analyseshave beencarriedout
using the computerprogramMacClade(vers.3.05,Maddison
and Maddison 1992).
Phylogenies
The plant phylogeny used in this study follows the rbclbased analysis of seed plant relationshipsby Chase et al.
(1993).They perfbrmedtwo diff'erentsearchesusing slightly
different taxon sampling and weighting procedures.These
searchesproducedvery similar trees.We havefbr the present
analysisusedthe tree producedby their search2 (or tree B),
which they judged to be the most reliable.Chaseet al. summarizedtheir findingsin a simplifiedcladogram(their {ig. 2)
in which most terminal taxa were given informal names,reflecting their approximatecorrespondenceto groupings in
previous classifications.This summary phylogeny is presentedin a modified fbrm in figure 2 (Chaseet al. 1993).
With minor modificationsto be noted,we haveusedthe Chase
et al. terminalcladesand nomenclatureas the characterstates
in our analysisand discussionof butterfly host associations.
For better resolution within their "Rosid l" clade, exceptionally important as butterfly-host plants, we have recognized six subclades,following the branchingpatternof their
"tree B," which we labeled "Rosid lA . . . F," Unlessotherwise specified,the term "plant clade" in this paperrefers
to these terminal clades and subcladesin the Chase et al.
phylogeny. There are problems with this analysis,mainly
arising from the computationaldifficulties of analyzing a
datasetof this size.In a critique of the analysis,Baum (1994)
notes that although it is very likely that Chase et al. have
not found the most parsimonioustree, the final phylogeny
include many higher level groupingssuggested
by traditional
systematists,
and that eventhe unconventionalplacementsof
sometaxa ofien fit surprisinglywell with morphologicaldata.
In any case,it is the most comprehensiveattempt so far to
reconstructa phylogenyfor the seedplants as a whole, and
should be a better estimateof the true phylogeny than one
inferred fiom previoustaxonomy.For most of our analyses
we have only usedthe terminal cladesin this phylogeny,not
the deeperbranchings,which may be less reliable.In a few
cases where host plant families were not included in the
analysisby Chaseet al., their positionswere inferred from
the classificationof Cronquist (1981). These plant families
488
N. JANZ AND S. NYLIN
T r a p e z i t i n a(e1 )
t
Appendix 2, (the subfamily level relationshipsare shown in
Fig. 1).
There are perhapseven more uncertaintiesin the butterfly
M e g a t h y m i n a(e1 )
Hesperidea
(outgroup) phylogenythanin the plant phylogeny.Partsof the phylogeny
P y r g i n a e( 1 )
are poorly resolved. This is particularly true for the large
P y n h o p y g i n a (e1)
family Lycaenidae,but also tbr Pieridae,wherewe havecho^^^l^ti^^^
/i \
vucilduil
ras \ r/
sen to collapsebranchesabout which the pheneticanalyses
B a r o n i i n a (e1 )
by Geiger (1980) and Ehrlich and Ehrlich (1961) were in conif'- PP aa rpni lai osns ii ii nn aa((ee545) ) Papilionidaeflict. The basal structure of Nymphalidae is left unresolved
whereHarvey (1991)conflictswith Scottand Wright (1990),
D i s m o r p h i i n a(e1 )
but the nymphalid taxonomic groupings in our phylogeny
( 1' 1 )
Coliadinae
Pieridae
follow the taxonomyof Harvey ( 1991).WhereMiller ( 1987b)
conflicts with Hancock (1983) on the resolutionof species
P i e r i n a e( 2 8 )
(9)
Riodininae
Riodinidae groupsin Papilionini,we have followed Miller's more recent
study. We have also, when possible,tried to estimatethe
P o l y o m m a t i n a(e6 8 )
effect
of the phylogeneticreconstructionon our results.
T h e c l i n a e( 6 7 )
The choiceof outgroupto the butterfliesis somewhatprob( 1)
Lycaeninae
Lycaenidae
lematic.When tracing characterevolution on a phylogenetic
(1)
Poritiinae
tree, one shouldideally use severaloutgroups,including the
C u r e t i n a e( 1 )
sistergroup, to correctly reconstructancestralcharacterstates
Libytheinae(1)
(Maddisonet al. 1984).In our casethis criterion is difficult
A p a t u r i n a e( 15 )
to fulfill as the phylogeny of higher Lepidopterais almost
D a n a i n a e( 11 )
completelyunknown (Nielsen 1989).The skippers(Hesperiidae) are strong candidatesas the sister taxon to the true
M o r p h i n a e( 5 )
butterflies,but it has been suggestedthat the wholly South
B r a s s o l i n a(e1 )
American group Hedyloidea should be positionedbetween
S a t y r i n a e( 4 )
Nymphalidae
the true butterfliesand the skippers(Scoble 1986).HedyloRosid1B
C a l i n a g i n a (e1 )
idea is poorly known (including its host plant affiliations)
-i
ilr:r::: ::::::::.*r:x.'.
Charaxinae (1 8)
f: No
.l
and its position is controversial.The most recent analyses
H e l i c o n i i n a(e4 5 )
actually f'avorHesperiidaeas sistergroup to the Papilionoidea
I
.'! ves
L i m e n i t i n a(e4 5 )
(Weller and Pashley1995;de Jong et al. 1996;Scoble 1996).
1
:: : : : . . ,
(
4
4
)
N
y
m
p
h
a
l
i
n
a
e
For
this reasonwe have chosento excludeHedyloideafiom
NN Equivocal
the
study,
but we test the effect that this may have on the
FIc. 1. Simplified version of the phylogeny of Papilionoideaused
reconstructionof the ancestralhost plant association.
in the analyses, showing the major relationships within butterflies.
H e s p e r i i n a(e1 )
I
I
Total number of taxa in the phylogeny actually used in the analyses
is given within parenthesesbehind each taxon name. Utilization of
the plant clade Rosid 1B (Cannabidaceae,Fabaceae,Krameriaceae,
Moraceae, Polygalaceae, Rhamnaceae, Rosaceae, Ulmaceae, Urticaceae, and Zygophyllaceae) is traced on the phyiogeny. The assignment of states to branches follows optirnization on the complete
phylogeny.
Host Plant Datct
Data on host plant utilization were collectedfrom several
sources(Ehrlich and Raven 1964; Common and Waterhouse
1972;Larsen1974, l99l1' Smart 1975;JohnstonandJohnston
1 9 8 0 ;H i g g i n sa n d H a r g r e a v e s1 9 8 3 ;S c o t t 1 9 8 6 ;d e l a M a z a
Ramires 1987; DeVries 1987; Migdoll 1987; Miller 1987a1.
Ackery 1988; Parsons1991; Corbet and Pendlebury 1992;
Ebert 1993). Becauseit is dif{icult to evaluatethe validity
of literature data on host plant affiliations, false records certainly exist. We have thereforetried to be conservative,by
excluding anecdotalhost plant records.We included a host
plant record only if it was corroboratedby two independent
sources,if the plant was recordedfor more than one species
in the butterfly genus,if there were recordsof more than one
host plant genus from the same plant family, or if it was the
only plant recorded for this butterfly group. Atypical hosts
with little support were excluded.The risk we thereby run
of erroneously excluding host plant data that are correct is,
we believe, outweighed by the advantageof excluding a
greater number of records that are incorrect. The host plant
databaseis provided in Appendix 3.
(Aswere Salicaceae(placedin Rosid lA), Plantaginaceae
terid l), and Cactaceae(Hamamelid l).
There has not yet been any comparableattempt to perform
a combined phylogenetic analysis of the butterflies as a
whole. For this reasonwe have synthesizedresultsfrom several smaller studiesinto a single phylogeny for all butterflies
(Fig. I, Appendix 2). Sourcesfor the different parts of the
phylogeny,and the type of evidencethey presented,arelisted
in Table 1.
The phylogenyhas been resolvedto genericlevel in most
groups,the exceptionsbeing groups with little or no variation
in host use (e.g.,Satyrinae,which has beenresolvedto tribal
level) and groups with large variation in host use and for
which a detailedphylogenyis available(e.g.,Papilio, which
has been resolvedto specieslevel). The level of resolution
Character Coding
is likely to have some effect on the reportedpatterns(Sill6nTullberg 1993,seeResults).The completephylogenyconsists
Optimizing a complex charactersuch as host plant utiliof 431 ingroup taxa, and is given in parentheticalformat in zation is problematic.One major issue is to treat multiple
489
BUTTERFLIES AND PLANTS
T , q . e l - El .
Sources for phylogenetic information regarding the butterflies. Morph, morphological data; Behav, behavioral data; Tax,
,"-"""-t. tr""t'
'Iaxon/tuxa
Level of resolution used by us
Typc 01 data
Papilionoidea
Butterfly families and choice of outgroups
Hesperioidea
Papilionidae
Parnassiini
Graphini, Troidini
Subfamilies
Subfarnilies, trrbes
Genera
Genera, species
Species
Species
Genera and species groups
Morph
rRNA
Morph * mtDNA
Morph
Morph * Behav
Morph
Morph
Morph
Morph
Morph
Morph
Morph
Morph
lnTDNA
Morph
Allozymes
Morph (Tax)
Morph (Tax)
Morph
Morph
Morph (Tax)
Morph
Morph (Tax)
Morph (Tax)
Morph
Morph
Morph
Grctphiurn
Papilionini
Pctpilio
Pieridae
Species
Subfarnilies
Riodinidae
Lycaenidae
Tribes, genera
Genera
Subfamilies
Tribes, genera
Nymphalidae
Danainae
Acreini
Subfamilies, tribes
Tribes, genera
Genera
Genera
Genera
Species groups
associations.
Although most butterflytaxa in our analysisare
restrictedto one plant clade, 36Vouse plants belonging to
more than one clade. There is no method fbr charactercoding
capableof handlingthis problemin a completelysatisfactory
way. However, several approachescan be taken to work
around the problem, all with different problemsand advantages.
One method is to simply optimize plant clade use as an
unordered multistate character with multiple associations
treatedas ambiguity.This makesit easyto interpretpatterns
of multiple host use,but leadsto loss of infbrmation.It does
not help to code multiple associationsas polymorphisms,as
MacClade3 doesnot allow polymorphicstatesto be assigned
to internal nodes in the phylogeny.An alternativeis to optimize each plant clade as an independentbinary character.
This allows easy investigationof specifichost associations,
but makes it difficult to interpret multiple host use and can
leadto falsereconstructions
of nodesashavingno association
at all. Currently the only way to correctly handle multiple
associations,while allowing ancestralpolymorphisms,is to
code host use as a multistatecharactequsing a separatestate
for each host plant associationand, in addition, a separate
state fbr each possible combinationof plant clade associations, and then assign transfbrmationweights using a step
matrix (seeAppendix 4 and Maddison and Maddison 1992,
p. 83). A difficulty with this approachis that the numberof
statesgrows exponentiallywith the number of plant groups.
Limitation on how many statesthe current version (3.05) of
MacCladecan handlerestrictsanalysisto a maximum of four
plant groups simultaneously.The methodcan thereforeonly
be used on either a broad scaleor on subsetsof the butterfly
Referenccs
Kristensen1976
Martin and Pashley 1992
Weller and Pashley1995
de Jong et al. 1996
Scott and Wright 1990
Miller 1987b
H a n c o c k1 9 8 3
Milier 1987b
Munroe1961
Saigusaet al. 1982
Munroe 1961
Hancock 1983
Miller 1987b
Sperling 1992,1993
Ehrlich and Ehrlich 1967
Geiger 1980
Smart 1975
Smart 1975
Ackery 1984
Eliot 1973
Srnart 1975
Scott and Wright 1990
Harvey 199I
Smart 1975
Ackery 1984
Ackery and Vane-Wright 1984
Pierre 1987
hosts. In the following tests we have chosen different approaches to this dilemma, depending on the problem.
Ancestral Host Plant Association
To estimatethe phylogeneticsequenceof butterfly-host
plant associations
we createda multistatecharacterwith multiple host use coded as polymorphisms.To control for the
possibilityof erroneousancestralstateassignmentdue to the
fact that polymorphisms are only allowed at terminals in
MacClade3, we also optimizedthe samedata as a matrix of
binary characters.
Thereis somecontroversyover the degreeof certaintywith
which a characterstatecan be ascribedto an internal node
of a phylogeny (e.9., Frumhotf and Reeve 1994). Optimization of a characterthat changestoo quickly relative to the
branchingpatternof the phylogenywill not result in a plausible historicalreconstruction.To addressthis problem, we
performedseveralrandomizationsto test to what extent our
reconstructionof the ancestralstateis dependenton the frequencyand distributionof characterstatesamongextanttaxa
and on the structureof the butterfly phylogeny.These tests
fbllow the same logic and use the same null hypothesisas
the "permutation tail probability" test (Archie 1989; Faith
a n d C r a n s t o n1 9 9 1 ) .
First, if host plant utilization changestoo fast relative to
speciationthere will be no phylogeneticsignal in this character and ancestralhosts cannot be reconstructedby optimization on the butterfly phylogeny (Frumhoff and Reeve
1994).Reconstructionat the ancestralnode will in that case
be a function of the relative frequenciesof current host as-
490
N. JANZ AND S. NYLIN
sociations.If our reconstructionis an artifact of a host plant
associationbeing commonbecauseit is for somereasoneasy
to colonize,then this associationshouldbe more or lessrandomly distributedamong butterfliestoday. It follows that if
the reconstructionof the ancestralnode using the actualdistribution of character states among living taxa differs significantly from that under a random distribution, given the
samefrequenciesof characterstates,then the ancestralhost
assignmentis more likely to reflect the actual evolutionary
history. To addressthis question,a randomizationtest was
carried out by reassigningthe observedstates1000 times at
random and reconstructingthe ancestralnode each time. The
statesof all extanttaxa were randomized,including the outgroup.
Second,as the butterflyphylogenyis uncertainon several
points, we testedhow sensitiveour reconstructionof the ancestralhost was to the tree topologyby partiallyrandomizing
tree structure.First, we completelyrandomizedthe structure
within eachfamily, only retainingthe structurebetweenfamilies. We thenrepeatedthe randomizationsof brancheswithin
families,while retainingthe most basalbranchin eachfamily.
In eachcase,the ancestralcharacterstateswerereconstructed
for 1000 randomizations.
Afier the above analysishad indicatedthe ancestralhost
plant clade,we further separatedthis cladeinto smallerunits
in an attempt to determine the ancestralplant association
more closely.
Colonization and PhvlogeneticDistance
To test the prediction that host shifts and colonizationsare
more common between closely related host taxa we performed two different analyses.First we analyzedshifls from
the presumably ancestralclade (Rosid 18) to three major
plant groups.This test was carried out on very broad plant
categories,namely colonizationsor shifts from Rosid lB to
the threemajor plant groups "rosids" (otherthan Rosid 1B),
"asterids" and "other plants" (Chaseet al. 1993),making
the step matrix coding describedabove in CharacterCoding
appropriate.Each of these groups and each possiblecombination of them were treatedas separatestates,with 15 in
all. Gains and losseswere given equal weight, that is, a gain
and a loss of one of these plant groups both carried a cost
of one (Appendix 4). Becausestep matricescannotbe used
with unresolvedtrees,the test was carried out on 100 phylogenieswith randomly resolvedpolytomies.
The test hypothesis predicts that shifts from the ancestral
host plant should most commonly be to plants in the same
major clade (rosids)and least commonly to plantsmost distantly related to the original host (other seed plants). The
number of unambiguouschangesfrom feeding on Rosid lB
to feeding on the three groups were counted using the step
matrix and the "chart state changes" option in MacClade.
We testedthesenumbersagainstthe null hypothesisthat colonizations should be equally distributed among the three
groups (one-third to each), assumingthey are of approximately equal diversity and thus provide equally sized "targets" for random colonization. Accurate and meaningful
measuresof diversityfor thesegroupsarevery hardto obtain,
as thev do not easilv translateto the traditionaltaxonomical
groups.However,using the number of plant families in our
Appendix I as a crude measureof diversity,this assumption
seemsreasonable:thereare 54 rosid families.47 asteridfamilies, and 51 familiesof other seedplantsamongthe butterfly
hosts.
As there is always variation betweenresolutionswe first
neededto know whetherthe found differenceswere consistent over the resolutions.In this test and in similar teststhat
follow, we have used an approachto the problem of irresolution similar to Losos (1994) and Martins (1996).For each
randomresolutionwe calculatedthe pairwisedifferencesbetween number of colonizationsof rosids.asterids,and other
seedplants from the ancestralclade Rosid 1B. For instance,
if there were 35 colonizationsof other rosids.23 of asterids.
and 17 of other seedplantsin a given resolution,the pairwise
diff'erencebetween rosids and asteridswould be +12. betweenrosidsand other seedplants * 18, and betweenasterids
and other seed plants +6. The distribution of these differenceswere examinedto get an estimationof the number of
resolutionswherethe hypothesiswas corroboratedorrefuted.
ln our exampleall differenceswere in accordancewith the
hypothesis,as they were all positive.Note that this procedure
only gives an estimateof the consistencyof the found differenceover the examinedphylogenies,the magnitudeof the
diff'erencemay still be small enoughto be a result of chance.
We therefore performed chi-square tests of fit to the null
hypothesismentionedabove, both on the averagenumbers
of colonizationsover the 100 resolutionsand on each individual resolution.
ln the other analysiswe investigatedcolonizationswithin
and betweenthe two large sister groups "rosids" and "asterids." We comparedthe number of colonizationsto and
from plantsbelongingto the samegroup (rosidsor asterids)
with the numberof colonizationsbetweenthesegroups,making no assumptionabout the ancestralhost plant. To make
the test as conservativeas possible,we excludedall changes
within the terminal clades of the Chase et al. (1993) phylogeny (Rosid 1-3 and Asterid 1-5), counting only changes
betweentheseratherlargecladesas changes"within rosids"
and "within asterids." The numbersof unambiguouscoloagainstthe null hynizationswere testedby goodness-of-Iit,
pothesisthat the numberof changeswithin asteridsor rosids
and betweenthem shouldbe equal.The rationaleis that since
we assumethat the two groups are approximatelyequally
diverse.if colonizationsare random, half should be to the
other major clade.As the total numberof colonizationscould
be summed using binary characters,there was no need for
step matrix coding and henceno needto resolvepolytomies.
To identify groups of unrelatedplants that often occur together as butterfly host plants, we constructeda seed plant
phylogenyusing only presenceor absenceof butterfly taxa
as characters,using a method similar to what is often used
in cospeciationstudies(e.g.,Patersonet al. 1993).Thesewere
analyzedusing PAUP 3.1.1 (Swofford 1991) with detault
"factory" settings(simple addition sequence,one tree held
at each step during stepwiseaddition, tree bisection-reconnection[TBR] swappingalgorithm,MULPARS option in effect, no topological constraints).We compared the plant
groups suggestedby this analysis with the groups in the
Chaseet al. (1993) phylogeny.The butterfly charactersand
491
BUTTERFLIES AND PLANTS
plant taxa fbllow the matrix shown in Appendix 3. We used
a hypotheticalancestorwith zero (no butterfly associations)
as outgroup,as the origin of angiospermspredatesthe butterflies,and we did not want to put a constrainton what plant
groups could be united by the analysis.Well-definedgroups
in this analysisthat are not supportedby the Chase et al.
phylogeny were interpretedas host shift tendenciesnot correspondingto plant phylogeny.
Plant Ph1'logen\versus Gron-th Fornt
To assesswhether plant phylogeny or growth form is the
more evolutionarilyconservativeaspectofbutterfly host use,
we comparedthe number of stepsneededto trace each on
the butterfly phylogeny,using a coding that assignsan equal
number of statesto both features.The categorieswe used
were for plant groups, "rosids," "asterids," and "other
plants," and for growth forms, "herbs," "vines," and "trees
and shrubs." Plant group and growth form were coded as
multistate characterswith one state for each category and a
separatestatefor eachcombinationofcategories(sevenstates
in total). Transformationweights were assignedusing step
matrices(seeaboveand Appendix 3). The numbersof steps
neededto trace the two characterswere averagedover 100
phylogenieswith randomly resolvedpolytomies,and tested
by goodness-of'-fit
againstthe null hypothesisof equal numbers,that is, thatplant growth form andphylogenyareequally
conservativeaspectsof the association.To investigatethe
consistencyof the diffbrencesin the numberof statesneeded
to tracethe two characters.their distributionswere examined
over the randomresolutions(seeabove).This analysiscould
potentiallybe influencednot only by the diversityofthe plant
clades(seeabove),but also by the "availability" ofdiffbrent
growth forms for colonization. However, at least on the family level, the distribution of growth forms does not seemto
be alarminglyunequal,amongthe families listedin the Cronquist system on the Flowering Plant Gateway website,282
containtreesor shrubs,9l vines,and 208 herbs(Watsonand
Dallawitz 1992).
The characterstatesare very unequalin frequency,with
strong bias toward feeding on the plant group "rosids" and
the growth form "trees." The question therefbrearisesof
whether there is a different tendencyto shift betweenplant
groups while feeding on trees or to shift between growth
forms while feeding on rosids.
We used two proceduresto test this. First, the possibility
that changesin plant clade use are more likely to occur on
a certain growth form (and vice versa)was testedwith the
concentrated-changes
test (Maddison 1990; Maddison and
Maddison 1992). Specificallywe testedif major host shifts
were concentratedto branchesreconstructedas woody-plant
1-eeders,
and if shifls between growth fbrms are concentrated
to branchesreconstructedas feedingon rosids.Thus the test
was repeatedtwice, using "plant group" and "growth form"
in turn as the dependentvariable.As the test cannothandle
multistatecharactersthe characterswere recodedas binary.
All taxa that use rosids were coded as rosid feeders(regardlessof what else they feed on). The other statethus represent
butterfly taxa that do not feed on rosids. Growth forms were
treated the same way. To ensurethat only complete shifts
were included in the analysis(i.e., when a colonizationis
followed by specialization
on the novel plant),we only counted a shift if no speciesin the butterfly taxon hasretainedthe
old association.Shifts were countedusing stepmatrix coding
of plant use and the "chart all changes"option in MacClade.
The concentrated-changes
test can only be used on completely resolvedphylogenies,so the test was iteratedover
100 butterfly phylogenieswith randomly resolved polytomies. MacClade(Maddisonand Maddison 1992)usesa simulation algorithm to calculatethe statisticson large phylogenies and our calculationswere basedon 1000 such simulations for each randomly resolved tree, using the default
settingsof the program (allowing either stateto be ancestral
and using actual changes).
As an alternativetest of the associationbetweentree-feeding and tendencyto host shift, we counted the number of
terminalplant cladesfrom the Chaseet al. (1993)phylogeny
that were used as hostsby the terminal taxa in our butterfly
phylogeny.If this measureof "host use diversity" is treated
as a continuouscharacter,its correlationwith tree feeding
can be assessedby the approachof independentcontrasts
(e.g. Fefsenstein1985; Pagel 1992; Purvis and Rambaut
1995).As the butterfly terminal taxa are most often genera,
taxa with multiple host plant associations
may representcollectionsof speciesspecializedon differentplants,polyphagy
of individual species,or both. In any case, "host use diversity" should approximatethe number of host colonizations
that havetakenplacewithin that taxon,regardlessof whether
thesehave led to increasesin host rangeor to divergencein
h o s t p l a n t u s e a m o n gs p e c i e s .
Thesedatawereanalyzedwith ComparativeAnalysisusing
IndependentContrasts(CAIC; Purvis and Rambaut 1995),
using the "brunch" algorithm, which is designedto test if
changesin a discretecharacter(like tree feeding) are associated with changes in a continuous character (like host
range).This test differs fiom the concentratedchangestest
in simply testingwhetherchangesin two charactersare correlated,without identifying one characteras logically independentor causative.One advantageof CAIC is that it has
an algorithmfor handlingpolytomies.Detailsof the testscan
be fbund in Pagel (1992) and Purvis and Rambaut (1995).
The contrastsgeneratedby CAIC were testedboth qualitatively with a sign test or quantitativelywith a one-samplettest (seeHoglund and Sill6n-Tullberg1994).
CAIC usesexplicit assumptionsabout branchlengthsand
offbrs two default alternatives:all branch lengths assumed
equal (correspondingto a punctuationalmodel of evolution)
or ages of taxa assumedproportionalto the number of included species(correspondingto a gradualmodel of evolution), using an algorithm by Grafen (1989). As we had no
information on branch leneths.we used both.
Rgst,'L.ls
Patterns o.fHost Plant Utilization
Butterflies use almost all major seed plant families, and
even a few nonseedplants, some speciesdo not even 1-eed
on plants(Fig. 2). Nevertheless,
as waspointedout by Ehrlich
and Raven(1964),the useofplant cladesappearsnonrandom.
Some families are heavily utilized by many butterfly groups
492
N. JANZ AND S. NYLIN
Cycads (4)
Ginkgo (0)
Pinaceae(3)
otherconifers(1)
Gnetales(0)
Ceratophyllum (0)
paleoherb'l (16)
Laurales(29)
monocots(36)
Magnoliales(20)
paleoherb2 (0)
ranunculids(8)
hamamelid1 (9)
hamamelid2 (0)
Gunnera (0)
(44)
caryophyllids
asterid5 (3)
asterid4 (2)
asterid3 (38)
asterid2 (32)
asterid 1 (90)
Luthrodesin Polyommatinae).Other examplesinclude Asteraceae,which, consideringits size,is also relativelyrarely
used by butterflies.
Further supportingEhrlich and Raven's(1964) contention
that relatedbutterfliestendto feedon relatedgroupsofplants,
the Chaseet al. (1993) phylogenysuggeststhat some ostensibly disparatehost plant assemblages
are more phylogenetically homogenousthan previous taxonomy suggested.For
example, under the classificationof Cronquist (1981), the
diversehost list of the nymphalidtribe Nymphalini includes
as dominant themes three families in the order Urticales
(Hamamelidae),two in Rosales(Rosidae),one in Salicales
(Dillenidae),and two in Fagales(Hamamelidae).In addition
there are speciesfeedingon Ericales(Dillenidae),Asterales.
(Asteridae)Rhamnales(Rosidae),Malvales(Dillenidae),and
L i l i a l e s ( L i l i i d a e ) .I n t h e C h a s ee t a l . ( 1 9 9 3 )a n a l y s i sh, o w ever, the most tiequently used groups; Rosaceae(Rosales),
Urticales, Salicales,Fagales and Rhamnales,formerly in
three subclasses.
all fell within the Rosid I subclade.while
Malvales belongedto the sister group Rosid 2 and Grossulariaceae(Rosales)to Rosid 3. Thus, the nine plant orders
of main nymphalidhosts,previouslyscatteredover five subclassesprobably have relatively close affinities.
Ancestral Host Plant Association
The optimizationof host use as a multistatecharacterand
as binary charactersboth suggestthat the ancestralhost plant
r o s i d2 ( 1 1 0 )
was located within the clade we have called "Rosid lB"
(Fig. 1, seeAppendix I for a list of includedfamilies).This
rosid 1F (0)
was the only cladethat even camecloseto being drawn back
rosid 1E (22)
to the root of the butterfly phylogeny. Plant taxa such as
rosid 1D (4)
Rosid 2 and Asterid I (seeAppendix l), also usedby many
butterflies(Fig. 2), are apically distributedon the butterfly
rosid 1C (32)
phylogeny.
r o s i d1 B ( 1 7 0 )o
Hesperioidea,the most likely sister group, was used as
rosid 1A (92)
outgroup for the optimizations.This group is mainly associatedwith Fabaceae(Rosid lB) and monocotyledons.
Ftc;. 2. Phylogenetic relationships among seed pltrnts fiom Chase
If the
et al. (1993). The names Ros 1A to E are assignedby us to subclades
sister group to Papilionoideainsteadturns out to be Hedyidentified by Chase et al. (1993). Numbers of associatedtaxa in the
loidea,the reconstructionwill be somewhatmore uncertain.
butterfly phylogeny are indicated after plant clade names. The dot
plant data on Hedyloideaare scarcebut indicate that
Host
d e n o t e s p r e s u r . n e da n c e s t r a l h o s t p l a n t c l a d e .
Sterculiaceae(Rosid 2) is the most important host plant
group. Sterculiaceae
is a memberof Malvales,suggestedby
(e.g.,Fabaceae),while othersare dorninanthostsfbr partic- Ackery (1991) to be the ancestralhost group for butterflies.
ular subsetsof butterflies.In Papilionidae,Aristolochiaceae However,even in this caseour optimizationssuggestthat the
(Paleoherb1) and Rutaceae(Rosid 2) are clearly dominating colonizationsof Rosid 2 by Hedylidae, some Hesperiidae,
themes,while Pieridae are typically associatedwith plants and some Papilionidaeare independentevolutionaryevents
in Brassicaceae
and relatedfamilies, such as Capparidaceae and that Rosid lB is the most probableancestralhost plant
(Rosid 2). Fabaceae(Rosid 1B) are common hosts in Rio- group. The same is true if Hedyloideais used as the only
dinidae and Lycaenidae,and perhapsUrticales (Rosid 1B) outgroupor if no externaloutgroupis used at all.
The reconstructedancestralnode appearsto be signifitogetherwith Passifloraceae
and relatedfamilies (Rosid 1A)
could be said to be predominanthosts in Nymphalidae,al- cantly different from reconstructionsunder random character
thoughthis is a very large generalization.
Other,equallycon- stateassignment.Rosid lB f-eedingwas ascribedto the anspicuousplant groups.however,are only rarely usedas hosts cestralnode in only 49 of 1000randomizations(P : 0.049).
by butterflies.The very largefamily Orchidaceaeis only used The relatively high proportion of Rosid 1B feeding on the
by a few generain Lycaenidae(Hypolycaenaand Chliaria in phylogeny (39c/o)is thereforenot a sufficient explanationfor
Theclinae)and Nymphalidae (Faunis in Amathusiinae).Like- its reconstructionas the ancestralstate.Moreover,as the diswise, gymnospermsare usedby only a handful of speciesin tribution of Rosid 1B f'eedersin the phylogenyis nonrandom,
Pieridae(Neophasiain Pierinae)and Lycaenidae(Callophrys, it is unlikely that the widespreaduse of this group (and the
Eumaeus. and Stn'mon in Theclinae. and Theclinesthesand reconstructionof the ancestralstate) should simplv fbllow
r o s i d3 ( 1 3 )
493
BUTTERFLIES AND PLANTS
120
50
100
a 4 0
c
@
c
.F
.N
o
^^
{tl
6 " "
o
,N
C
o
o
BO
60
O
o o n
40
c
o)
> 1 0
20
RosidsAsterids Other
plants
Within Between
Ftrt. 3. Association between likelihood of colonizations and phylogenetic distance. (a) Mean number of colonizations from the ancestral
clade Rosid lB to three large plant groups of approximately equal diversity, averaged over 100 phylogenies with randomly resolved
polytomies (see text). (b) Number of colonizations of plants belonging to the same major clade, that is, from a rosid to a rosid or an
asterid to an asterid, compared wrth the number of coionizations liom an asterid to a rosid or vice versa.
from the fact that Rosid lB might fbr some reason be easy
to colonize. A historical explanation of the utilization of at
least this plant group is therefbre more probable.
It follows from this result alone that the phylogenetic structure must be more important fbr the reconstruction of the
ancestral association than simply frequency of use of each
plant taxon. A thousand random resolutions of all branches
within the butterfly families (keeping only the structure between families) generated only 30 cases where utilization of
Rosid lB was ascribed to the butterfly ancestor (P : 0.03),
indicating that this reconstruction is a very improbable outcome if we regard the phylogenetic relationships within butterfly families as completely unknown. Howeveq it is enough
to retain the most basal branch in each family and randomly
resolve all branches above to make Rosid lB the most likely
ancestral association. This means that the reconstruction of
the basal branches within each family is critical for the reconstruction of the ancestral butterfly-host plant association.
Atler optimizing a multistate character where Rosid lB
had been further divided into two subclades (1B1: Fabaceae
and Polygalaceae; 182: remaining families, see Appendix l)
indicated that the ancestral host plant family most likely was
Fabaceae (as Polygalaceae is a very minor host). The result
of this more detailed analysis was not used further, and for
this reason it was not put through the randomization tests
described above.
Colonization cmd Phylogenetic Distance
A comparison of the number of unambiguous changes fiom
what we suggest to be the original host plant clade (Rosid
lB) to the three large groups of roughly equal size, "other
rosids," "asterids," and "other seed plants," gives some
support to the hypothesis that the probability of a host shift
is related to the phylogenetic distance between the plant
groups involved (Fig. 3a). There was no single resolution
where the number of colonizations of rosids was fewer than
the number of colonizations of asterids and "other plants."
In only fbur resolutionsout of 100 there were more colonizationsof "other plants" than of asterids.Thus, we conclude that the differenceswere consistentover the phylogenies with randomly resolvedphylogenies.
The numberof colonizationsof other rosidsaveraged35.0
(range:25-49) over 100phylogerrieswith randomlyresolved
polytomies.Colonizationsof asterids,sistergroup to the rosids, averaged23.2 (16-33), while colonizationsof any plant
outside the rosids and asteridsonly averagedL7.4 (8-23).
These averagenumbers depart significantly from the null
hypothesisof equal proportions (Xr : 6.3S, df : 2, P :
0.041). Howeveq among the 100 actual resolutions,there
were 39 where the diffbrencesin colonizationsproved nonsignificant.Thus we cannot safely conclude that colonizations are not evenly distributedamong groups in the "true"
resolution,even though they are consistentin direction.
Comparisonsof shilisjust betweenrosidsandasteridsgave
strongersupportto the hypothesis(Fig. 3b). The total number
of colonizationsbetweenthese groups accountedfor 62 of
173 unambiguouschangeson the phylogeny,or 35.87o,while
11I shifts occurredamong plant cladeswithin rosids or ast e r i d s ( 1 r - 1 3 . 8 8 ,d f : I , P < 0 . 0 0 1 ) .T h i s w a s a v e r y
conservativetest, as shifts within the terminal plant clades
o f C h a s ee t a l . ( 1 9 9 3 )w e r e n o t c o u n t e d .
Ehrlich and Raven (1964) noted that some groups of genealogicallyunrelatedplantsoften occurredtogetherashosts,
suggestingan underlyingchemicalconvergence.
Our analysis
using the butterfly clades as characterssuggestsadditional
such convergencesextendingEhrlich and Raven'sobservation. The most notablegrouping of unrelatedplants by butterflies was utilization of plants in Asterid I togetherwith
plants in Rosid I and 2. Twenty-sevenbutterfly taxa in our
analysisuse plants in all thesecladesas hosts,representing
between l8 and 23 independentevolutionary events.The
plant families usedwere most often Rosaceaeand Ulmaceae
in Rosid 1, Rutaceae,Tiliaceae,and Sapindaceae/Sterculiaceaein Rosid 2, and Oleaceae,Rubiaceae,and Verbenaceae
494
N. JANZ AND S. NYLIN
40
24
JC
l l
a
E
@
18
@
a
G
30
tc
C
25
c
o
E
c
C)
o
c)
E
C
20
6
c
'o
d
tt
I
6
tc
10
5
3
0
0
Onwoody Onherbaceous
plants
plants
Positive
b
Ftr;. 4. Association between feeding on woody plants and host shifts or colonizations. (a) Mean number of shifts between rosids and
other plants when feeding on woody versus herbaceous plants, averaged over 100 phylogenies with randomly resolved polytomies. (b)
Independent contrasts between tree- and nontree-feeding lineages. Positive contrasts are those in which the tree-feecling iineage uses
m o r e _h o s t s c l a d e s p e r b u t t e r f l y t a x o n . B l a c k b a r s r e p r e s e n t r e s u l t s u s i n g s c a l e d b r a n c h l e n g t h s , a n d s h a d e d b a r s r e p r e s e n i r e s u l L s u s i n g
equal branch lengths (see text).
in Asterid l Other, less strongly supported groupings in our
analysis that differed from those of Chase et al. (1993) include Rosid 3 united with Asterid 3 (5-6 independent evenrs),
and Asterid 2 united with ranunculids (3 independent events).
Plant Phy-log€n\t urrtu"- Growth Form
The mean number of steps needed to trace the character
"major plant groups" (rosids, asterids and other plants) on
100 phylogenies with randomly resolved polytomies was
217.1 (range 210-226), while it was 168.6 (range 163-174)
fbr the character "growth form." This result departs significantly from the null hypothesis that changes in growth form
and plant clade are equally common (X2 : 0.10, df : l, P
: 0.014). The difference is consistent as the distributions of
numbers of steps for the two characters acrossrandomizations
do not overlap. This indicates that, even at this low level of
plant group resolution, growth form may in fact be a more
evolutionarily conservative aspect of butterfly-host plant associations than plant phylogeny.
There is, howeveq one plant clade and one growth fbrm
that are much more widely utilized than the others. Rosids
are used as host plants by 69c/oof the butterfly taxa, while
asterids and other plants are used by only 30o/o and 35Vo,
respectively. Likewise, 73% of the butterfly taxa feed on trees
and/or shrubs, whrl,e 2lo/o f'eed on vines and 31Vo feed on
herbs. Thus, a higher tendency to shift between plant clades
when feeding on trees would increase the apparent average
rate of host plant shifi. Likewise, a lower tendency to shift
between growth forms when f'eeding on rosids could decrease
the apparent rate of shifts among growth forms.
Interpretation of the concentrated changes test is complicated, as the results fiom the 100 random resolutions polytomies vary substantially (Fig. 4a). For a majority of resolutions (61), the number of major host shifts taking place on
branches characterized by feeding on trees and shrubs is higher than expected by chance (P < 0.05). However, the P-values
ranged from 0.002 to 0.349.
Not surprisingly,therewas no evidenceat all for the complementaryhypothesis,that feeding on rosid plants should
discourage shifts between different growth forms, as in no
randomizationwas the P-value below 0.1.
The independentcontraststest showeda strongassociation
betweentree feedingand use of an increasednumberof host
plant cladesby butterfly taxa, giving further supportto the
hypothesisthat colonizationsof new host plants are facilitated by f'eedingon trees (Fig. 4b, exemplified in Fig. 5).
This relationshipwas highly significantunder both rhe sign
test and the one-sample/-test (Table2). The trend was consistentover all butterfly families, althoughnot significantin
all families,probablydue to small numbersof contrasts.The
two algorithmsfor estimatingbranchlengthsgave very similar results.
A possibleproblemwith thesetestsis that diversityofhost
use and taxon size are confounded.One can not entirelyrule
out the possibility that tree-feedingtaxa in generalcontain
more speciesthat herb-feedingtaxa.
DrscussroN
Ehrlich and Raven's(1964)main observations,
that related
butterfliesoften feed on related host plants, and that some
plant groups are commonly used by butterflieswhile other
large plant groupsare not, are upheldby our reanalysis(Fig.
2). As others have noted (e.g., Jermy 1976, 1984), these
patternscan be explainedby sequentialcolonizationof related plant groups without the coevolutionarytwist that the
plants have escapedand radiatedin the absenceof butterfly
f'eeding.Under a coevolutionaryinterpretation,underutilization of someplant groupsby butterflieswould be ascribed
to chemical det'encesevolved to exclude butterflies fiom
feeding.ln the light of our results,however,it is not likely
that such groups,which are also much older than the butterflies,have ever been usedby butterflies.Some may have
evolved chemicaldefensesagainstother herbivoresthat are
equally effectiveagainstbutterflies.Others,especiallythose
495
BUTTERFLIES AND PLANTS
Battus
Pharmacophagus
-.
-.
-.
*
Euryades
Cressida
Pachilopta
LOSana
-. Trogonoptera
'. Troides
-. Panosmia
-. Atropaneura
'.
'.
-.
Parides
P a p i l i oi n d r a
Papilio brevicauda
P a p i l i om a c h a o n
P a p i i l op o l y x e n e s
*
N P a p i l i oz e l i c a o n
- . P a p i l i ox u t h u s g r o u p
s P a p i l i op r i n c e p sg r o u p
-. Papilio drurya group
-' Papilio demolion group
r c P a p i l i o ' f u s c u sg' r o u p
- . P a p i l i o ' p r o t e n o rg' r o u p
r u P a p i l i o ' p o l y t e sg' r o u p
- P a p i l i oh e r a c l i d e sg r o u p
- - P a p i l i oc h i l a s ag r o u p
TeeLn 2. Relationship between tree l'eeding and number of plant
clades used by butterfly taxa, calculated using independent contrasts. Two different algorithrns have been used to estimate branch
lengths: (l) all branch lengths equal; or (2) branch lengths scaled
so that the ages of taxa are proportional to the number of species
t.:rcontain
a-lest
Taxon
Papilionoidea
Papilionidae
Pieridae
Riodinidae
Lycaenidae
Nymphalidae
Branch
I c n g lh \
Equal
Scaled
Equal
Scaled
Equai
Scaled
Equal
Scaled
Equal
Scaled
Equal
Scaled
/
df
Qirrn tecr
Probability
< 0.001
< 0.001
0.2s9
0 .1 6 6
0.201
0.2s5
*
0
*
2.91 t 9
0.009
2.29 1 9
0.034
2.93 2 6
0.007
2.88 26
0.008
4.23 6 2
3.77 62
1.27 5
1.62 5
l 4l
7
|.24
7
*
0
*
0
pos/rrcg
conrrasrs
Prohability
34/8 < 0.001
36/6 < 0.001
2/r
l 000
3/O
0.248
2/o
0.480
3/O
0.248
/
0
*
*
l/0
tt/3
0.061
It/3
0.061
t]/4
0.009
18/3
0.002
'i Could
not be calculated.
fects butterfly host selection.Another prediction,that may
be easierto test, is that if theseplants do sharesuch a trait,
they should also be linked in the diets of other groups of
P a p i l i ot r o i l u s
phytophagous
insects.
P a p i l i op i l u m n u s
Consideringthe variation in dominatinghost taxa among
Papilio palamedes
- P a p i l i oc a n a d e n s i s
butterfly families, it is somewhatremarkablethat the most
- P a p i l i og l a u c u s
basal branchesin each family feed on plants in Rosid lB,
s P a p i l i om u l t i c a u d a t u s
such as Fabaceae,Urticaceae,Ulmaceae,or Rosaceae.The
N P a p i l i oe u r y m e d o n
Rosid I B cladeis also by f ar the most likely ro haveincluded
r P a p i l i or u t u l u s
the ancestralbutterfly host. The characterstaterandomizations strongly suggestthat this pattern does reflect evoluFIc. 5. Association between tree feeding and propensity fclr host
shifts and colonizations exemplified by Troidini and Papilio in Pa'
tionary history, not this plant group being fbr some reason
pilionidae. White branches indicate herb-feeding, black branches
unusuallyeasy to colonize.Even if Rosid 1B were easierto
woody-plnnt-feeding lineages. Number of plant clades from Chase
colonize,
therecould still be a historicalexplanation:all but(
1
9
9
3
)
et al.
used as hosts by each butterfly taxon are indicated
terfliesmay be literally preadaptedto the chemicaland other
below butterfly names. One of the two to three shifts between rosids
and other plants, taking place on herb-feeding branches, and three
traitsof theseplants,simply becausetheir ancestorsfed upon
of the 13 32 such shitis taking place on woody plant-feeding
plantscontainingthem. Furthersubdivisionof Rosid I B corbranches, are shown in this figure (bars on branches). The figure
roboratesScott's (1986) suggestionthat Fabaceaewas the
also shows three of the ,12independent contrasts between herb- and
most likely ancestralhost plant family.
t r e e - f e e d i n g t a x a . I n a l l 3 c o n t r a s t s s h o w n t h e t r e e - 1 ' e e d i n gt a x a a r e
The strongconservatismof butterfly associationwith maassociated with higher numbers of plant clades than are the herbfeeding taxa.
jor plant cladesdoes not precludefrequentshifts among related host species.Indeed,many butterfliesfeed on several
speciesor generawithin the same plant family, suggesting
very distantlyrelatedto the original butterflyhost,may sim- that therehavebeenmany colonizationeventsbetweenplants
ply not have been "discovered" yet by butterflies,as such too closely relatedto be distinguishedby our analysis.
distant colonizationsappear to have been rare. Given the
Restrictionof most colonizationsto related plants could
current stateof knowledge,it is impossibleto give any def- reflectconstraintson geneticvariationin the capacityto feed
inite explanationto thesepatterns.
on novel hostplants,making a shift to an ancestralhost plant
Our analysisalso extendsEhrlich and Raven'sobservation more likely than to a completelynovel plant (Futuymal99l).
that some groups of unrelatedplants repeatedlyco-occurin There is some evidencefor this in chrysomelidleaf beetles
the host lists of butterfly taxa, which they interpretedas re- (Futuymaet al. 1993, 1994, 1995)and in burterflies(N. Janz
flecting chemicalor other underlyingsimilarity.The clearest and S. Nylin, unpubl.).In fact, one of the examplesEhrlich
prediction liom our analysisis that plants in Rosid I (most and Raven (1964) give in their paper on colonizationof reimportantly Rosaceaeand Ulmaceae),Rosid 2 (most impor- lated plant groupsby relatedbutterfliescould, in the lighr of
tantly Rutaceae,Tiliaceae and Sapindaceae/Sterculiaceae),
the new plant phylogeny,be better understoodas a recolonand Asterid I (most importantly Oleaceae,Rubiaceae,and ization of the ancestralhost plant clade: the switch of one
Verbenaceae)share some chemical or other feature that af- genus in Parnassiini (Htpermnestrd) from the dominant Ar-
ru
-.
-.
-.
P a p i l i oe l e p p o n eg r o u p
Papilio pyrrhostictagroup
496
N. JANZ AND S. NYLIN
AsnnnsoN, R. S. 1993. Weevils and plants phylogeneticversus
ecologicalrnediationof evolution of host plant associationsin
Curculioninae(Coleoptera,Curculionidae).Mem. Entornol.Soc.
Can. 165:197-232.
AncHtn, J. W. 1989. A randomizationtest for phyiogeneticinforrnationin systematicdata. Syst. Zool. 38:239-252.
Brur,t, D. 1994. rbcl- and seed-plantphylogeny.TrendsEcol. Evol.
9:39-41.
B n N s o NW
, . W . , K . S . B n o w s , . t x n L . E , .G r r - e r : n r .[ 9 7 5 . C o e v o lution of plants and herbivores:passionllower butterflies.Evolution29:659-680.
BrnsNseur,t,M. R. 1983. Coumarinsand caterpillars:a case for
coevolution.Evolution 3'7:163 119.
Br.nNevs,E. A., eiru M. Gneuen. 1988. On the evolutionof host
specificityin phytophagousarthropods.Ecology 69:886 892.
Bnoors, D. R., eNo D. H. Mcl-ulrN,qr. 1991. Phylogeny,ecology,
and behavior.Univ. of Chicago Press,Chicago.
C u e s p ,M . W . , D . E . S o r - r r sR
, . G . O r - r r s r n e oD
, . MonceN,D. H.
LEs, B. D. MrsHr-ln,M. R. Drrv,c.r-r-,
R. A. Pnrca,H. G. Hrlrs.
Y . Q r u , K . A . K n o N , J . H . R n r r r c , E . C o x r r , J . D . P e r . u e n .J .
R. MeNuenr, K. J. SvrsNl,q,H. J. Mrcuells, W. J. Knuss, K.
G . K e p o r . ,W . D . C r - . a , nM
r , . H n o n E u ,B . S . G r u r , R . K . J a N s n r ,
K . K r r r rC
, . F W r r a p e rJ, . F S m r l r r ,G . R . F u n N r e nF, H . S l n e u s s ,
S. E.
Q. Xr,+Nc,G. M. Pr LrNKErr,P. S. Sor-rrs,S. M. SrvnNsr-N,
Wrr,r-r,qr,rs,
P. A. Geoex, C. J. QurNN,L. E. Er;t-'r,cRru,
E. GoLnNSERG
G,. H . L n a p N ,J n , S . W . G n , r H r : v ,S . C . H . B a r < n n r r ,
Ar-t) V. A. AlBEnr. 1993. Phylogeneticsof
S. D,cvls,ANoA^-,
from the plastid
seedplants:an analysisof nucleotidesequences
gene rbcl. Ann. Mo. Bot. Gard. 80:528-580.
hypotheses.
CoootNt;roN,J. A. 1988. Cladistictestsof adaptational
Cladistics4:3-22.
1994. The roles of hornologyand convergencein studies
of adaptation.Pp. 53 78 in P.Eggletonand R. Vane-Wright,eds.
Phylogeneticsand ecology.AcadernicPress,London.
C o n u c r N ,I . E B . , , r N o D . F W r r E n H o r r s r . ,1. 9 7 2 . B u t t e r f l i e so f
Australia. Angus and Robertson,Sydney, New South Waies,
Australia.
1992. The butterfliesof
Conesr, A. S., eNo H. M. Pl,r.'or-r,suRy.
the Malay Peninsula.Malayan Nature Society,Kuala Lumpur.
CorrnrNsv,S. P. 1984. Habitat versusfoodplantselection.Pp. 8990 in R. I. Vane-Wrightand P R. Ackery. eds. The biology of
butterflies.AcadernicPress,Orlando,FL.
CnoNeutsr, A. 1981. An rntegratedsystem of classificationof
flowering plants.Columbia Univ. Press,New York.
o e J o N c ,R . , R . I . V a N n - W n r c H r , r x o P R . A c r e n i " . 1 9 9 6 .T h e
higher classificationof butterflies(Lepidoptera):problemsand
p r o s p e c t sE. n t o m o l .S c a n d .2 7 : 6 - 5l 0 l .
on t-r Mrzr R,rvines, R. 1987. Mariposasmexicanas.Fondo de
Cultura Econ6mica,Mexico City.
DEVRIEs.P.J. 1987. The butterfliesof CostaRica and their natural
history, Papilionidae,Pieridae,Nyrnphalidae.Princeton Univ.
Press,Princeton,NJ.
EsLnr, G. 1993. Die Schrnetterlinge
Baden-Wiirtternbergs.
Verlag
Ac:rNowleociML,N'rs
Eugen Ulmer, Stuttgart, Germany.
EHnrrcn, P R., nNo A. H. Eunlrcu. 1967. The pheneticrelationWe wish to thank O. Eriksson, C. Mitter, B. Tullberg, C.
shipsof the butterflies.I. Adult taxonomyand the non-specificity
Wiklund, and three anonymous referees fbr valuable comh y p o t h e s i sS. y s t .Z o o l . 1 6 : 3 0 1 - 3 1 7 .
ments on the manuscript. This work was supported by a grant Eunlrcu, P. R., eru P. H. Rnves. 1964. Butterfliesand plants: a
study in coevolution.Evolution 18:586 608.
from the Swedish National Science Research Council to SN.
ELtor, J. N. 1973. The higher classificationof the Lycaenidae
(Lepidoptera):a tentative arrangement.Bull. Br. Mus. (Nat.
Lrrrneruns Ctt'pn
H i s t . ) E n t o m o l .2 8 : 3 7 3 5 0 6 .
Acxsnv, P R. 1984. Systernaticand faunisticstudieson butterflies. E u v e r , T . C . , M . C . M r N N o ,e N o B . A . D n u r . n t o N o .1 9 9 2 . F l o r issantbutterflies:a guide to the fossil and present-dayspecies
Pp. 9 2l rrrR. I. Vane-Wrightand P R. Ackery, eds.The biology
of centralCoiorado.StanfordUniv. Press.Stanford,CA.
of butterflies:their cladisticsand biology. British Museum (NatE n r x s s o NO
, . , e N o B . B n e m s n .1 9 9 2 .P o l l i n a t i o ns y s t e m sd, i s p e r s a l
u r a l H i s t o r yt . L o n d o n .
modes, life fonns, and diversification rates in angiospermfama review ofnyrnphalid
1988. Host plantsand classification:
ilies. Evolution 46:258-266.
b u t t e r f l i e sB. i o l . J . L i n n . S o c . 3 3 : 9 5 - 2 0 3 .
D. P.,ero P.Cn,qNsroN.1991. Could a cladogramthis short
1991. Hostplantutilizationby Atiican and Australianbut- F,A.rrH,
have arisenby chancealone?On permutationtestsfor cladistic
terfliesB
. i o l . J . L i n n . S o c .4 4 : 3 3 5 - 3 5 1 .
structure.Cladistics1 : l-28.
Acrsny, P. R., r.ro R. I. VeNr-Wnrcul. 1984. Milkweed butterflies: their cladisticsand biology. British Museum (NaturalHis- FnnNv,P. 1975. Biochemicalcoevolutionbetweenplantsand their
insect herbivores.Pp. 5-19 itt L. E. Gilbert and P. H. Raven,
tory), London.
istolochiaceae (Pal l) and Rutaceae (Rosid 2) theme to feed
on Zygophyllaceae (Rosid 1B), which Ehrlich and Raven
claimed to be closely related to Rutaceae. If such recolonizations are common, it means that a high number of host
shifts could go unnoticed in a phylogenetic study, because
they tend to shift back to the original host, thus making the
overall pattern look more conservative than it is on a finer
level. Or, put in another way, an opportunistic pattern of host
plant utilization on a microevolutionary scale may well result
in a conservative pattern on a macroevolutionary scale.
The fact that plant growth form was the more conservative
aspect of host associations in our analysis suggeststhat other
factors than plant chemistry, such as habitat or community
structure, play an important role in shaping the large-scale
patterns of butterfly-host plant association. A similar conclusion was reached in a recent phylogenetic study ofweevils
and their host plants (Anderson 1993). The influence of
growth fbrm is accentuated by the elevated rate of host shifts
in lineages feeding on trees, the most frequently used plant
growth form among butterflies, while changes among
growths form appear to have occurred independently of the
host plant clade. This, in turn, can perhaps be explained by
Feeny's (1976) distinction between the different kinds of de1-enses
utilized by "apparent" trees and "unapparent" herbs.
As the mature foliage of trees with diff'erent taxonomic origins will have a convergent chemical def-ence,evolving a
capacity to feed on mature leaves of a particular tree will
preadapt the insect to feed on mature leaves from other trees
(Feeny l9'76, l99l). It fbllows that these aspects of plant
chemistry should in tact be better correlated with plant
growth form than with phylogeny.
ln this study we have explicitly focused on the patterns
and determinants of host shifts, through colonization (when
a new plant is added to the host plant range) and specialization (narrowing of the host range, in the case of a host shift
to include only the novel plant), as these seem to be the most
important processes shaping the association between butterflies and their host plants. Even if the patterns that emerge
on this taxonomic level cannot themselves have been caused
by coevolution, the general mechanisms behind host shifts
are of great importance for understanding the dynamics of
the coevolutionary process.
BUTTERFLIES AND PLANTS
491
eds. Coevolution of animals and plants. Univ. of Texas press,
l u t i o n o f t w o b i n a r y c h a r a c t e r s : a r e g a i r . r so r l o s s e s c o n c e n t r a t e d
Austin.
on certain branches of a phylogenetic tree? Evolution 14:539'
1976. Plant apparency and chemical def'ence. Rec. Adv.
55'7.
Phytochem. 10:1-40.
M e n o r s o NW, . P . ,e r o D . R . M e n o r s o N1. 9 9 2 M
. a c C l n d ea:n a l y s i s
I 991 . Chemical constraints on the evolution of swallowtail
ofphylogeny and characterevolution.Vers. 3.05. Sinauer.Sunbutterflies. Pp.315-340 in P. W Price, T. M. Lewinsohn, G. W.
derland. MA.
Fernandes, and W. W. Benson, eds. Plant-animal interactions:
M.roorsoN,W. P., M. J. DoNoc;nr_'r:,
ANDD. R. MeoorsoN. 19g4.
evolutionary ecology in tropical and temperate regions. Wiley,
Outgroup analysisand parsimony.Syst. Zool. 33:83-103.
New York.
Menrm, J. A., eNn D. P. Pasrrr-Ev.1992. Molecular svstematic
FnlslNsrnr:-, J. 1985. Phylogenies and the comparative method.
analysis of butterfly family and some subfamily relarionships
Am. Nat. 125:1-15.
(Lepidoptera:Papilionoidea).Ann. Entomol. Soc. Am. 85:127FIr:olEn. K. 1995a. Lycaenid butterflies and plants: hostplant reI 39.
lationships, tropical versus temperate. Ecotropica I :5 1,58.
MenrrNs,E. P. 1996. Conductingphytogeneticcomparativestudies
1995b. Lycaenid butterflies and plants: is myrmecophily
when the phylogeny is not known. Evolution 50:12-22.
associated with particular hostplant preferences? Ethol. Ecol.
Mtcoor-r-,I. 1987. Field guide to the butterfliesof southernAlrica.
Evol.7:107 132.
C. Struik Publishers,Cape Town, South Africa.
Fnur.tror,r-, P C., ,cNn H. K. Rnr.vu. 1994. Using phylogenies to
MtLrsn, J. S. 1987a. Host-plantrelationshipsin the papilionidae
test hypotheses of adaptation: a critique of some current pro(Lepidoptera):parallel cladogenesisor colonization'lCladistics
posals.Evolution 48: 172-180.
3 : 1 0 5 - l2 0 .
F L :r ' u y r , r , qD
, . J. 1976. Food plant specialization and environmental
1987b. Phylogenericstudies in the papilioninae (Lepipredictability in Lepidoptera. Am. Nat. l1O:285-292.
doptera:Papilionidae).Ph.D. diss. Arn. Mus. Nat. Hist.. Niw
1991. Evolution of host specificity in herbivorous insects:
York.
genetic, ecological, and phylogenetic aspects.Pp.431-454 inp.
Mrrrnn., C., 'rNo D. R. Bnoors. 1983. Phylogeneticaspectsof
W. Price, T. M. Lewinsohn, G. W. Fernandes, and W. W. Benson.
coevolution.Pp. 65-98 in D. J. Futuym:rand M. Slatkin, eds.
eds. Plant-anirnal interactions: evolutionary ecology in tropical
Coevolution.SinauegSunderland,MA.
and temperate regions. Wiley, New York.
Mr-,NRoE,
E. 1961. The classificationof the papilionidae (LepiF r - r r r i y M A , D . J . , M . C . K n l s r : , A N r ) S . J . S c r r u p r . r _ p .1 9 9 3 . G e n e t i c
d o p t e r a )C
. a n . E n t o m o l .S u p p l . l 7 : 1 - 5 1 .
constraints and the phylogeny of insect-plant associations-reNtnt.srN,E. S. 1989. Phylogenyof major lepidopterangroups.pp.
sponses of Ophraella contmuna (Coleoptera, Chrysomelidae) to
281-294 ir B. Fernholm,K. Bremeq and H. Jornvall,eds. Tie
host plants of its congeners. Evolution 47:888-905.
hierarchy of lif'e. Elsevier, Amsterdam, The Netherlands.
Fuluyrvr.r,D. J., J. S. Wer-sH, T. Monlor, D. J. FuNr, eNo M. C.
Nvr-rN,S., .tNu N. WEollr-. 1994. Sexual size dimorohism and
Knusr. 1994. Genetic variation in a phylogenetrc context-recomparativemethods.Pp. 253 280 inP. Eggleton
R. V"nesponses of two specialized leaf beetles (Coleoprera, Chryso"nd
press,LonWright, eds. Phylogeneticsand ecology.Acadernic
nelidae) to host plants of their congeners. J. Evol. Biol. j:l2joon.
146.
Prcnl, M. D. 1992. A methodfor the analysisof comparativedata.
Furuvlre, D. J., M. C. Kllsn, eNo D. J. FuNx 1995. Genetic
J . T h e o r .B i o l . 1 5 6 : 4 3 1 - 4 4 2 .
constratnts on macroevolution: the evolution of host alfiliation
1994. The adaptationistwager.Pp. 29-52 in p. Eggleton
in the leaf beetle genus Ophroella. Evolution 49:79'l-809.
and R. Vane-Wright,eds. Phylogeneticsand ecology.Academic
Getcrn, H. J. 1980. Enzyme electrophoretic studies on the genetic
Press,London.
relationships of Pierid butterflies (Lepidoptera: Pieridae). I. EuPensoNs,M. 1991. Butterfliesof the Bulolo-Wau Vallev. Bishoo
ropean taxa. J. Res. Lepid. 19:181 195.
Museum Press,Honolulu, HI.
GnepnN, A. 1989. The phylogenetic regression. Philos. Trans. R.
P e t r n s o x , A . M . , R . D . G n e v , , r N oG . P .W e l l r s . 1 9 9 3 .p a r a s i t e s .
Soc. Lond. B Biol. Sci.326:119-156.
petrels and penguins:does louse presencereflect seabirdphyHaNcocx, D. L. 1983. Classification of the Papilionidae (LepiIogeny?Int. J. Parasitol.23:515-526.
doptera): a phylogenetic approach. Smithersia 2:l-48.
PIennn,J. 1987. Syst6matiquecladistiquechez les Acraea (LepiHARVEy, D. J. 1991. Higher classification of the Nymphalidae,
doptera,Nymphalidae).Ann. Soc. Enrornol.Fr. 23 ll-2j.
Appendix B. Pp. 255-273 in H. F Nijhout. The developmenr
Pr<rcE,
P. W., C. E. Bor.'roN,P. Gnoss, B. A. McprrEARSoN,
and evolution of butterfly wing patterns. Smithsonian Institution
J. N.
THOvpsoN,nNo A. E. Wers. 1980. Interactionsamong three
Press, Washington, DC.
trophic levels:influenceof plantson interactionsbetweeninsect
Henvnv, P. H., exo M. D. Per;nr_. 1991. The comparative method
herbivoresand natural enemies.Annu. Rev. Ecoi. Syst. 1l:41
in evolutionary biology. Oxford Univ. Press, Oxford, U.K.
65.
Htct;rxs, L. G., .qNoB. H,rncpervns. 1983. The butterflies of BritPunvrs,A., aNn A. Rrrasrrrr. 1995. Comparativeanalysisby inain and Europe. Williarn Collins and Sons, London.
Hoc;r-uNn, J., eso B. SrllEN-Tur-LBERC. 1994. Does lekking prodependentcontrasts(CAIC): an Apple Macintosh application
mote the evolution of rlale-biased size dimorphism in birds? On
for analysingcomparativedata.Comput. Appl. Biosci. ll 241t h e u s e o f c o m p a r a t i v e a p p r o a c h e s .A r n . N a i . 1 4 4 : 8 8 1 - 8 8 9 .
251.
Jenvv, T. 1976. Insect-host plant relationship-coevolution or seS e r r ; u s eT
, . , A . N e x n v r c H r ,H . S s r v e , e N o O . y n r . r . 1 9 8 2 . p h y quential evolution? Symp. Biol. Hung. l6:109-133.
logeny and geographicaldistribution of the swallow-tail sub1984. Evolution of insect/host plant relationships. Am.
gents Grophiun (Lepidoptera:Papilionidae).Entomol. Gen. 8:
Nat.124:609-630.
59 69.
JouNsloN, G., eNo B. JosssroN. 1980. This is Hong Kong: butSc:oet.s,
M. J. 1986. The structureand affinitiesof the Hedyloidea:
terflies. Hong Kong Government Publications, Hong Kong.
a new conceptof the butterflies.Bull. Br. Mus. Nat. Hist., EnKnrsrnNsurl, N. P 1976. Remarks on the family-level phytogeny
tomol. Ser. 53:251 286.
of butterflies (Insecta, Lepidoptera, Rhopalocera). Zeitschrili fiir
1996. In searchof butterflyorigins: morphologyand molZoologische Systernatik und Evolutionsforschung 14 25 33.
ecules.TrendsEcol. Evol. ll 274-275.
LensuN, T. B. 1914. Butterflies of Lebanon. National Council for
Scorr, J. A. 1986.The butterfliesof North America.StanfordUniv.
Scientilic Research. Beirut.
Press,Stanfbrd,CA.
1991. The butterflies of Kenya and their natural history.
Scolr, J. A., ,qNoD. M. Wnrr;Hr. 1990. Butterfly phylogenyand
Oxford Univ. Press, Oxford, U.K.
fossils. Pp. 153-208 irz O. Kudrna, ed. Butterfliesof Europe.
Losos, J. B. 1994. An approach to rhe analysis of compararive
AULA-Verlag, Wiesbaden,Germany.
data when a phylogeny is unavailable or incomplete. Syst. Biol.
S c n r e e pJ, . M . , e N o F , S l e N s r v . 1 9 8 1 .T h e n u t r i t i o n ael c o l o g vo f
43:l l7-123.
immature insects.Annu. Rev. Entomol. 26: 183-21l.
MeoorsoN, W. P 1990. A method for testing the correlated evoSlLr-ls-TULLeERG,
B. 1988. Evolution of gregariousness
ln apo-
498
N. JANZ AND S. NYLIN
sematic butterfly larvae: a phylogenetic analysis. Evolution 42:
293-305.
1993. The etfect of biased inclusion of taxa on the correlation between discrete characters in phylogenetic trees. Evol u t i o n4 7 : 1 1 8 2 - l l 9 l .
Sr,rent, P. 1975. The illustrated encyclopedia of the butterfly world.
Salamander Books. London.
Svrr.s.r, J. 1978. Plant chemistry and the evolution of host specifrcity: new evidence frotn Heliconius and Passiflora. Science
2OI:145-747.
S p r : n r - r N c .F , A . H . 1 9 9 2 . M i t o c h o n d r i a l D N A v a r i a t i o n a n d H a l dane's ruie in the Papilio glaucus and P. troilus species groups.
Heredity 1 l:227 -233 .
1993. Mitochondrial DNA phylogeny of the Papilio maclicon species group (Lepidoptera: Papilionidae). Mern. Entonol. Soc. Can. 165:233-242.
Sworr-ono, D. L. 1991. PAUP: phylogenetic analysis using parsimony. Vers. 3.1.1. Illinois Natural History Survey, Champaign,
IL.
W e r s o x , L . , e N o M . J . D , q r - r - e wt r z . l 9 9 l . T h e f a m i l i e s o f f l o w e r i n g
plants: descriptions, illustrations, identification, and information
retrieval. Vers. 16. March 1997. URL http://www.keh.ukans.edu/
plant
gateway,
hrrp:l/
URL
flowering
delta. In:
The
w w w . i s c .t a m u . e d u / F L O R A / c r o n a n g . h t m .
WnllEn, S. J., euo D. P. PesrtlEv. 1995. In search of butterfly
origins. Mol. Phylogenet. Evol. 4:235-246.
W n r - r - r : n , S . J . , D . P P , q s H I - r , v ,A N D J . A . M , c R ' t t N . 1 9 9 6 . R e a s sessment of butterfly family relationships using independent
genes and rnorphology. Ann. Entomol. Soc. Am. 89:184-192.
A. R., lNo M. R. BEIrpNerurl. 1993. Plant chemistry,
Zrscpnl,
insect adaptations to plant chemistry, and host plant utilization
patterns. Ecology 74:47-54.
Corresponding Editor: C. Mitter
Appr,Notx 1
List of important host families fbr butterflies included in the clades
recognized by Chase et al. (1993). Question marks denote families
that were not included in the analysis by Chase et al., and for which
positions have been inferred lrom the classification of Cronquist
(1e81).
Rosid l: (A): Celastraceae, Erythroxylaceae, Euphorbiaceae, Lina c e a e , M a l p h i g i a c e a e , O c h n a c e a e , P a s s i f l o r a c e a e ,S a l i c a c e a e ? ,V i o laceae; (B): Cannabidaceae, Fabaceae, Krameriaceae, Moraceae,
Polygalaceae, Rhamnaceae, Rosaceae, Ulmaceae, Urticaceae, Zygophyllaceae; (C): Begoniaceae, Betulaceae, Casuarinaceae, Cuc u r b i t a c e a e , F a g a c e a e ,J u g l a n d a c e a e ,M y r i c a c e a e ; ( D ) : O x a l i d a c e a e
(Oxalis): (E): Combretaceae, Melastomaceae, Myrtaceae, Punicaceae.
R o s i d 2 : A c e r a c e a e , A n a c a r d i a c e a e , B a t a c e a e , B o r n b a c a c e a e ,B r a s s i c a c e a e ,B u r s e r a c e a e , C a p p a r a c e a e ,C a r i c a c e a e , G e r a n i a c e a e , H i p p o c a s t a n a c e a e , M a l v a c e a e , O x a l i d a c e a e ( H y - p s e r o c h a r i s ) ,R e s e d a ceae, Rutaceae, Sapindaceae, Simaroubaceae, Sterculiaceae, Tiliaceae, Tropaoleaceae.
Rosid 3: Crassulaceae, Grossulariuceae (Ribes . . .), Hamamelidaceae, Saxilrag rceae ( Sar ifra ga).
Asterid 1: Acanthaceae, Apocynaceae, Asclepiadaceae, Bignoniaceae, Boraginaeae, Convolvulaceae, Cornaceae (Aucuba'), Gentianaceae, Gesneriaceae, Hydrophyllaceae, Lamiaceae, Loganiac e a e , O l e a c e a e , P l a n t a g i n a c e a e ' J ,R u b i a c e a e , S c r o p h u l a r i a c e a e , S o lanaceae. Verbenaceae.
Asterid 2: Apiaceae, Aquifoliaceae, Araliaceae, Asteraceae, Campanulaceae, Caprifoliaceae, Cornaceae (Corokia), Cornaceae (Grise linio), Cornaceae (.HeIw i ng ia ), Dipsacaceae, Menyanthaceae, Pitt o s D o r a c e a e .V a l e r i a n a c e a e .
Asterid 3: Diapensiaceae, Ebenaceae, Epacridaceae, Ericaceae,
Myrsinaceae, Primulaceae, Sapotaceae,Symplocaceae, Theaceae.
Asterid 4: Alangiaceae, Araliaceae, Cornaceae (Conlus), Nyssaceae, Hydrangeaceae.
Asterid 5: Dilleniaceae. Vitidaceae.
Ranunculids: Berberidaceae, Fumariaceae, Menispermaceire, Papaveraceae, Ranunculaceae.
Paleoherbs 1: Aristolochiaceae, Piperacetre.
Monocots: Arecaceae, Bromeliaceae, Commelinaceae, Cyperaceae,
Dioscoreaceae, Heliconiaceae, Liiiaceae, Musaceae, Orchidaceae,
Poaceae, Smilacaceae, Zingiberaceae.
Laurales: Hernandiaceae, Lauraceae, Monimiaceae.
Magnoliales:
Paleoherbs
Annonaceae, Cannelaceae, Magnoliaceae, Wintera-
2: Chloranthaceae, Illiciaceae, Nympheaceae.
1: Platanaceae, Proteaceae, Sabiaceae, Aizoaceae,
Hamamelid
Amaranthaceae. Cactaceae?
Caryophyllids : Caryophyllaceae, Chenopodiaceae,Nyctaginaceae,
Olacaceae, Plumbaginaceae, Polygonacetre, Portulaceae, Santalaceae. Viscaceae/Loranthaceae.
Pinaceae.
(other) conifers: Cupressaceae, Podocarpaceae, Taxaceae, Taxodiaceae.
Cycads: Cycadaceae, Zamtaceae.
AppnNorx 2
Description of the butterfly phylogeny used in this study. Clades
are identified by parantheses, numbers refer to taxon # in Appendix
3 The phylogeny can be obtained from the authors in electronic
form upon request.
( ( ( 4 4 0 , ( 4 3 8 , 4 3 e ) ) , ( 4 4 r . ( 4 4 2 , 4 14 ,3()()s),,((3( , ( 2 , 4 ) ) )1,8( (, 1
(e ) , ( 2 0 .
, (12,370, )(, 2 8 ,
( 2t , ( 2 2 , ( (s3, 3 6 ) , ( ( 2(62, 3 , ( 2 s , 2 4 ) ) ) , ( ( 3 2 , ( 3 3 , 3 1 () () 3
2e)))))))))),(17,((6,(7,(((e,8),(11,10)),((12,13),(16,14,15)))
, 8)),
( . ( 4 e , 4 6 ) , ( 4 78),) ) ,Q e , ( ( 4 s . 4 4 ) , ( 4 0 , ( 4 3 , ( 4 t , 4 2 )()()6)0) (), 3
, 73
( 5 e (, ( s 6(, 5 8 , s 7 ) ) , ( ( s 2I ), 5
, ( 5 5 , ( s 4 , s 3 ) ) ) ) ) ) ) ) ) ) )1) , ( ( (6e 0 ,I e) , ( e 3 ,
8 ,8 7 ,8 e ) , ( 8 0 ,2( 1, ' s7 , ( 8 4 , 8 s ) ) ,
e 2 ) , 9 4 , 9 5 , ( 9 6 , 9 7 ) 1, e080,,(9 9 ) ) , ( ( 8 6
, 74 ) ) ,(6 7 ,6 9 ,68) ,6 s ,66 ,( 6 3 ,6 4 ) ,
9 , ' 7 ' 76, 7
8 2 ,83 ,( (8 1 ,7 1 ,7O )(, 18 , 73 , ' 7
6 2 ) ) ) )( ,(1 0 91, 0 2(,1 0 11, 0 3 )(,1 0 81, 0 71, 0 61, 0 51, 0 4 ) )( ,( (l 8 0 ,( ( 18 1 ,
| 3,(229,(227,228)),
19,220,225,224,226)),2
182),((Q2r,222,223),(2
4,23s,
r,23O,232,233,23
202,((195,196),( 198,197)),183,(244,247,23
246,245),((2I 6,2 I s),2 | 8,2| 4,
236,23'7,
238,239,240,24r, 242.243,
2r1), (212,(209,2rO),2 l 1), (206,205,(207,208),(204,203)),(201,
( 1 e 9 , 2 0 0 )( )1,9 0(,1 9 1 1, 9 2 ) )(,1 8 e(,1 8 6 1
, 8 4 1, 8 5 1
, 8 71
, 8 8 ) )1, 9 3 ,
l s ) , ( 11 2 , ( 1 4 r, 1 3 ) ) ) , ( ( 1 5 8 , ( 1 s 6 l,s145,13 5, 51, s 7 ) ) ,
1 9 4 ) ) ) , ( (1(61, 1
( 1 76 , 1 7
1 ,75 ,1 6 1 . 137, r 11 ,17 4 ,16 8 ,16 5 ,16 9 ,16 6 ,12 0 ,
0,17
2 , 16 3 , 1 6 4
1 6 2 . 1 6 7 , r 2 r6
, (01,1 7 81,7 7) ) ) , (I 5 9 , (I 2 8 ,12 7 ) ) , (1(2 9 ,I 3 1 ,I 3 0 ) , 1( 3 2 ,
4 0 ) , 1 14) , ( ( 1
13 3 ,l 3 5 , 1
3 4 ) , ( (1(3 6 , 13 8 , 1 3 73, 19 , 1
( 4 2 , 1 4 3 I)5, (l , 1 5 2 ,
I 45)))))),((tzs, ( | 24,(r22,r23))),
l 50)),( 144,(( 149,148),(t 47,1,16,
12 6 ,( 118 ,( l l 7 , l l e ) ) ) ) I, 7 e ) ,I I 1 ,1I 0 ) ,( 4 3 7(,(4 3 64, 3 s (, ( ( 4 3 3 , 4 3 4 ) ,
( 4 3 r , 4 3 2 ) () (, 4 2 6 , 4 2 1()4, 2 8(, 4 3 04, 2 e ) ) ) ) ) , 4 2(s(,( 4I 8 ,4 I 9 ) ,( 41s ,
((393,(3e9,3es,394,396,397,
(4r7,4 16))),(420,(422,42
r,423,424))),
8 8 , 38 9 ) , ( I44 ,
8 3 , 38 4 , 38 s ) ) , ( 3I9, 3 9 0 , 3 e 2 ) ,8(73, 38 6 ) , 3
3 9 8 ) ) , (832 , ( 3
( 404,403),(40r, 4O2,400),405,4 10,409,( 408,4O1,406),(41|, 4 12),
((257,2s6,2s5),
t,265'),
4t3),(((267,Q66,259,260.264,263.262,26
(249,((268,((269,27O),((2'7t, (272,
2s 8,254,253,252,2sr,2sO),248),
2 7 3 ) ) , ( ( 247, ( 2 1s , 2 ' 6
7 ) ) , ( 2 7.'(72 78 , 2 79 ) ) ) ) ) )(,2 8 0(,2 8I , ( ( 2 8 2 , ( 2 8 4 ,
I ) ) ) ) ) ) ) ) ) )()3, 3 7 ,
2 8 3 ) )(,2 e 2(, ( 2 8 s(, 2 8 62, 8 7 ) )(,2 8 8 (, 2 8 e(, 2 9 0 , 2 9
(340,(344,343'),(34r,342)),339,345
(338,(35l .(350,349),348),
,346,
, 5 s ,3 s 6 ,3 s 3 ,
, 7 9 ,3 7 8 ) )(,( ( 3 5 83, s 9 ) ,( 3 5 7 (, 3 5 2 3
3 4 7 ) (, 3 8 1 (, 3 8 0 3
3s 4 )) ,36 4 (, (36 0 3
, 6 1) ,( 36 2 ,3 6 3))) ,3 6 9(,368 ,36 7 , 3 6 6 ,63s ) , ( 321, 3 7O ,
3 11 , 3 '67, 3 7s , 3 74 , 3 73 , 3 1t ) ) ) ,((( ( 2 95 ,( 2 96 , 2 98 ) ) , 249, 2 9 7
) ,2 9 33, 0 0 ,
( 3, 0 3(,3 0 4 , 3 0 s )()()3,I 6 , 3I 3 ,3 I 7 , 31 s ,3 r 4 , 3 1 2 , ( 3 0 9 ,
2 9 9 3, 0 1 , 3 0 2
3 0 8 )3, 0 73, 0 6 (, 311 ,3 10 ) ) ,( ( ( 3 2 33, 2 2 , 3 2 r3. 2 0 )3, 2 7 (, 3 2 83, 2 6 , 3 2 s ,
3 2 4 ) , l39 ) , 31 8 , ( 3 3 2 , 13,33 3 3 , 3 3 53,43, 3- r 0 ,3- r6 , 3 2 9 ) ) ) ) ) ) ) ) ) ) ) ;
499
BUTTERFLIES AND PLANTS
o\
O
E
-c
6
€
-
a
a)
t
. , - o
x
.!-.
€ r t O O - . 1 - , t r € r - C a - a l - j 6 € r € l C - d 6 f n € r 4 l C - . r - J r @ r - 5 O - N - + h € r ! 6 O - d - i r @ r € 5 O
6 3 a 3 c - - c - o o : : : - : = = = = - - - * : 1 1 : : !
d
q
j
*
a
s
: i l
:
-:
a -
i-t
'' 'c
, o
3
. ja = =27'
E
_ U
r
=---=
?
P
=
=
1
i t- ? -
-
- ;: -aE3?sa
:^ .-:7a.'
E1.
= \
!
; : i . - a i i r -
: 4
= i == = i G t i i = : 1=;
=-7i:a::iririZ
t
+
:
i
i
,
Z
i
4
;=
=-i
iz i 27i 12i
; ; l ; z+jZ>Ei
i'ti / r-: i7=u
li 1= ;i i t Ei ; z2?i2 i2 . ; : t' E
;dE:i5: : . . . . , , ,
' ,
. ., ..:
O . :
3j; f(;i*:-:J=-;;
o
d
; =
=
:
L ;
F
.
:
-
c
F U
A
:'
Q
. = a
g
.Oe
O *
=
^;
X
x €
! o
: s
;<
cr !?
;
r
o
c
-
d
-i
a
n
€ r
€ o
a
-
d
-
+
n
€ r
€ 6
o
-
d
-
t
n
€ r
€ o
c - d - $ n €
t t * + i + + i + J
r
r
3
s00
N. JANZ AND S. NYLIN
.b
1a
; a
-
2 i ^ i -
t - a = z E
I
L
^ L
! - + F i E +! E t ; : !
-e .. t -!=
iai .41-t?
E i 2 ; j 2 . ) - t E E A { .a
=
L
! -
=
-
:
Y
.
;
i
;
:
:
=
.
!1211i i
1
-
.
=
_
.
n
,
-
:
2 i : a d
.
!
r
E
E
:'
E
r
O
x
:
z
a
._
.g
3
l
u
ill
c
<
':
=
d
-
:
:
;
;
'a o
:
a
i a
:
J U
n
d
a
a
i d
3
E
q
- c r - i n G r - 3 C + a l - o - - + - i t
! 3 0 - a r - +
J J r n n
n
€ r
€ t a - . t - , 1
6 €r { 6 c
6 n € € @ c o @ @ € € o r r r
.
4
:
_a
? 7 - . ; t ' -E== = E : =
r - -8.-= ^:;.:,.
i : , , ; - , = = - e" =
F=;;:
:tj-i
;:i ; ae :: : i : r,
- :ii
> ,
- r l
a f n € r - ]
r r r r r r r
c - a r - J 6 € r F o - - . t - t 6
c o 3 3 5 0
501
BLITTERFLIES AND PI-ANTS
:.
=
-
d
l g e
!
! i = =
1
E =
i i '
z
:
= - . - ; : i . : : , : = - - t -
.- !r '=ij :?j e= =i = -= t= ;=; ! :?- = : 4 , d = . = = = - i= Q ; , - - : i z i; =; t7 =
-- iz : L= <a.,=a j -=E' =
v v = 7 - = v =- Fi = ': =irE-: c6 -i5a3 " Z: =- i : E a rr r 1 n n - z. i i
'-s
:
:
l 1
- \ , r
;
d
.
il
:
E :
u
d
!
"o
=
!:
U
L a€ r
r a
-
- . r
-
t
n
€ r - 5 c
- c r - a r @ r €o
t i t * 1 1 1 + + + L . L . L . ' .
- . r
-
- * . r - 4 6 € r { o a @ o c @ G c c € € € r r
- J ' a € r -
t n € r r O - - . 1
o
-
z
r
r
a.
Z
=
E
:
F
: :
= . 1
a
-
a
t
D
l
a
- . r . r r i t r r
_ . r
E
_ . 1
o
- . r - - - + +
- ' = = - - Z = - . = = = = . = = = 2
' t :
lz tL, : . eY f -r ' <! =
-.iii1.:,i
4 < < < < < <
,
^
=
= =
t
=
4
!
--=
L
n
a
<
:
4
Z Y
d
-
t L : . L >
e Y
;
'
t
E
L
3 C - d - f
n
€ r
- O
O -
d
- / 1 6 e r € O C - d
d d d d d . t d d d N d a r d d d
-
f
n
C
r
= 4 ,
:
>
=
E
Z
E
z
-
- = E
= - i = - r ; = ,
. z : . = = ' 1 !=i : , i E - a ? ?
++t ; ; : i i = t = = = - a = a = :
< - < - < - < , < 4 , 4<. < 4 4 < < . > E a >
E
- . r - , f n € r - o a - . r - 4 6 € r
€ € c c € € c c € r
!
,
a y F , . ; i E a =
= . = ! , = Z fZ a 1 : i : = - = .
- 2 1 - Z + Z a = ' ! . e = z
-
Z-
?
>
d A ,
-
502
N. JANZ AND S. NYLIN
A P P L , N D I X4
Descriptions of step matrices used in this paper.
=
=
t r
Step matrix "a" was used to describe transformation costs between the l5 possible combinations of "Rosid 1B," "other rosids"
(all rosids except Rosid lB), "asterids," and "other seed plants."
State numbers translate as follows: 0, Rosid lB; I, other rosids; 2,
asterids; 3, other seed plants; 4, Rosid 1B * other rosids; 5, Rosid
1B * asterids; 6, Rosid lB * other seed plants; 7, other rosids f
asterids; 8, other rosids * other seed plants; 9, asterids + other
seed plants; A, Rosid 1B f other rosids f asterids; B, Rosid 1B
+ other rosids * other seed plants; C, Rosid 18 + asterids * other
seed plants; D, other rosids i asterids + other seed plants; E, all
groups.
Step matrix "b" was used to describe transfbrmation costs between the seven possible combinations of the plant groups "rosids,"
"asterids," and "other seedplants." The same step matrix was also
used for the seven possible combinations of the growth forms
"herbs," "vines," and "trees and shrubs." State numbers translate
as follows: 0, rosids (herbs); I, asterids (vines);2, other seed plants
(trees and shrubs); 3, rosids * asterids (herbs f vines); 4, rosids
* other seed plants (herbs + trees and shrubs); 5, asterids + other
seed plants ( vines * trees and shrubs); 6, all groups.
; c c - .
1 = = =
:
=
I
:
to: u
From: 0
0
1
1
0
3
::
0
5
:E
U
r €o -
- d - a 6 o r
-
3 a - . r -
i r , €r €t
T
a - d -
+=+113!tt??!!TiiTT??+TT++++
r;
/.
2
6
0
2
2
0
2
1
I
3
3
1
3
'|
0
3
3
'I
1
'I
3
2
3
2
0
I
2
I
2
3
I
2
0
I
0
2
:1: -!-.2:tE
-i:.+;+2i=
ji+E-.r:;i
t
a
2
E
c -
- . r - a r €r
€l
a
- . 1 - i
- + t f - t i + t
r €
r
€ o - - d - J ' . €
- o - * - - - - f J a t t a + + +
5
6
J
3
I
1
'I
1
1
1
2
2
1
1
0
1
1
1
1
0
1
4
4
0
I
3
T o : 0 1 2 3 4 5 6
1
1
I
I
2
0
3
0
1
1
0
3
3
2
3
2
2
3
3
3
2
2
A
2
2
2
4
J
2
2
3
2
0
2
3
0
3
3
0
3
2
0
2
2
3
J
J
1
3
2
'I
I
2
3
A
I
J
From: 0
2
1
1
2
I
J
'|
3
3
J
3
,|
0
I
E
3
3
J
3
1
2
2
o
2
1
1
2
J
2
t
2
7
B
2
1
8
A
Z
r
r
1
2
9
c
1
2
3
3
3
2
1
1
1
,l
0
0