Molecular Ecology (2008)
doi: 10.1111/j.1365-294X.2008.03898.x
OPINION
Blackwell Publishing Ltd
Homoploid hybrid speciation in animals
J E S Ú S M AV Á R E Z * and M AU R I C I O L I N A R E S †
*Centro de Ecología, Instituto Venezolano de Investigaciones Científicas, Apartado 20632, Caracas 1020-A, Venezuela, †Instituto de
Genética, Universidad de los Andes, Carrera 1E No 18A — 10, PO Box 4976, Santafé de Bogotá D.C., Colombia
Abstract
Among animals, evidence for homoploid hybrid speciation (HHS, i.e. the creation of a
hybrid lineage without a change in chromosome number) was limited until recently to the
virgin chub, Gila seminuda, and some controversial data in support of hybrid status for the
red wolf, Canis rufus. This scarcity of evidence, together with pessimistic attitudes among
zoologists about the evolutionary importance of hybridisation, prompted the view that HHS
is extremely rare among animals, especially as compared with plants. However, in recent
years, the literature on animal HHS has expanded to include several new putative examples
in butterflies, ants, flies and fishes. We argue that this evidence suggests that HHS is far
more common than previously thought and use it to provide insights into some of the
genetic and ecological aspects associated with this type of speciation among animals.
Keywords: animals, homoploid, hybridisation, introgression, molecular markers, speciation
Received 17 April 2008; revision received 9 July 2008; accepted 14 July 2008
In 1853, W. C. Hewitson was confronted with a challenge
while describing a new species of Heliconius butterfly, whose
fore- and hindwings strikingly resembled those of Heliconius
erato and Heliconius charitonia, respectively (Fig. 1). Heliconius
species were named with reference to Greek mythological
personages, a tradition going back to Linnaeus. Hewitson
decided to name it Heliconius hermathena, and he could not
have made a more elegant choice. In antiquity, a Hermathena
was a composite or hybrid bust with two heads: the
Greek gods Hermes and Athena — for an example see
www.mnarqueologia-ipmuseus.pt/?a=3&x=2&i=59. One
hundred and fifty years later, it has been suggested that H.
hermathena is a case of homoploid hybrid speciation (hereafter
HHS) (Beltrán et al. 2007), which if confirmed will make
Hewitson’s providential choice of name one of the oldest,
if not the oldest recognition of hybridism in an animal
species. Nonetheless, HHS among animals has remained
highly controversial, both for theoretical and empirical
reasons (Coyne & Orr 2004). In recent years, the number of
suggested/confirmed cases has increased dramatically
(Table S1, Supplementary material), opening the door to
an analysis of the variety of ecological and evolutionary
contexts associated with this speciation process.
Correspondence: Jesús Mavárez, Fax: +58 212 504 1088; E-mail:
[email protected]
© 2008 The Authors
Journal compilation © 2008 Blackwell Publishing Ltd
Definitions
During HHS, a stable, fertile and reproductively isolated
hybrid lineage arises without change in chromosome number
(Rieseberg 1997; Coyne & Orr 2004). This definition implicitly
assumes that hybridisation is fundamental to speciation
and reproductive isolation must rise during or after hybridisation (e.g. Fig. 2a), but not before it (i.e. adaptive introgression, Fig. 2c). However, in theory, reproductive isolation
might appear much later when the initial evolution of the
HHS lineage occurs in geographical isolation (Buerkle et al.
2000). The definition also needs to be expanded to incorporate ‘without changes in chromosome number and mating
system’, in order to avoid the inclusion of asexual diploid
putative hybrids such as the soft coral Alcyonium hibernicum
(McFadden & Hutchinson 2004), whose shift in mating system
creates immediate isolation and hence is more closer to
polyploid hybrid speciation. The remaining aspect of the
definition, hybridisation, is discussed in detail below.
Molecular markers
Recent evidence of putative HHS in animals (Table S1) is
based principally on analyses of variation at molecular
markers. Three main lines of evidence rooted in conflicting
information about the ancestry of the hybrid species are
used: (i) genealogical discordance, in which different markers
2 J . M AVÁ R E Z and M . L I N A R E S
Fig. 1 From upper left: Pogonomyrmex rugosus var. fuscatus, Lycaeides sp. and Papilio appalachiensis. Second row: Heliconius heurippa,
Daphnia galeata mendotae and Rhagoletis sp.
Third row: invasive Cottus perifretum and Gila
seminuda. Bottom row: Heliconius erato, H. hermathena and H. charitonia. Photo credits: T.
Schwander (P. r. f.), C. Nice (L. sp.), D. Wright
(P. a.), M. Linares (H. h.), D. J. Taylor (D. g. m.),
D. Schwarz (R. sp.), J. Freyhof (C. p.), G.
Clemmer (G. s.), O. Delgado (H. e., H. h. and
H. c.).
Fig. 2 Schematic alternative speciation scenarios creating an admixed species H from species 1 and 2. (a) H is the product of a homoploid
hybrid speciation (HHS) between species 1 and 2. (b) H is the sister species of 1, sharing ancestral polymorphisms with 1 and 2. (c) H is the sister
species of 1, receiving subsequent adaptive introgression from 2. Scenario B implies no hybridisation, and in theory can be distinguished
from the other two, although with varying degrees of difficulty depending on the genetic markers and analyses used. Scenario C is intermediate between A and B and can be particularly difficult to distinguish from the first. However, if introgression from species 2 in scenario
C enhances isolation between species 1 and H, the latter could still be considered as a hybrid species. In fact, the reason why scenario
C does not constitute HHS is that hybridisation did not play a role during the initial development of reproductive isolation in species H.
suggest alternative ancestral relationships (e.g. Papilio); (ii)
intermediate frequencies, in which markers possess a mixture
of the alleles which may have either fixed (e.g. Gila, Heliconius)
or frequency differences (e.g. Cottus) in the parental species;
and (iii) recombinant sequences, which appear as the product
of recombination events between two parental haplotypes
[e.g. ITS–rDNA (internal transcribed spacer–ribosomal DNA)
and HSP90 in Daphnia]. However, most proposed cases of
HHS are based on a rather small number of markers, which
limit the rejection of alternative explanations (e.g. retention
of ancestral polymorphisms, Fig. 2) and imply that additional
morphological and ecological evidence are needed to corroborate instances of HHS. On the other hand, the combination
of molecular data and coalescent simulations in order
© 2008 The Authors
Journal compilation © 2008 Blackwell Publishing Ltd
H O M O P L O I D H Y B R I D S P E C I AT I O N I N A N I M A L S 3
to distinguish HHS from other scenarios is an excellent
approach that has been rarely used (see assumptions and
examples of the coalescent approach as applied to HHS in
Bertorelle & Excoffier 1998; Mallet 2005 and Salazar et al.
2008). Interestingly, several proposed instances of HHS
cannot be detected by mitochondrial DNA (mtDNA) alone
(Table S1), indicating that single-locus analyses with these
markers have very limited utility and may be useful only in
cases where females from both parental species have contributed to hybridisation and lineages have not yet sorted.
Morphology
With few exceptions, almost all taxa resulting from HHS
appear morphologically intermediate, although they may
resemble one of the parental species more closely than the
other, presumably reflecting a larger genetic contribution of
that species during the hybridisation process (e.g. Heliconius).
However, conclusions about parental contributions must
be made with care given that statistical analyses of multiple
morphological traits (e.g. geometric morphometrics) have
been performed in only a handful of taxa (i.e. Cottus, Gila,
Heliconius, Papilio). Instead, morphological evidence is often
based on either studies of single traits or on the authors’
knowledgeable intuition about the group. Nonetheless, the
suggestion that hybridisation usually leaves a morphological
signature has profound consequences for our ability to detect
hybrid taxa and, when coupled to molecular data, can provide invaluable evidence to distinguish HHS from alternative
scenarios (such as ancestral polymorphism, which does not
require morphological intermediacy). On the other hand,
laboratory crosses between parental species have been performed in order to reproduce a putative HHS taxon’s phenotype through controlled hybridisation (i.e. Gila, Heliconius,
Papilio, Xiphophorus) and to analyse the genetic basis of hybrid
traits (Heliconius). Such studies can provide an excellent
source of supplementary evidence in support of HHS and
should be performed when possible.
Ecological divergence
Gross & Rieseberg (2005) have noted that reproductive isolation between HHS and their parental plant species is almost
consistently associated with some form of ecological divergence, an observation that fits well with theoretical predictions (McCarthy et al. 1995; Buerkle et al. 2000). Although
such divergence is not as well documented in the animal
literature, some inferences can be made from the ecology
and the geographical distribution of HHS animals and their
parental species. For instance, a few HHS taxa are allopatric
(e.g. Gila, Lycaeides) or at least parapatric (Cottus) vis-à-vis
their parental species. In some cases, at least one ecological
adaptation contributing to isolation from the parental form
has been identified, such as a new host plant (Rhagoletis,
© 2008 The Authors
Journal compilation © 2008 Blackwell Publishing Ltd
Lycaeides), or a change in altitudinal range (Lycaeides) or temperature tolerance (Cottus). However, about half of the hybrid
taxa in Table S1 are sympatric with at least one parental
species, making ‘hybrid taxon × parental sp.’ hybridisation
not only theoretically possible but also observable in nature
(e.g. Daphnia, Heliconius, Pogonomyrmex). Thus, although ecological divergence from parental forms can be an important
feature in some hybrid taxa, its incidence is not general and
other behavioural or genetic factors must be invoked to
explain reproductive isolation.
Geographical analyses also reveal that most putative HHS
animal species have restricted distributions compared to
their parental species. This is consistent with hybridisation,
since this process usually occurs at a local geographical scale
(Barton & Hewitt 1985; Arnold 1997) and alternative hypotheses to HHS such as incomplete lineage sorting do not predict that species harbouring ancestral polymorphisms should
necessarily have restricted geographical distributions. More
importantly, the restricted distribution of most putative HHS
taxa allows for a test based on geographical discordance:
when markers are shared between the putative HHS and
sympatric populations of one or both parental species, but
undetected in allopatric populations, interspecific allele
sharing due to retention of ancestral polymorphisms is
unlikely. Incomplete lineage sorting does not predict such
geographical pattern, whereas it is expected from hybridisation and HHS (Harrison 1990; Barton 2001). For example,
in Heliconius, of all the 40 or so colour races found in the
parental species, the combination of colour pattern genes
observed in the putative HHS Heliconius heurippa only occurs
in the two races with distributions geographically adjacent
to H. heurippa (Mavarez et al. 2006). Similarly, in Daphnia,
the putative HHS shares some ITS haplotypes with one
parental species in the sympatric but not the allopatric
range (Taylor et al. 2005).
Reproductive isolation
Analyses of reproductive isolation are crucial to understanding HHS: the main theoretical objection to this type of
speciation focuses on the difficulty of evolving barriers to
gene flow between hybrids and their parapatric/sympatric
parental species (Coyne & Orr 2004). Although only a few
studies have performed such analyses in animals, available
data suggest a key role for different forms of prezygotic isolation. For instance, assortative mating appears to largely
drive isolation in Heliconius, plays a significant role in Pogonomyrmex and Lycaeides, and probably contributes in Xiphophorus and the Cichlids. Furthermore, allochrony appears
to be important in Papilio and Pogonomyrmex, since these
species occur in sympatry but never or rarely mate simultaneously. Finally, host-based mate choice is certainly the
isolating barrier in Rhagoletis and may contribute in Lycaeides.
However, postzygotic factors have also been shown to gene-
4 J . M AVÁ R E Z and M . L I N A R E S
rate isolation through sterility in Heliconius, inviability in
Pogonomyrmex, and might also play a role in Daphnia and
Cottus. Whether this difference in the importance of pre- vs.
postzygotic isolation reflects a biological trend or a bias in
the studies remains to be seen, but prezygotic barriers appear
to be stronger in the two cases where both isolating mechanisms have been analysed using a comparative approach
(Heliconius and Pogonomyrmex).
Are HHS taxa rare or just difficult to detect?
HHS is undoubtedly an infrequent process compared to
bifurcating speciation (Coyne & Orr 2004). Yet several lines
of evidence suggest that its importance has been underestimated, at least in animals. First, similar numbers of
instances of HHS have been confirmed in plants and animals
(Gross & Rieseberg 2005), and the number of cases in animals
could easily double if additional evidence is found in support
of the less well-established examples in Table S1. Second,
HHS seems quite common in groups known for their high
propensity for hybridisation, such as freshwater fishes and
butterflies (Scribner et al. 2000; Mallet 2005). Furthermore,
several cases have been proposed in Heliconius, Gila and
Papilio, which if confirmed, will provide fundamental
support to the idea that hybridisation can play a significant
role during adaptive radiation (Dowling & DeMarais 1993;
Seehausen 2004; Mallet 2007). Third, almost all of the proposed cases belong to well-known groups of organisms
studied by an active research community combining molecular, morphological and ecological data. This leaves open
the possibility that many more examples could be waiting
to be discovered in lesser-known groups. Therefore, although
HHS will probably remain relatively uncommon compared
to bifurcating speciation, we expect that forthcoming years
will certainly witness an increase in the importance given
to this process in animal evolution.
Nonetheless, detecting hybrid species will remain a challenge, for several reasons. First, by definition, hybrid taxa
are lineages with mixed ancestry; thus, their recognition
will depend therefore on the identification of lineages with
genetic polymorphisms typical of or associated with two
different parental species. However, the probability of a
successful interspecific cross decreases with divergence
between the hybridising species (Arnold 1997; Coyne & Orr
1997; Edmands 2002; Presgraves 2002; Price & Bouvier 2002;
Bolnick & Near 2005). Therefore, most hybrid taxa will have
closely related parental species, which share a significant
amount of genetic variation, and in which exclusive polymorphisms will be rare. Second, some scenarios of HHS
require of certain degree of backcrossing to a parental
species, which significantly reduces the genetic contribution
of the other parent to the hybrid genome, making detection
a challenge (e.g. Heliconius, Cottus). Finally, one or both
parental species may go extinct after the hybridisation pro-
cess (e.g. Mercenaria). Thus, it seems clear that most HHS cases
will be confirmed, or rejected, only after a sound statistical
genetic analysis of several markers has been performed in
combination with morphological, geographical and ecological information.
In summary, HHS taxa can (and should) be identified
using a multidisciplinary approach combining genetic, morphological, geographical and ecological data. The genetic
data in particular should be tested against predictions based
on coalescent simulations and/or the geographical discordance approach. Although HHS is unique among speciation
mechanisms in being amenable to experimental approaches,
this feature has been rarely exploited to date, and should be
employed in future studies in order to recreate the possible
routes to hybrid speciation and phenotypes and the genetic
architecture of hybrid traits. Moreover, research on animal
HHS has now reached a level of sophistication that will
allow it to shift from focusing simply on whether or not it
occurs in nature, to asking more nuanced questions about
the ecological and genetic conditions under which it occurs.
References
Arnold ML (1997) Natural Hybridization and Evolution. Oxford
University Press, Oxford, UK.
Barton NH (2001) The role of hybridization in evolution. Molecular
Ecology, 10, 551–568.
Barton NH, Hewitt GM (1985) Analysis of hybrid zones. Annual
Review of Ecology and Systematics, 16, 113–148.
Beltrán M, Jiggins CD, Brower AVZ, Bermingham E, Mallet J (2007)
Do pollen feeding, pupal-mating and larval gregariousness have
a single origin in Heliconius butterflies? Inferences from multilocus
DNA sequence data. Biological Journal of the Linnean Society, 92,
221–239.
Bertorelle G, Excoffier L (1998) Inferring admixture proportions from
molecular data. Molecular Biology and Evolution, 15, 1298–1311.
Bolnick DI, Near TJ (2005) Tempo of hybrid inviability in centrarchid
fishes (Teleostei: Centrarchidae). Evolution, 59, 1754–1767.
Buerkle CA, Morris RJ, Asmussen MA, Rieseberg LH (2000) The
likelihood of homoploid hybrid speciation. Heredity, 84, 441–451.
Coyne JA, Orr HA (1997) ‘Patterns of speciation in Drosophila’
revisited. Evolution, 51, 295–303.
Coyne JA, Orr HA (2004) Speciation. Sinauer & Associates, Sunderland, Massachusetts.
Dowling TE, DeMarais B (1993) Evolutionary significance of introgressive hybridization in cyprinid fishes. Nature, 362, 444–446.
Edmands S (2002) Does parental divergence predict reproductive
compatibility? Trends in Ecology & Evolution, 17, 520–527.
Gross B, Rieseberg L (2005) The ecological genetics of homoploid
hybrid speciation. Journal of Heredity, 96, 241–252.
Harrison RG (1990) Hybrid zones: windows on evolutionary processes. In: Oxford Surveys in Evolutionary Biology (eds Futuyma D,
Antonovics J), pp. 69–128. Oxford University Press, Oxford, UK.
Mallet J (2005) Hybridization as an invasion of the genome. Trends
in Ecology & Evolution, 20, 229–237.
Mallet J (2007) Hybrid speciation. Nature, 446, 279–283.
Mavarez J, Salazar C, Bermingham E et al. (2006) Speciation by
hybridization in Heliconius butterflies. Nature, 441, 868–871.
© 2008 The Authors
Journal compilation © 2008 Blackwell Publishing Ltd
H O M O P L O I D H Y B R I D S P E C I AT I O N I N A N I M A L S 5
McCarthy EM, Asmussen MA, Anderson WW (1995) A theoretical
assessment of recombinational speciation. Heredity, 74, 502–509.
McFadden C, Hutchinson MB (2004) Molecular evidence for the
hybrid origin of species in the soft coral genus Alcyonium (Cnidaria:
Anthozoa: Octocorallia). Molecular Ecology, 13, 1495–1505.
Presgraves DC (2002) Patterns of postzygotic isolation in Lepidoptera. Evolution, 56, 1168–1183.
Price TD, Bouvier MM (2002) The evolution of F1 postzygotic
incompatibilities in birds. Evolution, 56, 2083–2089.
Rieseberg L (1997) Hybrid origins of plant species. Annual Review
of Ecology and Systematics, 28, 359–389.
Salazar C, Jiggins CD, Taylor J, Kronforst MJ, Linares M (2008)
Hybrid speciation and the genealogical history of Heliconius
heurippa. BMC Evolutionary Biology, 8, 132.
Scribner KT, Page KS, Bartron ML (2000) Hybridization in freshwater fishes: a review of case studies and cytonuclear methods
of biological inference. Review of Fish Biology and Fisheries, 10,
293–323.
Seehausen O (2004) Hybridization and adaptive radiation. Trends
in Ecology & Evolution, 19, 198–207.
Taylor DJ, Sprengher HL, Ishida S (2005) Geographic and phylogenetic evidence for dispersed nuclear introgression in a daphniid
with sexual propagules. Molecular Ecology, 14, 525–537.
© 2008 The Authors
Journal compilation © 2008 Blackwell Publishing Ltd
Supporting Information
Additional Supporting Information may be found in the online
version of this article.
Table S1 Proposed examples of HHS in animals. Cases confirmed
with molecular markers are in bold. Mitochondrial DNA refers to
the parental source of mtDNA haplotypes in the hybrid taxon.
Nuclear DNA refers to the type of nuclear marker used in support
of HHS. Morphology refers to morphological categories considered for HHS taxa: identical to one parental, intermediate between
both or transgressive. Distribution refers to the geographical distribution of HHS taxa: restricted or widespread in relation to
parental spp., sympatric (S), parapatric (P) or allopatric (A). R.I.
refers to reproductive isolation between a HHS taxon and its
parental spp.
Please note: Blackwell Publishing is not responsible for the content
or functionality of any Supporting Information materials supplied
by the authors. Any queries (other than missing material) should
be directed to the corresponding author for the article.