Download PeterMoranPhDThesis - St Andrews Research Repository

A BEHAVIOURAL AND GENOMIC APPROACH TO STUDYING
THE EVOLUTION OF REPRODUCTIVE ISOLATION:
A CONTACT ZONE BETWEEN CLOSELY RELATED FIELD
CRICKETS IN THE GENUS TELEOGRYLLUS
Peter Moran
A Thesis Submitted for the Degree of PhD
at the
University of St Andrews
2017
Full metadata for this item is available in
St Andrews Research Repository
at:
http://research-repository.st-andrews.ac.uk/
Please use this identifier to cite or link to this item:
http://hdl.handle.net/10023/10260
This item is protected by original copyright
This item is licensed under a
Creative Commons Licence
A Behavioural and Genomic Approach to Studying
the Evolution of Reproductive Isolation:
A Contact Zone between Closely Related Field
Crickets in the Genus Teleogryllus
Peter Moran
This thesis is submitted in partial fulfilment for the degree of PhD
at the
University of St Andrews
Date of Submission
23/09/2016
Abstract
What processes contribute to the evolution of reproductive isolation and the
coexistence of interfertile species in the same habitat? This thesis investigates the relative
roles of species interactions and intraspecific processes in contributing to reproductive
isolation. I combine behavioural and genomic approaches to test hypotheses about what
mechanisms maintain the general species boundary between two closely related field cricket
species: Teleogryllus oceanicus and T. commodus. These species are a classic study system
for sexual communication and readily hybridize in the laboratory, however little is known
about species interactions in sympatric populations. I examine patterns of geographic
variation in two key sexual traits: calling song and cuticular hydrocarbons (CHCs), and the
geographic distribution of genetic variation across a broad sample of allopatric and sympatric
populations. I test whether X chromosomes play a pronounced role in population divergence
and reproductive isolation. Using close range mating trials and hybridization experiments I
identify numerous pre-mating and post-mating barriers between the species. The results
indicate that the species are currently reproductively isolated and the pattern of population
differentiation does not strongly support contemporary species interactions contributing to
phenotypic diversity. Numerous barriers exist between the species, in particular hybrid
females are sterile in both cross directions, while hybrid males are relatively fertile. This
provides a rare exception to Haldane’s rule which is central to many genetic theories of
speciation. Established theory predicts that X chromosomes should play a pronounced role in
the evolution of both pre- and postzygotic barriers. Contrary to this, I found no evidence that
X chromosomes contribute to hybrid female sterility. Moreover, X-linked loci exhibited an
unexpected pattern of reduced population differentiation within species, but increased species
divergence compared to autosomal loci, which may indicate selective sweeps or sex-biased
processes. Taken together, the results suggest that the causes and consequences of X
chromosome evolution, in particular among XO taxa, may contradict some of the established
theories.
i
Acknowledgements
Firstly, I would like to sincerely thank my supervisor Nathan Bailey for his continued
guidance and support throughout my four years in his lab. He has sharpened my scientific
thinking and was an ideal supervisor in providing the encouragement, guidance and patience
to allow me to grow as a research scientist. I would also like to thank my co-supervisor Mike
Ritchie for his valuable guidance and assistance throughout my time in the HMB. I felt lucky
to have two supervisors that always made the time to talk (science or non-science topics) and
that were inspiring in their passion for evolutionary biology.
Thanks go aswell to my lab mates, Sonia and Will, for making cricket maintenance and lab
meetings a lot more fun. I would like to especially thank Sonia for her advice and continued
assistance.
While they are too numerous to mention by name here, I would like to thank all my office
mates and the staff throughout my time in the HMB, for making it a truly enjoyable place to
work. In addition, I would like to thank the members of my committee, Christian Rutz and
David Ferrrier, for their advice and support.
Next, I thank all the friends that I’ve made in St. Andrews for the many fun evenings, swims
in the sea and for making sure I didn’t get stuck in the ivory basement. Special thanks go to
Nora, Adalvan, Braulio, Kieran, Illinca, Lisa, Diana, Lizzie, Riccardo, Behrouz, and Marco.
In particular, thank you Nora for always being there, sharing hours of free flowing
conversations and pushing me intellectually.
Finally, a deep thank you to my family for their endless support, without them none of this
would have happened.
My research was funded by the Natural Environment Research Council (NERC) and a small
field work grant from the Orthopterists’ society for which I am very grateful.
ii
Co-authorship statement
1. Chapter one – Independent review of the field of speciation research under the guidance
of Nathan Bailey and Mike Ritchie.
2. Chapter two – Cricket sampling in the wild was organised by Nathan and we were
greatly assisted by two field assistants, Katie Holmes and Tony Ly. I designed and
conducted the experiments. For the song data, I took all the recordings and analysed the
data. The work on CHCs was done in collaboration with John Hunt and Melia Burdon at
the University of Exeter. I prepared and extracted the CHCs. Melia ran GCMS on the
sample extracts and devised the methods for peak integration. I measured the CHC peak
abundances and conducted the statistical analyses. I wrote the chapter under the guidance
of both Nathan and Mike.
3. Chapter three – I conducted the experiments and performed the statistical analysis. The
work on CHCs was done in collaboration with John Hunt and Chris Mitchell at the
University of Western Australia. I prepared and extracted the CHCs. Chris ran GCMS on
the sample extracts and integrated the data. I performed statistical analyses on the
spreadsheet of the peak integrated data he gave me. I wrote the chapter under the
guidance of both Nathan and Mike.
4. Chapter four - I designed and conducted the experiments, performed the analysis and
wrote the manuscripts under the guidance of both Nathan and Mike.
5. Chapter five – The genomics component was done in collaboration with a number of
people. I collected the samples in the field. Sonia Pascoal (now at the University of
Cambridge) prepared the DNA for RAD-sequencing. Edinburgh Genomics did the
sequencing and Tim Cezzard and Judith Risse provided valuable bioninformatic support.
In particular Tim produced the three different de novo assemblies and provided me with a
VCF file, from which I did the downstream analysis. The candidate X-linked markers
were identified using a script written by Tim and modified by Judith. I conducted the
population genomic analysis and wrote the chapter under the guidance of both Nathan
and Mike.
6. Chapter six - Independent summary of the findings and the broader implications for
speciation research, written under the guidance of Nathan and Mike.
iii
Table of Contents
Abstract ...................................................................................................................................i
Acknowledgements ................................................................................................................ii
Co-authorship statement ....................................................................................................... iii
Table of Contents ..................................................................................................................iv
List of Figures .......................................................................................................................vi
List of Tables ........................................................................................................................vii
Declarations ........................................................................................................................ viii
Chapter 1 Introduction....................................................................................... 1
What are the processes that drive population divergence? .................................................... 2
Evolutionary outcomes of secondary contact......................................................................... 4
Selection against hybrids and the genetics of species differences ......................................... 8
Is there a special role for sex chromosomes in the evolution of reproductive isolation? ...... 9
The study of hybridization and reproductive isolation: evolving approaches ..................... 11
Study system ........................................................................................................................ 13
Objectives ............................................................................................................................. 20
Chapter 2 Geographic Variation: Long vs. Short Range Mating Signals ... 22
Abstract ................................................................................................................................ 22
Introduction .......................................................................................................................... 23
Material & Methods ............................................................................................................. 30
Results .................................................................................................................................. 38
Discussion ............................................................................................................................ 57
Supplementary...................................................................................................................... 69
Chapter 3 Sexual Isolation and the Evolution of Sexually Selected Traits . 82
Abstract ................................................................................................................................ 82
Introduction .......................................................................................................................... 83
Methods ................................................................................................................................ 92
Results ................................................................................................................................ 103
Discussion .......................................................................................................................... 115
Supplementary.................................................................................................................... 123
iv
Chapter 4 A Rare Exception to Haldane’s Rule: Are X Chromosomes Key
to Hybrid Incompatibilities? .......................................................................... 131
Abstract .............................................................................................................................. 131
Introduction ........................................................................................................................ 132
Methods .............................................................................................................................. 136
Results ................................................................................................................................ 142
Discussion .......................................................................................................................... 150
Supplementary.................................................................................................................... 157
Chapter 5 Genomic Discordance: Unusual Patterns of Differentiation at
Autosomal vs. X-linked Loci .......................................................................... 158
Abstract .............................................................................................................................. 158
Introduction ........................................................................................................................ 159
Material & Methods ........................................................................................................... 166
Results ................................................................................................................................ 175
Discussion .......................................................................................................................... 186
Supplementary.................................................................................................................... 194
Chapter 6 General discussion ........................................................................ 205
What can the contact zone between T. oceanicus and T. commodus tell us about
speciation? .......................................................................................................................... 205
What prevents hybridization and fusion of interfertile species? ........................................ 206
Is ecological differentiation a requirement for species coexistence? ................................. 208
Do sex chromosomes play a pronounced role in species isolation? .................................. 210
Identifying hybrid genetic incompatibilities ...................................................................... 214
Testing the unusual patterns of X chromosome differentiation ......................................... 216
Conclusions ........................................................................................................................ 218
References ........................................................................................................ 220
v
List of Figures
Figure 1-1.The distribution of the three putative species across Australia .............................. 15
Figure 2-1 Distribution of the sixteen field sites in eastern Australia ..................................... 30
Figure 2-2 Calling song oscillograms for both species. ........................................................... 32
Figure 2-3 Typical gas chromatography profile of the two species......................................... 34
Figure 2-4 PCA for calling song: field and lab recordings ...................................................... 40
Figure 2-5 PCA and DAPC for T. commodus CHCs - (incls strange samples) ....................... 42
Figure 2-6. T. oceanicus song population differentiation (DAPC) .......................................... 45
Figure 2-7. T. oceanicus CHC population differentiation (DAPC) ......................................... 48
Figure 2-8. T. commodus song population differentiation (DAPC)......................................... 50
Figure 2-9. T. commodus CHC population differentiation (DAPC) ........................................ 52
Figure 2-10. T. oceanicus song, CHCs and geographic distance............................................. 54
Figure 2-11. T. commodus song, CHCs and geographic distance ........................................... 55
Figure 3-1. Calling song oscillograms: parental species and the reciprocal hybrids ............... 96
Figure 3-2 Courtship song oscillograms: parental species and the reciprocal hybrids ............ 97
Figure 3-3 Typical chromatogram of the two parental species................................................ 99
Figure 3-4 Close range courtship trial results: conspecific and heterospecific pairs ............. 104
Figure 3-5 Close range courtship trial results: backcrosses with hybrid males and females . 107
Figure 3-6 Calling song inheritance pattern........................................................................... 109
Figure 3-7 Courtship song inheritance pattern....................................................................... 111
Figure 3-8 CHCs inheritance pattern ..................................................................................... 113
Figure 4-1 Schematic of the cross design .............................................................................. 139
Figure 4-2 Results from all three generations of crosses ....................................................... 145
Figure 5-1 DNA sample collection sites ................................................................................ 167
Figure 5-2 FastStructure analysis........................................................................................... 176
Figure 5-3 Genomic clines analysis ....................................................................................... 177
Figure 5-4 Principal component analysis ............................................................................... 179
Figure 5-5 Hierarchical clustering trees ................................................................................. 180
Figure 5-6 Genetic differentiation and geographic distance .................................................. 185
Figure 6-1 Selective paternal X-inactivation? ....................................................................... 212
vi
List of Tables
Table 2-1. T. oceanicus calling song loadings (DAPC)........................................................... 44
Table 2-2. T. oceanicus: Pairwise population comparisons for song discriminant functions
(LDs) ........................................................................................................................................ 46
Table 2-3. T. commodus calling song loadings ........................................................................ 49
Table 2-4. T. commodus: Pairwise population comparisons for song discriminant functions
(LDs) ........................................................................................................................................ 51
Table 3-1 Courtship trials: cross types and sample sizes for F1 crosses and backcrosses ...... 93
Table 3-2 Sexual trait inheritance: Cross types and sample sizes ........................................... 96
Table 3-3 Reproductive isolation index ................................................................................. 104
Table 3-4 GLM results courtship trials amongst conspecific and heterospecific pairs ......... 105
Table 3-5 GLM results courtship trials amongst backcross pairs .......................................... 107
Table 3-6 Summary data sexual trait inheritance pattern ..................................................... 114
Table 4-1 F1 crosses: GLM results ........................................................................................ 143
Table 4-2 BC1 crosses: GLM results ..................................................................................... 146
Table 4-3 BC2 crosses: GLM results ..................................................................................... 149
Table 5-1 Pairwise population FST: Autosomal vs. X-linked loci ......................................... 183
Table 5-2 Population genetic summary statistics for autosomal and X-linked loci .............. 184
Table 5-3 Mantel test results .................................................................................................. 185
vii
Declaration
viii
ix
“All the happiness the earth possesses in not being broken down into
matter and spirit was contained in the unique sound of the cricket.”
Step. E. (1916) Marvels of insect life; a popular account of structure and habit. New York, R.M. McBride and
Co. (Not in copyright: Source http://www.biodiversitylibrary.org/bibliography/9336#/summary)
Quote: Stéphane Mallarmé (1867)
x
xi
Chapter 1 Introduction
Understanding the processes that contribute to the origin and maintenance of species
boundaries is a central aim of evolutionary biology. One of the most influential ideas of the
modern synthesis, which was extensively developed by Ernst Mayr (1942) and Theodosious
Dobzhansky (1935, 1937) was that speciation occurred almost entirely through geographic
isolation and the restriction of gene flow. This led to what became known as the biological
species concept, which emphasized a central role of pre- and post-mating isolating barriers in
sealing off distinct gene pools (Mayr, 1942). Reproductive barriers were believed to be
unable to evolve in populations that interbreed due to the homogenising effect of gene flow.
Since then, extensive theoretical and molecular work has confirmed that allopatric divergence
is pervasive (Coyne & Orr, 2004). It is widely accepted as the main mode of speciation,
although species may sometimes arise and persist in the presence of gene flow (Via, 2001;
Mallet, 2005; Mallet et al., 2007; Feder et al., 2012; Gagnaire et al., 2013; Keller et al., 2013;
Seehausen et al., 2014; Schumer et al., 2014). However, closely related species often come
into contact (i.e. parapatry or sympatry) before reproductive isolation is complete. Therefore,
what are the mechanisms and evolutionary forces that maintain species boundaries upon
secondary contact (Harrison & Larson, 2014)? In particular, what is the relative role of
prezygotic and postzygotic barriers in reproductive isolation and do certain genetic
architectures, such as sex-linkage of reproductive barriers, facilitate species isolation?
1
What are the processes that drive population divergence?
Allopatric speciation is conceptually clear: given enough time, divergence among
geographically isolated populations is inevitable. Species differences are predicted to
accumulate independently within each lineage due to selective and stochastic processes
(Turelli et al., 2001). These processes are likely to interact and include ecological adaptation
(Schluter, 2001; Funk et al., 2006; Nosil et al., 2009), sexual selection (Lande, 1981;
Kirkpatrick, 1982; Payne & Krakauer, 1997), mutation and genetic drift (Nei et al., 1983;
Shuker et al., 2005) resulting in multiple reproductive barriers that contribute to species
isolation (Butlin, 1998; Howard et al., 2002; Coyne & Orr, 2004; Tyler et al., 2013). The role
of divergent natural selection in speciation is well-supported, and it can arise from numerous
environmental agents, encompassing both biotic and abiotic elements of the habitat (Schluter,
2001, 2009; Nosil, 2012). Environmental differences have the potential to influence a broad
range of behavioural and physiological traits resulting in pre- and post-mating barriers, in
allopatry or sympatry. In contrast, there is much less consensus about the role of sexual
selection in speciation (Panhuis et al., 2001; Ritchie, 2007; Svensson & Gosden, 2007;
Kraaijeveld et al., 2011; Safran et al., 2013).
Sexual selection
Darwin (1871) first proposed sexual selection as a cause of sexual dimorphism and
possibly speciation. However, during the modern synthesis the role of sexual selection in
speciation was largely ignored, because behavioural isolation was seen either as a pleiotropic
byproduct of divergent natural selection, or as a result of selection for “species recognition”
(Coyne & Orr., 2004, p213). Sexual traits involved in “species recognition” were believed to
change principally through natural selection, in particular through reinforcement to reduce
2
maladaptive hybridization (Dobzhansky, 1940). It was only in the early 1980s, almost a
century after Darwin first proposed it (Darwin, 1871), that the role of sexual selection in
population divergence and speciation was fully appreciated due to the realization that mate
choice and species recognition likely reflect a continuum (West-Eberhard, 1983).
Rapid signal divergence amongst populations may occur through a runaway Fisher
process, due to a genetic correlation between signal and receiver traits (Fisher, 1930; Lande,
1981, 1982; Kirkpatrick, 1982). The seminal sexual selection models by Lande treated female
preferences as effectively neutral, primarily evolving as a correlated response to selection on
male traits, and slight perturbations such as environmental heterogeneity, drift and variation
in selection on males could initiate runaway sexual selection independently in different
populations. Numerous models of sexual selection exist and tend to focus on female mate
preferences (reviewed extensively in Kirkpatrick & Ryan, 1991; Mead & Arnold, 2004;
Kokko et al., 2013). Nearly all models can lead to divergent behavioural isolation among
populations, however some forms of mate preferences may also constrain speciation. Sensory
bias, where the preference for a trait evolved in a different context than mate choice, such as
prey interactions, can lead to “open ended” preferences which could cause females to mate
with heterospecifics (Basolo, 1990; Kirkpatrick & Ryan, 1991; Ryan & Cummings, 2013).
Interaction between sexual and ecological selection
In most cases natural and sexual selection are likely to interact (Blows, 2002) and
distinguishing their relative role in population divergence and speciation is a current area of
focus (Doorn et al., 2009; Maan & Seehausen, 2011; Safran et al., 2013). The distinction
between natural and sexual selection may not be clear as most sexually selected traits are
likely to be influenced by both processes (Svensson & Gosden, 2007). In some cases the
3
distinction may be important as they are likely to operate in different ways and may oppose
or reinforce each other during divergence (Safran et al., 2013). Divergent ecological selection
can alter sexual communication systems, such as signal divergence to maximize transmission
properties in different environments (Boughman, 2001), which can have knock on effects on
mate recognition and mating systems. Comparative studies which compare species diversity
with estimates of the strength of sexual selection have indicated that sexual selection may
play a more pronounced role in early stages of divergence, in particular during adaptive
radiations (Kraaijeveld et al., 2011).
Evolutionary outcomes of secondary contact
A number of important evolutionary outcomes can occur when closely related species,
which are incompletely reproductively isolated, come into secondary contact, with the two
extremes being: hybridization and introgression, or reinforcement and reproductive character
displacement (Harrison, 1993; Coyne & Orr, 2004). Distinguishing between secondary
contact following allopatric divergence versus divergence with gene flow (i.e. parapatry) is
often challenging and taxa may come in and out of contact multiple times blurring the
distinction. A critical objective in speciation research is determining under what conditions
pre-existing behavioural or ecological species differences break down due to gene flow and
recombination, as opposed to becoming reinforced and strengthening species boundaries
(Abbott et al., 2013).
Hybridization and introgression
Hybridization and introgression can erode species boundaries (Harrison & Larson,
2014) resulting in the extinction of the rarer species (Rhymer & Simberloff, 1996; Seehausen
4
et al., 1997), the formation of stable hybrid zones (Barton & Hewitt, 1985; Harrison, 1993) or
even hybrid speciation (Mavárez et al., 2006; reviewed by Mallet, 2007; Keller et al., 2013;
Schumer et al., 2014). Allopolyploidy is one mechanism through which hybrid speciation can
occur and is more common among plants than animals (Mallet, 2007; Wood et al., 2009).
Investigations into the causes and consequences of hybridization began in the earlyeighteenth century with the work of Thomas Fairchild (1717; as noted by Zirkle, 1934) and
later by Linnaeus, who believed new species could arise from hybridization between
previously created forms (Arnold, 2006). Over the following century, experimental
hybridization studies often revealed negative fitness costs with hybrids suffering reduced
fertility and viability, compared to parental species. This led Darwin and his contemporaries
to place little importance on the role of hybridization and genetic exchange in evolutionary
diversification. The iconic view of the tree of life, with organisms evolving discretely and
independently in a branch-like manner became the dominant view during the modern
synthesis (Mayr, 1963). However, since the time of Darwin evidence had begun to emerge
that introgressive hybridization could be an important driver of evolutionary diversification
and had played an important role in the evolution of plant species (Anderson, 1949; Anderson
& Stebbins, 1954; Stebbins, 1959). Indeed, hybridization can be beneficial due to reducing
deleterious homozygosity through outcrossing (Grant et al., 2003), contribute genetic
variation faster than mutational processes, with some of these variants potentially having
adaptive or novel functions (Mallet, 2005; Rieseberg, 2009; Hedrick, 2013; Martin et al.,
2014) and can confer selective advantages under certain environmental conditions (Pfennig,
2007; Schmidt & Pfennig, 2015). Increasingly molecular studies are revealing that
hybridization and introgression may be frequent and can play an important role in animal
diversification (about 1-10% of all animal species and 25% of plants suggested to hybridise
with at least one other species; Mallet, 2005).
5
Reinforcement & reproductive character displacement
If sibling species overlap geographically and heterospecific matings are costly,
selection is predicted to enhance assortative mating through reinforcement by driving signal
or preference divergence (or both) (Dobzhansky, 1940; Coyne & Orr, 2004). Here I adopt the
“broad sense” definition of reinforcement, “the evolution of prezygotic isolating barriers in
zones of overlap or hybridization (or both) as a response to selection against hybridization”
(Howard, 1993, p46). Selection against interspecific mating’s has traditionally focused on
intrinsic incompatibilities, such as reduced viability or fertility of hybrids (J. A. Coyne &
Orr., 2004; E. M. Lemmon & Lemmon, 2010), and to a lesser extent on extrinsic
incompatibilities (Nosil et al., 2003; Servedio, 2004). However, reinforcement may occur
even when hybrids are viable and fertile but produced in fewer numbers or when interspecific
matings incur a direct fitness cost, such as reduced longevity or fecundity on the mating
partners ( Servedio, 2001; Servedio & Noor., 2003). The main signature of reinforcement is
reproductive character displacement, which is a pattern of enhanced prezygotic isolation in
sympatry compared to in allopatry (Brown & Wilson, 1956; Howard, 1993). However,
reproductive character displacement can manifest due to processes other than reinforcement,
such as signal interference, which can often be difficult to distinguish (Howard, 1993). In
addition, species interactions can have outcomes other than reproductive character
displacement, such as selection for a reduction in hybrid dysfunction (Ritchie et al., 1992;
Virdee & Hewitt, 1994), increased ecological differentiation (Schluter, 2001; Pfennig &
Pfennig, 2009) or polyandry and conspecific sperm precedent (Marshall et al., 2002).
Widespread acceptance of reinforcement has waxed and waned since it was first
suggested by Dobzhansky (1940), due to several major theoretical obstacles, but today the
use of more complex models suggest it is plausible and even probable under certain
6
conditions (Liou & Price, 1994; Kelly & Noor, 1996; Turelli et al., 2001; Kirkpatrick &
Ravigné, 2002; Servedio & Noor., 2003; Coyne & Orr., 2004). One of the main theoretical
objections to reinforcement is that gene flow and recombination should break down the
genetic association between pre-zygotic barriers arising from the process of reinforcement
and post-zygotic barriers that generate selection against hybridisation in the first place
(Felsenstein, 1981). However, if the underlying genes reside in regions of low recombination,
such as sex chromosomes or in chromosomal inversions, this can help maintain linkage
disequilibrium (Lemmon & Kirkpatrick, 2006). Another argument against reinforcement is
that it is “self-defeating”, as any increase in pre-zygotic isolation reduces the strength for
further reinforcement. Liou & Price, (1994) showed that sexual selection can increase the
likelihood of reinforcement as two forces now drive the evolution of female preference;
natural selection on female preferences to reduce costly mismatings and indirect selection on
females due to sexual selection on males.
Convincing empirical examples that reinforcement operates come mainly from three
types of studies: observations from nature contrasting the strength of pre-zygotic barriers
among allopatric and sympatric populations (Saetre et al., 2003; Servedio & Noor., 2003;
Nosil et al., 2007; Hopkins et al., 2014; Rundle & Dyer, 2015), selection experiments (Higgie
et al., 2000), and comparative approaches (Coyne & Orr, 1989, 1997). Few comparative
studies exist outside of Drosophila, and broader taxonomic representation is needed to
confirm how generalizable the pattern is.
7
Selection against hybrids and the genetics of species differences
The nature of selection against hybrids and the genetic architecture of traits
contributing to pre-zygotic and/or post-zygotic barriers are both likely to be important in
determining the outcome of species interactions (Servedio & Saetre, 2003; Servedio, 2004;
Ortiz-Barrientos et al., 2009). Hybridization in general is likely to have negative fitness
effects due to hybrids exhibiting both genetic and ecological dysfunction (Coyne & Orr.,
2004). Post-zygotic isolation may manifest in hybrids as a result of inviability or intrinsic or
extrinsic sterility. Intrinsic hybrid sterility encompasses two forms: physiological, where
hybrids fail to produce gametes (Maheshwari & Barbash, 2011), and behavioural, where
hybrids suffer a defect which renders them incapable of normal courtship behaviour (Clark et
al., 2010). Extrinsic postzygotic isolation also encompasses two forms; ecological inviability,
where hybrids are maladapted to the environment (Gow et al., 2007; Nosil et al., 2005) and
behavioural sterility, in which hybrids exhibit intermediate courtship behaviours that fail to
elicit mating (Naisbit et al., 2001). The distinction between intrinsic and extrinsic behavioural
isolation is subtle but may be important, as the former represents a physiological or
neurological defect, suggesting intrinsic behavioural sterility may conform to more common
patterns of intrinsic genetic incompatibilities such as Haldane’s rule (Davies et al., 1997;
Coyne & Orr., 2004). However, too few studies have examined hybrid courtship behaviour in
detail to test if this is indeed the case.
8
Is there a special role for sex chromosomes in the evolution of
reproductive isolation?
Sex chromosomes (X or Z) have been suggested to play a pronounced role in the
evolution of both pre-zygotic and post-zygotic barriers (Charlesworth et al., 1987; Reinhold,
1998; Saetre et al., 2003; Servedio & Saetre, 2003; Coyne & Orr., 2004; Qvarnström &
Bailey, 2009). The asymmetry in inheritance and expression of sex chromosomes means
evolutionary forces may act differently on them compared to autosomes (Charlesworth et al.,
1987; Vicoso & Charlesworth, 2006). In particular, recessive mutations which increase the
fitness of males (i.e. heterogametic sex) are expected to accumulate on the X-chromosome as
recessive mutations are immediately exposed to selection in the heterogametic sex
(Charlesworth et al., 1987). In addition, dominant alleles which are beneficial to females
(homogametic sex) are expected to accumulate on the X, as the X spends ¾ of its time in
females. This may be especially important for sexually selected traits as they are usually sexlimited and may have antagonistic sexual effects (Rice, 1984; Andersson, 2004).
One of the central questions in speciation research, is how the traits involved in
reproductive isolation remain associated in the face of gene flow. Physical linkage of the
associated genes can suppress or reduce recombination. In particular, sex-linkage can
promote strong linkage disequilibrium among traits as recombination on sex chromosomes is
restricted to the homogametic sex (Qvarnström & Bailey, 2009). Strong linkage
disequilibrium between mate recognition traits may mitigate the effects of gene flow and
recombination in sympatry and prevent the breakdown of assortative mating. Moreover, if
both prezygotic and postzygotic barriers are sex linked this can increase selection for
assortative mating i.e. reinforcement, which can eventually complete speciation (Noor et al.,
2001; Servedio & Saetre, 2003; Lemmon & Kirkpatrick, 2006). Two interesting examples of
9
the role of sex chromosomes in reproductive isolation comes from hybridizing pied
flycatchers and Gouldian finches in which genes causing hybrid incompatibility and genes
influencing male plumage are both on the Z chromosome (Sæther et al., 2007; Pryke &
Griffith, 2009; Pryke, 2010). This strong genetic association between traits involved in
reproductive isolation is likely to have promoted reinforcement and reproductive character
displacement in male plumage and female preference among sympatric flycatchers.
How common is sex-linkage of mate recognition traits (i.e. signals and preferences)
and how often are both prezygotic and postzygotic traits sex linked? Empirical results for
sex-linkage of prezygotic barriers have been mixed, with some studies supporting sex-linkage
in female heterogametic species (Z-linkage in Lepidoptera and some birds) but generally not
for male heterogametic species (Reinhold, 1998; Ritchie & Phillips, 1998; Prowell, 1998;
Orr, 2001; Saetre et al., 2003; Qvarnström & Bailey, 2009). In contrast, sex chromosomes
show a consistent and disproportionate involvement in post-zygotic incompatibilities. J. B. S.
Haldane was the first to note that in crosses between closely related species, if either sex of
the offspring suffers disproportionate fitness costs, such as reduced fertility or viability, it
will be the heterogametic sex (Haldane, 1922). Since then this pattern has been observed
across a broad range of taxa (e.g. mammals, birds, reptiles, amphibians, fish, insects,
nematodes and the plant genus Silene), irrespective of whether males or females are
heterogametic, implicating sex chromosomes rather than gender per se in post-zygotic
incompatibilities (Coyne & Orr., 2004; Schilthuizen et al., 2011). Haldane’s rule is important
as it suggests a surprising uniformity in the underlying mechanisms of intrinsic hybrid
dysfunction and it ignited extensive research which contributed to our understanding of the
genetic basis of post-zygotic isolation (Maheshwari & Barbash, 2011). Why are there no
similar generalizations for prezygotic traits? The genetic basis of prezygotic barriers have
been less well studied but may exhibit less consistent patterns, such as sex-linkage, due to the
10
complexity of traits involved, whose development, expression and fitness is often dependent
on the environment (Grant & Grant, 1993; Pfennig, 2007).
The study of hybridization and reproductive isolation: evolving
approaches
Areas where species overlap or hybridize (or both) provide an important natural
laboratory for identifying the mechanisms and processes involved in the divergence and the
maintenance of species boundaries (Hewitt, 1988). The advent of new molecular genetic
techniques in the 1980s promoted a renaissance of interest in the genetic basis of species
differences and hybrid zone research, which continues unabated today (Harrison, 1993;
Butlin, 1998; Abbott et al., 2013; Gompert & Buerkle, 2016; Harrison & Larson, 2016).
Advancements in next generation sequencing, in particular, reduced representation methods
(i.e. RADseq), provide cost-effective opportunities to examine intraspecific and interspecific
patterns of genetic diversity in non-model organisms (Luikart et al., 2003; Davey & Blaxter,
2010; Andrews & Luikart, 2014; Andrews et al., 2016). The power of RADseq lies in its
ability to discover and genotype individuals for thousands of genetic markers, which are
more-or-less randomly distributed across the genome (including both coding and non-coding
regions) (Baird et al., 2008; Davey & Blaxter, 2010). In the absence of a reference genome,
the short reads (~100 – 500 base pairs) can be mapped against a de novo assembled cluster of
reads and single nucleotide polymorphisms (SNPs) called (Catchen et al., 2013). The
availability of genome-wide markers provides a fine scale resolution to identify regions of the
genome which may contribute to local adaptation and pre- and post-zygotic barriers (Nosil &
Feder, 2012).
11
Two main population genetic methods have been used to examine patterns of
introgression in hybrid zones; geographical clines, which examine the relationship between
geographic distance and allele frequencies (or phenotypes), and genomic clines, which
examine the change in allele frequencies along a genome-wide admixture gradient (i.e. hybrid
index which is the proportion of alleles that were inherited from a pair of hybridizing
species). Genetic regions which are involved in reproductive isolation can be inferred based
on comparing the shape and widths of geographical clines (Barton & Hewitt, 1985) and the
concordance of genomic clines with a null model (Rieseberg et al., 1999; Carling &
Brumfield, 2008; Gompert & Buerkle, 2009; Teeter et al., 2010; Rieseberg et al., 1999).
However, geographic clines are highly dependent on the spatial scale of sampling and the
geography of hybrid zones can be highly variable, ranging from discrete junctions with steep
multilocus clines (Payseur et al., 2004; Kawakami et al., 2009; Janoušek et al., 2012; Vines,
2014) to mosaic hybrid zones with a complex patchwork of parental genotypes (Harrison &
Rand, 1989; Harrison, 1993). In the latter case genomic clines may be more suitable as they
avoid some of these spatial problems. However, defining the appropriate null model in
genomic cline analysis is an area of debate (Gompert & Buerkle, 2010, 2011; Gompert et al.,
2012; Fitzpatrick, 2013).
The pattern that is emerging from genomic analysis of hybrid zones is that species
boundaries are often semi-permeable with high variability in differential introgression across
the genome (Harrison, 1993; Payseur et al., 2004; Nosil et al., 2007; Nosil et al., 2009;
Harrison & Larson, 2014; Harrison & Larson, 2016; Roux et al., 2016). X-linked loci often
introgress much less compared to autosomal loci, consistent with X-chromosomes harbouring
genes involved in pre-zygotic and post-zygotic barriers (Saetre et al., 2003; Payseur et al.,
2004; Janoušek et al., 2012; Maroja et al., 2015). However, distinguishing differential
introgression, ancestral polymorphisms and variation in selection or recombination across the
12
genome remains a pressing issue in many cases (Bierne et al., 2011; Cruickshank & Hahn,
2014; Harrison & Larson, 2016; Roux et al., 2016).
Study system
Crickets as a model system for speciation
Crickets have long played a key role in research on sexual selection and speciation
(Otte, 1989; Gray & Cade, 2000; Broughton & Harrison, 2003; Mendelson & Shaw, 2005;
Maroja et al., 2009; Veen et al., 2012; Maroja et al., 2015; Gray et al., 2016). Studies have
generally focused on their acoustic behaviour due to their conspicuous, and generally speciesspecific, advertisement songs (Otte, 1992). Tuning in on cricket song offers the opportunity
to identify putative species which are otherwise morphologically indistinguishable (i.e.
cryptic species) and to easily map the geographic and ecological distribution of a species
(Otte & Alexander, 1983; Otte & Cade, 1983; Otte et al., 1987; Otte, 1992). In addition, their
tractability for rearing in the lab and their relatively large neurons make them highly
amenable to neurophysiological experiments (Huber et al., 1989).
Acoustic communication plays a primary role in cricket social behaviour, by
attracting conspecific females and mediating aggression and territoriality (Alexander, 1961,
1962; Otte, 1992). Males express sex-limited singing behaviour. They rub together a file and
scraper on their forewings by raising the wings and then repeatedly opening and closing them
(stridulation), with each closure resulting in a single sound pulse. Males generally produce
three types of calls: calling song which attracts conspecific females from a distance, courtship
song used at close range to stimulate the female to mount the male, and aggressive song to
maintain social hierarchy and territory (Alexander, 1961). All three songs are believed to
have evolved from an ancestral courtship song (Alexander, 1962; Otte, 1992).
13
Natural history of Teleogryllus species
Male crickets of the genus Teleogryllus produce some of the most complex calling
songs of all true crickets (family Gryllidea) with alternating chirp and trill components
(Figure 1-1 (below)) (Otte, 1992). The genus Teleogryllus is believed to have originated in
South Africa and spread eastwards throughout Asia, Australia and the Pacific, encompassing
ca. 35 species (although no molecular phylogeny exists) (Otte & Cade, 1983; p67-68). The
genus was revised by Chopard (1961) initially based on male genitalia and for many years a
unitary Australian species of Teleogryllus was recognized. Since then three Australian
species have been proposed, based on differences in their physiology, karyotype, geographic
distribution and acoustic behaviour: T. commodus, T. oceanicus and T. marini (Fontana &
Hogan, 1969; Otte & Alexander, 1983). T. oceanicus occurs in tropical and subtropical
regions throughout Oceania and northern coastal regions of Australia, while T. commodus is
believed to be restricted to the southern and eastern regions of Australia (Figure 1-1) (Otte &
Alexander, 1983). Both species coexist across a broad area of sympatry in southern
Queensland (~23.37ᵒS – 26.38ᵒS latitude (Hill et al., 1972) while T. marini has only been
observed in a restricted area of north eastern Queensland (Figure 1-1) (Otte & Alexander,
1983).
Little is known about the ecology of the species and if they have divergent habitat
preferences or resource use. All three species are believed to be ecological generalists, living
in cracks or simple burrows, in grassy and weedy areas, and mainly active at night (Otte &
Alexander, 1983). Morphologically, T. commodus and T. oceanicus, are almost
indistinguishable, but males can sometimes be identified based on differences in stridulatory
file numbers. T. commodus is a univolitine species, with the potential to egg-diapause as an
adaptation to the colder winters of the southern regions, while T. oceanicus breeds
continuously (Fontana & Hogan, 1968; Hogan, 1971). Very little is known about T. marini,
14
which has only been previously described once (Otte & Alexander, 1983). Aside from its
distinct calling song (primarily distinguished by a low carrier frequency ca. 3kHz and triplets
of pulses in the trill, Figure 1-1), the males have reportedly divergent genitalia, suggesting
reproductive isolation may be complete. Early studies highlight confusion over the taxonomy
of these species and a modern revision of their genetic relationship and geographic
distribution is needed.
Figure 1-1.The distribution of the three putative species across Australia
T. oceanicus is found across the north (represented by light grey area), T. commodus is
mainly restricted to the south-eastern coast (dark grey areas), but has also been documented
in the south-west and the red dot in the north indicates the site where we sampled the putative
closely related species T. marini (Otte & Alexander, 1983). T. oceanicus and T. commodus
overlap in an area ~400km long on the mid-eastern coast, highlighted by wavy white lines.
Oscillograms of male calling songs from each species are shown. A female T. commodus and
male T. oceanicus are also shown in close contact (photo credit Liam Dougherty).
15
Sexual traits and reproduction
Teleogryllus oceanicus and T. commodus, have been a model study system for the
genetics and neurophysiology of acoustic behaviour due in part to the complexity of their
songs (Huber et al., 1989). Seminal work by Hoy and colleagues examined the behavioural
and neurophysiological basis of calling song and female preferences, often using reciprocal
hybrid crosses to test the underlying genetics (Hoy, 1974; Hoy et al., 1977; Pollack & Hoy,
1979; Pollack & Hoy, 1981; Hoy et al., 1982; Doherty & Hoy, 1985). Their research
provided tentative evidence for the “genetic coupling” hypothesis, originally proposed by
Alexander (1962), that sexual signals and mate preferences may be under shared genetic
control i.e. pleiotropy (or linkage). Genetic coupling provides a conceptually clear
explanation for the coordinated evolution of signal and receiver systems, as changes in the
signal would be matched by parallel changes in the response. The alternative “coevolution”
hypothesis suggests that signal and receiver traits are genetically independent and are
coadapted through mutual selection and adjustment of each component (Greenfield, 2002).
Reciprocal hybrid crosses of T. oceanicus and T. commodus revealed that F1 hybrids
inherit intermediate song with some temporal parameters showing greater similarity with the
maternal species (Hoy, 1974). As crickets have an XO sex determination system, reciprocal
hybrid males differ in the maternal origin of their X-chromosome (and possibly in
cytonuclear components and maternal effects). The F1 hybrid females were found to not only
respond more to hybrid songs but to prefer songs from sibling males that shared the same
maternal species over reciprocal hybrids (Hoy et al., 1977). Both song and preference showed
an intermediate pattern and some components showed a greater similarity to the maternal
species, which suggests polygenic control and some sex-linkage (with selective paternal
inactivation in females) or maternal effects (e.g. cytoplasmic factors). This behavioural
coupling of song and preference traits in F1 hybrids was suggested to indicate a common
16
genetic basis to both traits. However, as both traits are polygenic and only F1 hybrids were
examined, the pattern of intermediate song and preference is consistent with both genetic
coupling and coevolutionary hypothesises (Doherty & Hoy, 1985; Butlin & Ritchie, 1989;
Boake, 1991). Segregation of the signal-response traits through further F2 or backcross
generations and genetic approaches, such as QTL mapping, would be needed in order to
attempt to distinguish these two processes.
Although the genetic coupling hypothesis has been a topic of interest for over fifty
years, empirical evidence supporting it is weak (Butlin & Ritchie, 1989; Shaw et al., 2011).
Early studies, using hybridization of sibling species, were constrained by experimental design
and genetic tools. Recent studies on the genetic architecture of sexual communication
systems across different taxa and sensory modalities - visual communication in Heliconious
butterflies and medaka fish (Kronforst et al., 2006; Fukamachi et al., 2009), acoustic
communication in Laupala crickets and the acoustic moth, Achroia grisella (Shaw &
Lesnick, 2009; Wiley et al., 2012; Limousin et al., 2012), and chemical communication in
Drosophila (Bousquet et al., 2012) (reviewed by Shaw et al., 2011) - provide tantalizing
results that have reignited interest in the role of genetic coupling in mate recognition systems.
However, nearly all of these studies cannot distinguish between tight physical linkage and
pleiotropy. The strongest case supporting genetic coupling so far comes from work on
chemical communication, in which genes of large effect are predicted to underlie the
emission and perception of signals (Ritchie & Philips, 1998; Bousquet et al., 2012). Bousquet
et al. (2012) demonstrated how a single gene, desaturase1, can simultaneously influence both
pheromone production and perception in Drosophila melanogaster through differential tissue
expression. At this stage it is difficult to determine if the paucity of cases supporting genetic
coupling is an artefact of the difficulty of distinguishing between physical linkage and
pleiotropy or if it actually reflects the underlying biological reality (Singh & Shaw, 2012).
17
In evolutionary terms, the distinction between tight physical linkage and pleiotropy
may not be clear, as both situations have the advantage of reducing recombination and
producing robust genetic correlations between sender and receiver traits. Most models of
runaway sexual selection depend upon a strong genetic correlation between signal and
preference traits (Lande, 1981; Kirkpatrick, 1982; exception Bailey & Moore, 2012). In the
absence of physical linkage, assortative mating can generate linkage disequilibrium between
signal and receiver traits. However, this association will rapidly breakdown under random
mating (Greenfield et al., 2014), therefore physical linkage of mate recognition traits may be
particularly important to constraining gene flow and maintaining species boundaries in
sympatry (Sæther et al., 2007).
Following on from Hoy's early studies, extensive research has focused on examining
intraspecific patterns of calling song, female preferences, and, to a lesser extent, courtship
song within T. commodus and T. oceanicus. These studies have provided important insights
on the calling song elements which contribute to mate choice (Simmons et al., 2001;
Simmons, 2004; Bailey, 2008; Simmons et al., 2013) and potentially species recognition
(Hennig & Weber, 1997; Bailey et al., 2016, in review). Considerable phenotypic and genetic
variation exists for calling song and female preferences, both within and between populations
(Zuk et al., 2001; Simmons, 2004; Hunt et al., 2007; Pitchers et al., 2013). In T. commodus
numerous calling song parameters have been shown to be under stabilizing selection,
possibly corresponding to sensory tuning in females (Brooks et al., 2005), while others, such
as calling effort, seem to be under directional selection, possibly reflecting a role as an
indicator of genetic quality (Bentsen et al., 2006). However, almost all studies have focused
solely on allopatric populations of each species. Only two studies have examined calling song
among sympatric populations (Hill et al., 1972; Otte & Alexander, 1983). These studies
18
suggested current hybridization is rare (Otte & Alexander, (1983, p69) tentatively suggest
they encountered a hybrid male based on its song) and found no evidence for reinforcement
in calling song or female preferences.
A growing number of studies are highlighting that chemical communication and
cuticular hydrocarbons (CHCs) play an important role in cricket social behaviour (Rence &
Loher, 1977; Balakrishnan & Pollack, 1997; Tregenza & Wedell, 1997; Mullen et al., 2007;
Simmons et al., 2013; Thomas & Simmons, 2008, 2009, 2010, 2011). Cuticular hydrocarbons
have been found to have undergone a rapid diversification, which parallels song divergence
amongst a species radiation of Laupala crickets (Mullen et al., 2007). Furthermore, male T.
oceanicus on two different Hawaiian islands appear to have independently lost the ability to
sing (Pascoal et al., 2014), an adaptation to avoid acoustically oriented parasitoid flies (Zuk et
al., 2006), suggesting alternative mating signals or strategies are important (Bailey et al.,
2007; Thomas et al., 2011; Gray et al., 2014; Simmons et al., 2014). Females exposed to
male-derived CHCs have been shown to increase their responsiveness to calling song (Bailey,
2011). Expression of cuticular hydrocarbons can also be highly flexible in response to cues in
the social environment (Thomas et al., 2011; Pascoal et al., 2015). However, little is known
about the nature of CHCs in T. commodus and the role chemical communication may play in
mate recognition and interspecific sexual isolation.
19
Objectives
In this thesis I use behavioural and genomic approaches to examine the reproductive
barriers maintaining the species boundary between T. oceanicus and T. commodus. Although
these species are a classic study system for sexual selection, surprisingly little is known about
the processes or mechanisms that maintain species integrity across an extensive area of
species overlap. This study therefore aims to bridge an important gap in our knowledge on
the barriers to gene flow between these species and more broadly to address questions related
to how interfertile species maintain their integrity upon secondary contact and whether sex
chromosomes play an important role in reproductive isolation.
Chapter 2 - Geographic Variation in Long vs. Short Range Mating Signals
In Chapter one I examine geographic variation in long and short range mating signals
across allopatric and sympatric populations. This allows for questions about the relative roles
of species interactions and geographic isolation in signal divergence to be examined. For
example, are long range signals more likely to exhibit divergence due to species interactions
than short range signals? Few studies have examined geographic variation, among regions of
sympatry and allopatry, in more than one sexual signal or signals from different sensory
modalities. However, features of both the modality and “reach” of signals may influence their
importance in reproductive isolation.
Chapter 3 - Sexual Isolation and the Evolution of Sexually Selected Traits
In the second chapter I examine the role of behavioural barriers and sexual selection
against hybrids. In addition, I test whether prezygotic barriers are sex-linked. Behavioural
isolation among hybrids may be an important barrier to gene flow but has received relatively
little attention in speciation research. Haldane’s rule is usually restricted to cases of hybrid
20
sterility or mortality but may be extended to behavioural traits which also reduce hybrid
fitness (Davies et al., 1997).
Chapter 4 - A Rare Exception to Haldane’s Rule: Are X-chromosomes Key to Hybrid
Incompatibilities?
In the third chapter I examine the role of sex chromosomes in a remarkably rare
exception to Haldane’s rule. Deviations from Haldane’s rule might be more prevalent than
previously assumed, in particular among taxa with unconventional sex determination systems
(i.e. XO sex determination in crickets). I test whether X chromosome incompatibilities, in
particular X-X incompatibilities, contribute to female (homogametic) sterility.
Chapter 5 – Genomic Discordance: Unusual patterns of differentiation at autosomal vs. Xlinked loci
In the fourth chapter I use population genomic approaches to test the relative role of
autosomal and X-linked loci in population differentiation and species divergence. I examine
the genetic relationship of the three putative species to determine if there is contemporary
hybridization and introgression. Taken together, the results reveal an unexpected pattern that
contradicts established theory that X chromosomes play a pronounced role in population
divergence and reproductive isolation.
21
Chapter 2
Geographic Variation in Long vs. Short Range Mating Signals
Abstract
Sexual signals and mate preferences can evolve rapidly and play a key role in the
formation and maintenance of species boundaries. Behavioural barriers to gene exchange are
likely to involve multiple mate recognition signals and sensory modalities. Geographic
comparisons can elucidate how variation in mate attracting signals arises and persists within
populations and ultimately contributes to species isolation. However, few studies examine
more than one sexual signal or signals involving different sensory modalities. Long and short
range signals may be under divergent ecological and sexual selection, which may interact to
promote or constrain signal diversification and population divergence. In addition, male and
female mating signals are likely to be under different selective pressures. In this study I
examined geographic variation in male calling song and male and female cuticular
hydrocarbons (CHCs), among allopatric and sympatric populations of two sibling species of
Australian field cricket species, Teleogryllus commodus and T. oceanicus. These are classic
study species for sexual selection and are sympatric across an extensive area of eastern
Australia. However, most research has focused solely on long distance calling song in
allopatric populations, and close range signals may play an important role in species
interactions and sexual isolation.
Here I examine the relative role of interspecific interactions influencing divergence in
long and short range signals. Male calling song showed a clear bimodal distribution of song
types in sympatry, which suggests current hybridization is rare or absent. There was weak
22
support for character displacement in only one out of four sympatric populations. These
results are more consistent with calling song having diverged sufficiently to prevent
mismatings prior to secondary contact. In contrast, the species boundary was not clearly
delimited by CHCs. In both species there were clear sex differences in CHC profiles. There
was a remarkably high level of CHC variation observed, specifically among females in
sympatry and also in the northern region, which overlaps with a third putative species, T.
marini, raising the possibility that species interactions may have contributed to CHC
diversity. The two signal traits (CHCs and song) traits differed in their relationship with
geographic distance. As predicted, geographic distance showed a stronger relationship with
divergence in CHCs than with song, in line with CHCs playing a role in environmental
adaptation. Divergent sexual and ecological selection are likely to contribute to the
discordance in trait divergence between both species and sexes.
Introduction
The incredible wealth of animal diversity is reflected in the spectacular range of
signals animals use to communicate with one another: visual, aural and olfactory. This is
particularly evident in mate recognition systems (MRS) (Paterson, 1985). MRS are composed
of sexual signals (e.g. calling song, pheromones, displays) that are processed by a receiver
resulting in mating (Greenfield, 2002). Mating signals can be very diverse across otherwise
cryptic species lineages, which has long raised debate as to the role of sexually selected traits
in speciation (Darwin, 1871; West-Eberhard, 1983, 1984; Gleason & Ritchie, 1998; Wells &
Henry, 1998; Panhuis et al., 2001; Mendelson & Shaw, 2005; Ritchie, 2007; Svensson &
Gosden, 2007). Understanding how intra- and interspecific variability in mate recognition
23
systems arises and is maintained across a species range is important in determining how
behavioural barriers contribute to interspecific sexual isolation.
Divergence in MRS may occur across a species range due to varying selective
pressures and stochastic processes (West-Eberhard, 1983; Endler, 1992; Foster & Endler,
1999; Blows, 2002; Wilkins et al., 2012). Rapid signal divergence amongst populations may
occur through a runaway Fisher process, due to a genetic correlation between signal and
receiver traits (Lande, 1981, 1982; Kirkpatrick, 1982). West-Eberhard (1983) highlighted
how social interactions between members of the same species could also drive rapid
evolution of socially-selected characters among different populations. These processes could
cause allopatric populations to diverge in different directions due to the arbitrary nature of the
factors involved. However, a contrasting prediction could be made that MRS may be
constrained and stable across populations due to the coevolutionary dynamics of signalreceiver systems (Paterson, 1985), with female preferences imposing strong stabilising
selection on signal traits (Brooks et al., 2005; Bentsen et al, 2006). In addition, divergent
ecological selection may result in predictable patterns of MRS divergence amongst
populations, corresponding to environmental factors (Boughman, 2001; Seehausen et al.,
2008).
If the MRS of sibling species have diverged sufficiently before secondary contact, it
may facilitate strong assortative mating and interspecific sexual isolation. Species
interactions may promote or oppose divergence in signal-receiver systems (Haavie et al.,
2004; Tobias et al., 2014; Greig et al., 2015). If female preferences are open-ended (e.g.
females prefer songs with higher frequencies) such interactions can lead to heterospecific
mate choice and asymmetrical introgression in areas of species overlap (Stein & Uy, 2006;
Baldassarre et al., 2014). Hybridization and introgression may result in convergence of
elements of both species communication systems leading to a continuum of species
24
boundaries (The Heliconius Genome Consortium et al., 2012). Alternatively, increased
fitness costs due to signal jamming or costly mismatings with sibling species may promote
divergence of signal-receiver systems in sympatry to maximize dissimilarity through
reinforcement or reproductive character displacement (Servedio & Noor, 2003; Coyne & Orr,
2004). Howard (1993) distinguishes between these two phenomena and here I treat
reinforcement in a “broad sense”, as a process of selection against costly interspecific
matings which can subsequently result in the pattern of reproductive character displacement.
Reproductive character displacement can manifest due to processes other than reinforcement,
and reinforcement can have consequences other than reproductive character displacement
(Noor, 1999). Most convincing examples of reproductive character displacement come from
studies on insects and frogs (reviewed in Butlin, 1987, 1989; Howard, 1993; Noor, 1999;
Servedio & Noor, 2003; Servedio, 2004; Gerhardt, 2013).
The relative importance of intraspecific and interspecific processes in driving
evolutionary change in MRS remains an area of considerable debate. Courtship behaviour is
generally a multi-modal process involving numerous sexual signals and sensory modalities
which operate over different spatiotemporal scales (Rowe, 1999; Candolin, 2008; Leonard &
Hedrick, 2009). Long range and short range signals have often been suggested to serve
different functions, namely species recognition and mate choice (Pfennig, 1998; Gray, 2005;
Zuk et al., 2008). The distinction between these functions may be misleading and instead
both processes potentially reflect a continuum of mate choice (Ryan & Rand, 1993;
Mendelson & Shaw, 2012, 2013; Padian & Horner, 2013). The same signals which mediate
intraspecific mate choice in many cases have been found to contribute to species level
assortative mating (Svensson et al., 2007; Higgie & Blows, 2007; Svensson et al., 2016).
However, the form and strength of selection acting on long vs short range signals is likely to
25
be different. Earlier acting courtship behaviours are expected to be under stronger selection
pressures to enhance assortative mating as they have greater potential to influence gene flow.
However, relatively few studies have examined geographic variation across allopatry and
sympatry, in both long and short range signals, especially from different sensory modalities
(Tregenza et al., 2000). Areas where closely related species overlap geographically (hybrid or
contact zones) provide an important opportunity to examine the relative importance of long
vs. short range signals in mediating species interactions. If long range signals
disproportionately contribute to species isolation, one may predict a greater divergence of
these traits in sympatry compared to short range signals. In addition, sexual signals (or cues)
that are under both divergent ecological and sexual selection may exhibit enhanced
population divergence.
Study System
The two closely related field cricket species Teleogryllus commodus and T. oceanicus
provide an excellent model system to examine (the causes & consequences of) divergence in
MRS, as their acoustic behaviour has been well characterized (Hoy, 1974; Casaday & Hoy,
1977; Hoy et al., 1977; Doherty & Hoy, 1985; Hennig & Weber, 1997). The species overlap
across an extensive area (~400km) in north-eastern Australia (~23.37ᵒS – 26.38ᵒS latitude;
Hill et al., 1972) in what is believed to be an area of secondary contact (Otte & Alexander,
1983). Both species are ecological generalists and are often found in very close proximity
with no obvious differences in resource use or habitat preferences (Otte & Alexander, 1983).
They readily hybridize in the laboratory and although F1 hybrid females are sterile, hybrid
males are fertile and capable of mating with females of the parental species, suggesting
hybridization and reinforcement may occur in sympatry (Hogan, 1971; Hogan & Fontana,
1973; Moran et al., in review). However little is known about species interactions in the area
26
of overlap as most research has focused almost exclusively on allopatric populations
(Simmons et al., 2001; Zuk et al., 2001; Tinghitella et al., 2011; Pitchers et al., 2013; Bailey
& Macleod, 2013).
The long-range, male calling song is the most distinguishable feature known to
differentiate the species and has long been believed to play a critical role in maintaining the
species boundary (Hill et al., 1972; Bailey & Macleod, 2013). However, only two previous
studies have examined calling song in both allopatric and sympatric populations (Hill et al.,
1972; Otte & Alexander, 1983). Hill et al., (1972) explicitly tested for reproductive character
displacement in male calling song and female preference and found no evidence for it.
However, they examined calling song components separately, while acoustic signals are
complex and multivariate in nature, therefore a multivariate statistical approach may be more
appropriate (Higgins & Waugaman, 2004). In other cricket species, such as Gryllus fultoni
and G. vernalis, calling song parameters have been found to exhibit character displacement,
which suggests song may play an important role in mediating species interactions
(Hondasumi, 2005; reviewed in Gerhardt, 2013). Calling song has been studied intensively
for its role in intraspecific mate choice and may encode information on male quality
(Simmons & Ritchie, 1996; Drayton et al., 2007; Drayton et al., 2010). Considerable genetic
variation in both calling song and female preferences has been found within and between
populations of these Teleogryllus species, which could facilitate song diversification across
the species range (Simmons et al., 2001; Simmons, 2004; Hunt et al., 2007).
Courtship behaviour is a multimodal process in these species and short range signals
may play an important role in mate preferences and species isolation. Upon physical contact,
individuals use their antennae to detect cuticular hydrocarbons (CHCs) which are involved in
27
sex recognition (Tregenza & Wedell, 1997) and mate recognition (Howard & Blomquist,
2005). CHCs are waxy molecules secreted on the cuticle of male and female insects which
are believed to have evolved for desiccation resistance but have been co-opted for social
communication (Balakrishnan & Pollack, 1997; Tregenza & Wedell, 1997; Smadja & Butlin,
2009). Variation in CHCs is often species-specific and likely to play an important role in
interspecific sexual isolation (Ferveur, 2005; Smadja & Butlin, 2009). CHC profiles in T.
oceanicus are known to be heritable (Thomas & Simmons, 2008a), sexually dimorphic
(Thomas & Simmons, 2008b) and under sexual selection (Thomas & Simmons, 2009)
through both male and female choice (Thomas & Simmons, 2010), potentially providing
information about a partner’s genetic compatibility (Thomas & Simmons, 2011). However,
little is known about the nature of CHCs in T. commodus and the role chemical
communication may play in mate recognition and reproductive isolation. Behavioural barriers
are likely to be important in maintaining the species boundary and as courtship involves
multiple mate recognition signals it is important to consider the joint role that song and CHCs
may play in species isolation.
Aims
The aims of this study were threefold:
1) First, I tested whether calling song clearly delimits the species boundary. If there
was current species admixture some sympatric individuals should exhibit intermediate
patterns of calling song, as we know F1 hybrid song is intermediate (Hoy & Paul, 1973;
Chapter 3). Alternatively, if the species were strongly or completely reproductively isolated
we would expect a bimodal distribution of song types corresponding to the pure parental
species.
28
2) Second, I tested the hypothesis that species interactions have resulted in
reproductive character displacement in calling song and/or CHCs in the area of overlap. I
predicted that if calling song or CHCs play a role in species isolation, sympatric individuals
would exhibit accentuated divergence in these signal traits compared to those from allopatry,
particularly in song components known to be important in female phonotaxis (Hoy & Paul,
1973; Hennig & Weber, 1997; Simmons, 2004). In addition, I predicted calling song would
be more susceptible to reproductive character displacement than CHCs, due to its earlier
acting role in courtship behaviour and a greater propensity for signal interference.
3) Third, I compared the patterns of geographic variation in song and CHCs among
populations within each species. The pattern of population divergence provides an insight
into the processes that contribute to variability of sexual signals. Clear geographic structuring
of populations, with spatial distance predicting patterns of signal variation, would support
genetic drift (i.e. isolation by distance) or adaptation to an environmental gradient (isolation
by environment). Strong divergence in song or CHCs among populations, with components
varying in arbitrary directions between populations would be consistent with some form of
sexual selection (Lande, 1981, 1982). Low levels of signal variation within and between
populations would be consistent with stabilizing selection or other selective constraints
preventing diversification of song or CHCs across the species range.
I predicted CHCs would show a stronger relationship with geographic distance,
reflecting adaptive divergence to a temperature gradient along the eastern coast of Australia.
In addition, as CHCs are likely to be under both ecological and sexual selection, due to their
dual role in desiccation resistance and sexual signalling, they may exhibit greater levels of
population divergence than song. A contrasting prediction could also be made that their dual
role may constrain their diversification.
29
Material & Methods
Study Populations
Adult male and female crickets were collected from 16 sites, from March 1st – April
2nd 2013, encompassing an extensive latitudinal transect (~2,500km across eastern Australia),
including both allopatric and sympatric populations (Figure 2-1). Areas of sympatry between
the two species were located based on published studies (Hill et al., 1972; D. Otte &
Alexander, 1983). As I was interested in natural mating dynamics, namely if the species
hybridize, females captured from the area of overlap were kept in sex-segregated boxes to
ensure offspring identity reflected natural matings prior to capture, rather than hybridization
during confinement in boxes.
Figure 2-1 Distribution of the sixteen field sites in eastern Australia
The orange squares indicate allopatric populations of T. oceanicus, the green diamonds
represent sympatric populations of both species, and the blue circles indicate allopatric
populations of T. commodus. The red dot in the north indicates the site where the third
putative species T. marini was sampled (Otte & Alexander, 1983). Oscillograms of male
calling songs from each species are shown.
30
Song Recording and Analysis
Male calling song recordings were taken in both the field and in the laboratory,
encompassing the same populations, using a Sennheiser ME66 microphone and an Olympus
LS-10 handheld recorder. Laboratory recordings were made using the F1 generation reared in
a common garden environment. Stock crickets were housed in 16-L plastic boxes of ca. 80
individuals in a 25 ⁰C temperature-controlled room on a 12:12 light:dark cycle. Song
recordings were conducted in a temperature controlled room (24.3⁰C sd ±0.686), dimly
illuminated by red light, with males ca. 10 -20 days past adult eclosion isolated in individual
small (118ml) plastic cups. Approximately fifteen males were recorded from each population.
A minimum of one minute of song was recorded per individual and for the field recordings,
ten song phrases were examined per individual, while for the laboratory recordings five song
phrases were examined. The individual’s average was used in further analysis.
A total of 18 call parameters, encompassing both spectral and temporal components,
were quantified per male and all songs were analysed using Sony Sound Forge (V.7.0). To
compare variation in call characteristics within and between species, I focused on 13
parameters, which are shared between both species and I assume are homologous (Figure
2-2). Overall the calling song analysis consisted of two datasets encompassing the 16
populations: field recordings (n = 205) reflecting variation among natural populations, and
lab recordings (n = 258). The lab recordings aimed to capture inherent population differences
by controlling for environmental sources of variation. As the lab individuals were reared in a
common garden environment, population differences in song or CHCs are likely to reflect
genetic differences among populations. Although as the individuals were the F1 generation I
cannot rule out maternal effects. Body size and temperature are known to influence song
traits in insects and frogs (Gerhardt & Huber, 2002), therefore temperature and body size
effects (weight and pronotum length (PL) on song parameters were examined using analysis
31
of covariance tests (ANCOVAs) and the strength of the correlation determined using
Pearsons correlation coefficient (r) (Table S1 & S2). Temperature differences between
recordings were controlled by using temperature-corrected residuals. The mean and
coefficients of variation for all song parameters were estimated across each population (Table
S3; S4). All statistical analysis was conducted using R (version 3.0.2).
Figure 2-2 Calling song oscillograms for both species.
Labels indicate the main song traits measured. Additional song elements measured but not
labelled included: number of pulses in the chirp, trills, and average pulses per trill section.
Illustration adapted from Bailey & Macleod (2013). “IPI” refers to inter-pulse interval and
“PL” indicates pulse length.
Cuticular Hydrocarbons
CHCs were extracted from males and females ca. 15-30 days after adult eclosion.
Individual crickets were isolated in small (118 ml) plastic containers for a minimum of 5
days, provisioned with Burgess Excel “Junior and Dwarf” rabbit food and water, before being
anesthetized by chilling and then placed in small glass extract vials (5ml) and stored at -20°C.
Prior to CHC extractions, cricket samples were thawed at room temperature for ca. 10
minutes. 4 ml of hexane was then pipetted into each vial and left for 5 minutes before
32
removing the cricket with clean forceps. 100 µl of extract was pipetted into 0.3ml fixed insert
vials (Chromacol LTD, Item # 11573680) and left to evaporate under a fume hood for ca.
twelve hours. The CHC residue samples were reconstituted using 100 µl of hexane
containing 10 ng µL-1 dodecane as an internal standard and analysed via GC-MS (Agilent
7890A gas chromatograph coupled to a 5975B mass spectrometer) equipped with a HP-5 MS
capillary column (30 m x 0.25 mm ID x 0.25 µm; Agilent J&W). A 2 µL injection of the
sample was injected into an inlet (4mm ID splitless inlet liner (Agilent 5062-3587)) held at
250°C in splitless mode for 1 min. The helium carrier gas flow was 1 ml min-1. The initial
oven temperature was held at 50 °C for 1 min, then ramped at a rate of 20 °Cmin-1 to 250 °C
followed by a 4 °Cmin-1 ramp to 320 °C and a 5 min hold at this temperature. Ionisation was
achieved by electron ionisation (EI) at 70 eV. The quadrupole mass spectrometer was set to
3.2 scans s-1, ranging from 40 to 500 Da.
The abundance of each CHC peak was estimated using MSD CHEMSTATION
software (version E.02.00.493; Agilent Technologies) by measuring the area under each peak
in the chromatograms. As T. commodus and T. oceanicus CHC profiles are highly divergent
with few shared peaks (Figure 2-3) two species-specific methods were used (developed by
Melia Burdon at the University of Exeter). Briefly, in each species, diagnostic ions were
identified for each chemical peak allowing for the relative abundance of peaks across samples
to be compared. The CHC peaks were measured blind to the individual’s population identity.
I assigned individuals to either of the species group based on an initial qualitative assessment
of their chromatograms and analysed them using the corresponding species-specific method.
Some peaks, in particular among T. commodus samples (peaks: 2, 3, 10, 25, 26), were found
to be highly variable and noisy and were therefore dropped from the analysis. In total 35
peaks were used in the analysis for T. commodus samples, and 20 peaks for T. oceanicus
samples. Prior to CHC analysis the data were standardized by dividing the abundance of each
33
peak by the internal standard (10 ng µL-1 dodecane) and normalized by a log10
transformation.
,
Figure 2-3 Typical gas chromatography profile of the two species
The X-axis shows the retention time (in minutes) and the Y-axis the peak abundance. The
CHC peaks were numbered in the order corresponding to the retention time, with the first
peak being the internal standard (10 ng µL-1 dodecane). There was a higher number of peaks
among T. commodus individuals compared to T. oceanicus. Moreover, the T. oceanicus peaks
generally had a higher retention time, which suggests longer-chained CHC components.
34
Statistical Analysis
Mating signals: Delimiting the species boundary
In the first part of the study, to test whether the species boundary is clearly delimited,
I focused on calling song data (both field and lab recordings), as most of the song units are
shared between the species allowing for a combined analysis of both species. In contrast,
cuticular hydrocarbon composition is highly divergent between the species with few shared
peaks (Figure 2-3), making a combined analysis of both species CHC profiles difficult. Thus,
I used the song data to examine both within and between species patterns of divergence,
whereas for the CHCs I focused on intraspecific patterns of population differentiation and sex
differences.
Due to the complexity of the signal traits, containing a large number of song elements
and CHC peaks which are likely to be highly correlated, I used multivariate statistical
methods to reduce the dimensionality of the data. To test the degree of signal divergence
between the species, which might inform about the current levels of hybridization, I used
principal component analysis on the field and lab calling song data. PCA was implemented
using the R package FactoMineR on the correlational matrix (Lê et al., 2008). Principal
components with eigenvalues >1 were retained (n = 4, for both field and lab datasets).
Species differences were tested using the individual PC scores in a MANOVA, with species
group used as a factor. To visualize the pattern of interspecific song variation the PC scores
were plotted for PC1 vs. PC2.
Testing Reproductive character displacement
To test for reproductive character displacement and patterns of population
differentiation, the calling song lab data and CHC data were split into two species-specific
35
subsets and analysed using principal component analysis (PCA) and discriminant analysis
(DA). For the calling song data sympatric individuals were classified into either parental
species group, informed by the results from the initial PCA on the total dataset. For the CHC
data individuals were assigned a priori to either species group based on visual inspection of
their chromatograms and analysed following the corresponding species-specific method. To
test for population differentiation and sex differences for song and CHC components,
principal components with eigenvalues >1 were retained, the individual scores extracted and
a multivariate analysis of variance (MANOVA) conducted with population, sex, and their
interaction.
If song or CHCs play an important role in mediating species interactions, sympatric
populations should show accentuated signal divergence compared to allopatric populations
(to reduce costly mismatings or signal interference). As I was specifically interested in
discriminating between allopatric and sympatric populations, discriminant function analysis
on the PCs (DAPC) was applied with population as the prior grouping factor, using the R
package adegenet (Jombart et al., 2010). This method was originally designed for genetic
markers but the reason I used it here was in order to integrate the results with genetic data at a
later date. DAPC performs a PCA on the initial data and then identifies a linear combination
of the variables that maximizes between-group variance while minimizing within-group
variance (i.e. discriminant analysis). Prior to analysis the data were standardized to have a
mean of 0 and a standard deviation of 1. Group membership was assigned based on each
individual’s population identity. One individual from the “HB” population was removed from
the song analysis as it had an extremely high frequency (5.9 kHz). The results were
qualitatively unchanged by the removal. The optimal number of PCs to retain is an important
consideration to ensure a fair trade-off between the power to discriminate and not overfitting. I used the cross-validation method in adegenet (using the xvalDapc function) to
36
identify the correct number of PCs to retain. This procedure performs DAPC on a subset of
the data (in our case 90% of observations from each population) and then finds the optimal
number of PCs to retain that maximizes the prediction success of assigning the remaining
10% of individuals to their correct group (i.e. population). Similar to the analysis for the PCs,
discriminant function scores were used to test the degree of population differentiation, using
MANOVAs, followed by ANOVAs and Tukey pairwise population comparisons on the
significant components using the R package multcomp (Westfall, 2008).
Geographic distance and signal divergence
As our sampling transect represents a clear latitudinal gradient, geographic distance
may accurately reflect colonization routes and patterns of population connectivity. To test if
spatial distance could predict variation in signal traits (e.g. due to isolation by distance),
Mantel tests based on the discriminant function coordinates were implemented in R using the
ade4 package (Dray & Dufour, 2007). Mantel tests are non-parametric tests that measure the
correlation between two matrices and control for the non-independence of populations that
are used in multiple comparisons. The song and CHC distance matrices were calculated based
on the mean population discriminant function coordinates for the first three axis, which were
then regressed against the geographic distance matrix, estimated from each site’s latitude and
longitude coordinates obtained from Google maps. The number of permutations of the
Mantel test was 10,000. To test whether song or CHCs differ in their relationships with
geographic distance I compared the slopes and intercepts by bootstrapping using the R
package boot (Canty & Ripley, 2008). I estimated and compared the frequency distribution of
1,000 slopes and intercepts for both traits (following the methods of Baselga, 2010). The
probability of one trait having higher parameters than the other was estimated by the
37
proportion of the 1,000 bootstraps for which the trait had higher parameters than the other.
For p-values I used the probability of obtaining the opposite result by chance.
Results
Males of both species were found calling in very close proximity across the contact
zone. The field recordings revealed a mixture of both species in sympatry, a small number of
males clustered with T. commodus and a larger number with T. oceanicus (Figure 2-4.A).
Strikingly, in the lab recordings all males from sympatric populations clustered among the T.
oceanicus call type (Figure 2-4.A). Interestingly, the field data suggested the presence of a
third putative species, T. marini (Fig. S1) (distinguished primarily by a low carrier frequency
(ca. 3 kHz) and triplet pulses in a proportionally longer trill; Figure 2-1- oscillograms) in the
north-eastern region around Daintree (DV & KH) (population labelled red in Figure 2-1).
There is only one previous description of this putative species, identified in a similar coastal
region (Otte & Alexander, 1983, p69). These T. marini individuals were removed to carry out
the formal analysis as the number of recordings were small (n=7) and the focus was on the
two principal species. Nevertheless, it is important to recognise that our population transect
covers at least two areas of sympatry, providing the possibility for interspecific effects on
mating signals in more than one contact zone.
Calling song clearly delimits the species boundary
Plotting the individual scores on the first two PC axes, for both the field and the lab
datasets, revealed a strikingly similar pattern; two distinct species-specific song clusters
(Figure 2-4). Principal component loadings were broadly similar for both song datasets
38
(loadings for both field and lab PCA in Table S5), so here the focus is on the lab data. The
first four principal components had eigenvalues >1 and cumulatively accounted for ~85% of
the variation. Both species were distinguished primarily on PC1 (ANOVA: F 1,256 = 2803.8, p
= <0.001) and marginally on PC2 (F 1,256 = 3.66, p = 0.057) (Figure 2-4), but not on any of the
other axes (all p values > 0.05). The loadings from PC1 highlight the main species
differences; T. oceanicus song consists of higher carrier frequency, greater number of trills
and longer duration of the chirp trill interval than T. commodus song (Figure 2-2; Table S5).
The highest loadings on the second component (PC2) are total duration and trill duration,
which are highly correlated (t = 92.663 df = 257, r2 = 0.982, p < 0.001). The third component
(PC3) is dominated by interpulse intervals, while the fourth component (PC4) contrasts trill
total duration with chirp duration (Table S5.).
Overall the species calling song was clearly differentiated in both the field and lab
datasets. The absence of intermediate individuals suggests contemporary hybridization is rare
or absent, as we know F1 hybrid songs are intermediate (Hoy, 1974). Whilst there were two
intermediate males on PC1 in the field data (Figure 2-4), these individuals were from a T.
oceanicus allopatric population (KH), a long distance from the primary contact zone.
Therefore, these intermediate male songs are not likely to reflect hybridization between the
species in question. The two divergent male songs from KH could reflect species interactions
with the third putative species T. marini, which co-occurs in the DV and KH area. However,
the high environmental variance (i.e. temperature differences) captured in the field recordings
could also have contributed to this pattern.
39
Figure 2-4 PCA for calling song: field and lab recordings
A) PCA scores (PC1 vs. PC2) of calling song from both field (n = 205) and laboratory
(n=258) recordings. Individuals are labelled based on species identity (T.com = T.
commodus; Mix = potential mixed species identity; T.oc =T. oceanicus) and geographical
grouping (allo= allopatric, sym= sympatric) (Colours and shapes correspond to Figure 2-1).
Confidence ellipses delimitate the 0.95 confidence interval. The two species are clearly
distinguished by PC1 in both the field and the lab datasets. In the field a small number of
individuals from sympatric populations clustered with T. commodus and a larger number with
T. oceanicus while in the lab no T. commodus like individuals were identified from sympatric
populations. B) Diagram illustrating the top five contributing factors to the first and second
principal component for both field and lab recordings.
40
Female-biased variability in cuticular hydrocarbon composition
In sharp contrast to the song data, the species boundary was not clearly delimited by
cuticular composition. A small number of individuals that originated from sympatry (n = 27)
and, more surprisingly, from the northern T. oceanicus allopatric populations (n = 18) were
analysed using the T. commodus CHC method, as they contained a higher number of peaks
than the standard T. oceanicus CHC profile (Figure 2-5, S2). It is important to recall that
sample chromatograms were assigned, blind to the samples population of origin, to either of
two predefined species methods based on a qualitative assessment of the number and
distribution of CHC peaks (the T. commodus CHC profile has a much larger number of peaks
with a lower retention time than the T. oceanicus profile (Fig. 2-3)).Strikingly, the 45
individuals in question were almost all females, with the exception of one male from the TS
population. Moreover, these samples from sympatry and T. oceanicus allopatry were clearly
differentiated in their CHC profiles compared to the standard allopatric T. commodus profiles
(Figure 2-5, S2, S3; MANOVA, examining females only for PCs with eigenvalues > 1 (n = 6)
and using geographic location as a factor (i.e. T. commodus allopatry vs. sympatry/T.
oceanicus allopatry): Wilks λ=0.291, F1,175 = 69.19 , P < 0.0001). Interestingly, no individuals
with a T. oceanicus like CHC profile were identified beyond the contact zone with T.
commodus.
41
Figure 2-5 PCA and DAPC for T. commodus CHCs - (incls strange samples)
A) Principal component scores (PC1 vs. PC2) of CHCs from males (blue) and females (pink),
with T. commodus like CHC profiles, across sixteen populations. Populations are arranged in
geographic order from AM in the south to KH in the north. The geographical regions of
interest are underlined with colours corresponding to Figure 2-1(T. commodus allopatry =
blue, Sympatry = green, T. oceanicus allopatry = orange). Sample sizes (n) for both females
and males are shown. B) Ordination plot of discriminant functions (LD1 vs. LD2) obtained
from DAPC on females only. Colour reflects geographic region as before. Clusters were
defined based on a priori population identity and the 0.95 confidence ellipse are shown. The
inset indicates the relative contribution of the discriminant functions in differentiating
populations. Discriminant function loadings are provided in Table S6 and chromatograms in
Fig. S2.
To ensure this unexpected pattern was not due to a sample mix up, GCMS was rerun
on stock CHC extracts from the 45 divergent samples and the results were qualitatively the
same. Although I cannot completely rule out a sample mix up or contamination, I do not
believe this is the case, as the samples in question are randomly dispersed across the dataset
in regards to the dates of extraction and analysis. In addition the CHC peaks are very
42
different between the geographic regions. A previous study also observed aberrant CHC
profiles (multiple peaks) in females from the Daintree region (S. Pascoal, personal
communication). As I cannot ascertain the species identity of the divergent individuals (in
particular those from sympatry) and little is known about the variability of CHCs in either
species, it is difficult to discern the basis of this unexpected pattern. To examine geographic
variation in CHCs I excluded these samples from the rest of the analysis, as otherwise the
analysis would be dominated by these differences. However, I elaborate on the significance
of this unexpected pattern in the discussion.
T. oceanicus song
To test whether species interactions promoted calling song divergence, I focused on
comparing discriminant functions for allopatric and sympatric populations of T. oceanicus,
due to the apparent absence of T. commodus males in the lab. The first song discriminant
component (LD1) explained 47.91% of the between population variation and was composed
predominantly of frequency and trill components (Table 2-1). The second axis (LD2)
explained 25.63% of the between population variation and was dominated by chirp and
structural elements. A multivariate analysis of the first three discriminant axes (LD1 – LD3)
revealed significant population differences in calling song (MANOVA: Wilks λ=0.55, F7,122 =
3.804, P < 0.001), primarily driven by the first two axes (LD1 – LD2).
43
Table 2-1. T. oceanicus calling song loadings (DAPC)
T. oceanicus song loadings for three discriminant functions (DAPC based on eight retained
PCs). Loadings in bold indicate main contributing factors.
Song element
Chirp.Duration
Chirp.Number
Total.Duration
Frequency
Trill.Total.Duration
Chirp.Trill.Interval
Chirp.Pulse.Length
Chirp.Interpulse.Interval
Trill.Pulse.Duration
Trill.Interpulse.Interval
Trill.1.2 interval
Intersong.Interval
Number of trills
Average.Pulses.per.Trill
% Variation explained
LD1
LD2
LD3
0.078
0.227
0.016
0.715
-0.002
0.202
-0.150
-0.146
-0.452
-0.542
0.119
-0.066
0.070
-0.142
47.91
-0.291
-0.418
-0.037
0.173
0.005
-0.046
0.183
0.237
0.146
-0.200
0.478
0.686
-0.232
0.258
25.63
0.222
0.312
-0.152
-0.557
-0.192
-0.065
-0.154
-0.067
-0.222
-0.277
0.242
0.296
-0.275
0.160
11.59
Broad comparisons between allopatric and sympatric populations revealed no
significant song differences between the two groups (MANOVA, using the geographic origin
of individuals (i.e. allopatric or sympatric) as a factor: Wilks λ=0.973, F1,128 = 1.173, P
=0.323). However, as our transect encompasses a large area and at least two contact zones
(KH & DV populations overlap with the third putative sibling species T. marini), not all
sympatric populations may be expected to exhibit character displacement or to have
responded in the same direction. To examine in more detail the effect of sympatry on song,
pairwise population comparisons (Tukey) were used.
44
Figure 2-6. T. oceanicus song population differentiation (DAPC)
Discriminant analysis of principal components (DAPC) results for song among T. oceanicus
populations. Boxplots of individuals DF coordinates for LD1-LD2 grouped by population and
arranged in geographic order from HB in the south to KH in the north. KH and DV were
sampled only a couple of hundred meters from each other and although allopatric in respect
to T. commodus, are sympatric with the third putative species T. marini.
All significant population comparisons for LD1 (8 out of 28 comparisons) involved
sympatric populations (Figure 2-7). Interestingly, the population at the leading edge of the
contact zone HB was the most highly differentiated, especially in comparison with the three
other sympatric populations (TS, RH, YP – p < 0.01) (Table 2-2). However, the sympatric
populations appear to have diverged in different directions. HB individuals had a higher
frequency and larger intervals than allopatric populations, while RH and YP individuals
showed a reduction in these song elements (Table 2-1). Interestingly, the lowering of calling
song frequency brings it closer into the range of that with T. commodus. For LD2, the only
population to show strong differentiation was TS, also from sympatry, however these
45
individuals exhibited shorter intersong intervals, making them more similar to T. commodus
(Table 2-1). No pairwise population comparisons were significant for LD3 (all p > 0.05).
Table 2-2. T. oceanicus: Pairwise population comparisons for song discriminant
functions (LDs)
Only significant comparisons are shown (P < 0.05).
Variable Population Comparison Estimate Std. Error t value Pr(>|t|)
LD1
LD2
HB-JC
-1.513
0.356
-4.246
0.001
HB-KH
-1.157
0.366
-3.161
0.04
HB-PL
-1.183
0.378
-3.13
0.044
HB-RH
-2.186
0.366
-5.972
<0.001
HB-TS
-1.455
0.361
-4.032
0.002
HB-YP
-1.921
0.361
-5.323
<0.001
DV-YP
-1.099
0.338
-3.249
0.031
DV-RH
-1.363
0.344
-3.968
0.003
TS-DV
-1.39
0.338
-4.11
0.002
TS-HB
-1.234
0.361
-3.42
0.018
TS-JC
-1.173
0.338
-3.468
0.016
TS-KH
-1.151
0.348
-3.305
0.027
TS-RH
-1.55
0.348
-4.45
0.001
YP-TS
1.208
0.343
3.522
0.014
T. oceanicus CHCs
The relative abundance of CHC peaks differed significantly among T. oceanicus
populations and between the sexes (MANOVA on the principal components scores (PCs):
using population as factor, Wilks λ=0.786, F7,341 = 2.999 , P < 0.0001; using sex as a factor,
Wilks λ=0.341, F1,341 = 163.179 , P < 0.0001; the interaction between population and sex,
Wilks λ=0.872, F7,341 = 1.684, P = 0.015). Discriminant function analysis using population as
46
the prior grouping factor reduced the overall sex differences (MANOVA, using sex as a
factor: Wilks λ=0.993, F1,341 = 0.779 , P = 0.507). This suggests that the factors that
differentiate the sexes are different from those that distinguish the populations as discriminant
function analysis maximizes between group variance while minimizing within group
variance. Although the difference between the sexes was reduced for the discriminant
functions, the interaction between population and sex was still significant (MANOVA,
Pop*Sex: Wilks λ=0.893, F7,341 = 0.893, P = 0.011). In line with the previous results,
examining the sexes separately revealed considerable differences in CHC peaks among the
populations and sexes (Figure 2-5).
There were differences between allopatric and sympatric T. oceanicus populations in
the proportion of CHC peaks for all three discriminant functions (MANOVA, using the
geographic region of origin of individuals (i.e. allopatric or sympatric) as a factor: Wilks λ =
0.857, F1,355 = 19.58 , P < 0.0001). There was no interaction between sex and the geographic
region (MANOVA, using geographic region*sex as a factor: Wilks λ=0.996, F1,353 = 0.447, P
= 0.72). Examining males and females separately indicated that the CHC profiles of both
sexes were differentiated between allopatry and sympatry (MANOVA, using geographic
origin (allopatry or sympatry) as a factor for females: Wilks λ = 0.809, F1,130 = 7.47 , P <
0.001; and for males: Wilks λ = 0.68, F1,223 = 25.85 , P < 0.001). Contrary to the prediction
that sympatric populations would exhibit consistent CHC differences compared to those in
allopatry, there was considerable variation among populations and between the sexes in the
extent and direction of differentiation (Figure 2-8). Most significant pairwise population
comparisons involved sympatric populations (P < 0.05, Table S7). However, certain
allopatric populations (i.e. PL and JC) were also highly differentiated (Figure 2-8). In
addition, the populations DV and KH at the northern edge of the transect, which overlap with
the third putative species T. marini, exhibited considerable differences compared to the other
47
populations (Figure 2-8). The significant population comparisons are shown in Table S7 and
the loadings for males and females in Table S8.
Figure 2-8. T. oceanicus CHC population differentiation (DAPC)
Male and female scores for LD1-LD3 grouped by population and arranged in geographic
order from HB in the south to KH in the north. The main geographic regions of comparison
are underlined; sympatry in green and allopatry in orange.
48
T. commodus song
There were differences among the allopatric T. commodus populations in their calling
song (MANOVA on the first three discriminant function scores and using population as
factor: Wilks λ=0.536, F7,116 =3.792, P < 0.001). The first discriminant axis (LD1) explained
41.05% of the variation between populations (ANOVA: F7,116 =5.782, P < 0.001) and was
composed mainly of frequency and trill elements, in particular the number of trills and the
number of pulses in the trill (Table 2-3). The second axis (LD2) explained 22.68% of the
variation (ANOVA: F7,116 =3.194, P = 0.004) and was composed mainly of chirp elements.
The third axis explained 18.86% of the variation (ANOVA: F7,116 =2.656, P = 0.014) and was
dominated by a negative correlation between intersong interval and chirp elements.
Table 2-3. T. commodus calling song loadings
Three discriminant functions based on seven retained PCs. Loadings in bold indicate main
contributing factors.
Song element
Chirp.Duration
Chirp.Number
Total.Duration
Frequency
Trill.Total.Duration
Chirp.Trill.Interval
Chirp.Pulse.Length
Chirp.Interpulse.Interval
Trill.Pulse.Duration
Trill.Interpulse.Interval
Intersong.Interval
Number of trills
Average.Pulses.per.Trill
% Variation explained
LD1
LD2
LD3
0.186
0.076
-0.009
0.510
-0.046
0.483
0.341
-0.139
0.211
-0.084
0.070
-0.554
0.473
41.05
0.312
0.282
0.167
0.220
0.119
0.191
-0.067
0.143
-0.159
0.185
0.239
-0.039
0.113
22.68
-0.022
-0.110
0.301
-0.019
0.310
0.334
0.233
0.049
0.299
-0.124
-0.365
0.369
-0.120
18.86
49
Tukey pairwise comparisons revealed that for LD1 the most differentiated populations
were the southern populations AM and BN, which were also highly differentiated from each
other (Figure 2-9; Table 2-4). For LD2 the only significant population differences involved
the BN population (BN – CC: P = 0.012; BN – SV: P = 0.036) while for LD3 only AM vs.
CH was marginally significant (P = 0.047). There was no evidence that the northern
populations (e.g. UQ or SV), closer to the contact zone with T. oceanicus, exhibited
accentuated song divergence which could reflect species interactions. Overall the T.
commodus populations were not strongly differentiated by song, with the few significant
population differences dominated by the same populations, AM and BN respectively.
Figure 2-9. T. commodus song population differentiation (DAPC)
DF coordinates for LD1-LD3 grouped by population and arranged in geographic order from
AM in the south to SV in the north.
50
Table 2-4. T. commodus: Pairwise population comparisons for song discriminant
functions (LDs)
Only significant (P < 0.05) comparisons are shown.
LD1
AM-BL
AM-BN
AM-CC
BN-CH
BN-MV
BN-SV
BN-UQ
LD2
BN-CC
BN-SV
LD3
AM-CH
Estimate
Std. Error
t value
Pr(>|t|)
-1.375
-2.145
-1.241
1.398
1.180
1.131
1.699
0.387
0.382
0.382
0.339
0.344
0.348
0.392
-3.551
-5.616
-3.250
4.121
3.434
3.246
4.338
0.013
<0.001
0.031
0.002
0.018
0.032
<0.001
-1.259
-1.116
0.354
0.348
-3.562
-3.205
0.012
0.036
1.145
0.369
3.105
0.047
T. commodus CHCs
There were differences among the T. commodus allopatric populations and the sexes
in the relative abundance of CHC peaks (MANOVA on the PC scores, using population as a
factor: Wilks λ=0.553, F7,295 = 4.37 , P < 0.0001; sex: Wilks λ=0.458, F1,295 = 57.3 , P <
0.0001; interaction between population and sex: Wilks λ=0.823, F7,295 = 1.374, P = 0.057). In
contrast to T. oceanicus populations the differences in CHCs between the sexes were still
significant using discriminant function analysis (with population as the prior group)
(MANOVA, using sex as a factor: Wilks λ=0.897, F1,295 = 8.35 , P < 0.001) (Figure 2-10).
Moreover, the interaction between population and sex increased (MANOVA, Pop:Sex: Wilks
λ=0.829, F7,295 = 2.009, P = 0.001). This suggests that some of the factors that differentiate
the sexes are similar to those that distinguish the populations, as discriminant function
analysis maximizes between group variance while minimizing within group variance.
51
Figure 2-10. T. commodus CHC population differentiation (DAPC)
Male and female scores for LD1-LD3 grouped by population and arranged in geographic
order from AM in the south to SV in the north.
Discriminant function analysis, conducted on the sexes separately, indicated that CHC
profiles in both males and females differed across populations (MANOVA, using population
as a factor: for females, Wilks λ = 0.809, F1,130 = 7.47 , P < 0.001; and for males, Wilks λ =
0.68, F1,223 = 25.85 , P < 0.001) (Figure 2-10). In both sexes the first discriminant axis (LD1)
exhibited a steadily increasing pattern of population differentiation from south to north
(Figure 2-10). Interestingly, the major CHC peaks associated with LD1 differed between the
sexes (Table S10). For the other discriminant components (LD2- LD3) there were
52
considerable sex differences in the pattern of population differentiation, in line with the
significant interaction effect (Figure 2-10).
In contrast to the pattern of song variation among T. commodus populations, a larger
number of pairwise population comparisons were significant and were not restricted to
populations at the species’ southern range margin (Table S9). There was greater population
differentiation based on CHCs among T. commodus populations in comparison to T.
oceanicus, which may partly reflect the broader geographic sampling area. The significant
population comparisons are shown in Table S9.
Discordance between song, CHCs and geographic distance
Geographically closer populations did not generally share a greater similarity in song
traits for either species (Figure 2-11A, 2-11A). The only marginal association between song
and geographic distance was among T. commodus populations, in which song elements
associated with LD3 (main loadings include intersong interval and number of chirps; Table
2-3) showed a weak increase from south to north (Pearson correlation r = 0.327, Mantel test,
p = 0.064; Figure 2-12.A). The lack of a spatial signal suggests song diversification is largely
independent of geographic distance.
In contrast, variation in cuticular hydrocarbon composition was associated with
geographic distance in both species (Figure 2-11B, 2-11B). Among T. oceanicus populations
CHC peaks associated with the second discriminant axis (LD2) showed a steady increase
with geographic distance in females but not males (Figure 2-12B). Intercepts were not
significantly different, but the slope was higher in females (female slope – 1.3 x 10-3, male
slope -1.8 x 10-4; significance tested using 1,000 bootstraps, p = 0.001). Among T. commodus
populations the first discriminant axis (LD1) showed a sharp increase with geographic
53
distance for both sexes (Figure 2-12.B). The slope was higher in males (2.2 x 10-3) than
females (1.2 x 10-3) (p = 0.01) but the intercept was marginally higher in females (0.48) than
in males (0.01) (p = 0.05). None of the other discriminant axes showed an association with
geographic distance.
Figure 2-11. T. oceanicus song, CHCs and geographic distance
Relationship between geographic distance (Euclidean) and T. oceanicus population
differentiation (measured as dissimilarity in discriminant function scores (LD)) for A) song,
B) CHCs. The sexes were examined separately and males highlighted by filled circles and
solid lines, while females are represented by open triangles and dashed red lines.
54
Figure 2-12. T. commodus song, CHCs and geographic distance
Relationship between geographic distance (Euclidean) and T. commodus population
differentiation (measured as dissimilarity in discriminant function scores (LD)) for A) song,
B) CHCs. The sexes were examined separately and males highlighted by filled circles and
solid lines, while females are represented by open triangles and dashed red lines.
Overall song exhibited a lower level of population differentiation and geographically
closer populations did not show a strong similarity among either species. In contrast CHCs
were highly variable among populations. Nonetheless, there was a clear geographic signal,
with CHC dissimilarity increasing with geographic distance. In particular, T. commodus
males and females showed a cline of CHC divergence from south to north, whereas only
female T. oceanicus exhibited an association with geographic distance. Examining the CHC
55
loadings, for both males and females, indicated that the pattern of population differentiation
was influenced by a large number of peaks, with very few peaks showing a major
contribution (Table S10, S8). Interestingly the peaks which contributed most to population
differentiation among males were not the same as those in females even though the overall
geographical trend was similar, in particular for LD1 in T. commodus (Table S10, S8).
56
Discussion
How and why do sexual signals vary?
Understanding how variation in mate recognition systems arises and persists within
populations and ultimately contributes to population divergence and species isolation is
central to speciation research. In this study I tested whether long vs. short range signals
exhibit similar patterns of divergence within and between species. In particular, I tested
whether species interactions have enhanced divergence in calling song and cuticular
hydrocarbon composition through reinforcement or alternatively led to convergence through
recent hybridization and introgression. Reinforcement has been suggested to most readily
occur in mosaic hybrid or contact zones, as initial levels of assortative mating are likely to be
high, and strong genetic associations between loci affecting mate choice and hybrid fitness
are likely to arise (Cain et al., 1999; Jiggins & Mallet, 2000). In addition, an important
question is how consistent are the effects of species interactions on the two different signal
modalities? For example, do both signals show enhanced divergence in sympatric populations
or does the pattern vary across sympatry?
Sex differences in mating signals and species boundaries
Male and female sexual signals may be under divergent natural and sexual selection
and therefore may differ in their contribution to population divergence and maintenance of
species boundaries. In this study, calling song, which is a male-specific trait, was clearly
delimited between the species in sympatry based on both field and laboratory recordings. The
apparent absence of T. commodus males amongst the sympatric F1 laboratory populations is
striking, considering genomic data revealed that the female progenitors of the laboratory
57
colonies encompassed individuals of both species (exception YP population which contained
only T. oceanicus individuals, Chapter 5). More surprisingly, the CHC data (from the same
F1 lab populations that song was recorded from) revealed a number of females (n = 45)
exhibiting a T. commodus like CHC profile among sympatric populations and also among T.
oceanicus allopatric populations (Figure S2). I refer to these females as T. commodus-like in
their CHC profiles as they had a large number of peaks; however, these females were distinct
from T. commodus samples from allopatric populations (Figure S3). Little is known about
CHCs in these species and, as I cannot ascertain the species identity of the sympatric
individuals it is difficult to elucidate the basis of this unexpected finding. Previously
published studies on CHCs in Telogryllus have focused solely on T. oceanicus. Moreover,
these studies have focused either only on males (Pascoal et al., 2015) or on a single
population from western Australia that is very far away from our sampling transect (Thomas
& Simmons, 2008a, 2008b, 2009, 2010; Simmons et al., 2013).
There are three most likely explanations, which are not mutually exclusive, for the
divergent T. commodus-like CHC samples and the discordance with song in delimiting the
species boundary.
1) Females which exhibited a T. commodus-like CHC profile outside the allopatric
species range may belong to either (or both) the T. commodus and T. oceanicus species and
exhibit divergent CHC profiles due to character displacement. The individuals in question
originate from a broad area encompassing at least two contact zones; T. commodus in the
south and the third putative species T. marini in the north. Nothing is known about CHC
profiles for the third putative species T. marini. The multiple contact zones provide numerous
opportunities for species interactions to have accentuated divergence in mating signals.
Reinforcement is expected to drive divergence in male mating signals and/or enhance female
discrimination (or, less commonly, male mate preferences (Albert & Schluter, 2004), but very
58
few examples of reproductive character displacement in a female mating signal exist
(exception Waage, 1975). Why species interactions may differentially influence female
mating signals in our study system is not clear. Fitness costs of heterospecific matings are
likely to be higher in females, but, as female crickets have strong control in selecting a
mating partner (as they must mount the male), selection would be expected to enhance female
discrimination rather than female attractiveness to conspecific partners.
2) Hybridization and introgression among each or all of the three putative species
across the two contact zones has resulted in a diverse array of cuticular profiles. Population
genomic analysis suggests current hybridization is rare or absent, however, historical
introgression could have contributed to CHC variability.
3) CHCs are highly variable, especially in females, and may not be species-specific
(Jallon & David, 1987; Ferveur, 2005). The use of two species-specific methods to measure
CHC profiles may not be sufficient to capture the variability. The expression of CHCs can
also be highly flexible and dependent on developmental or environmental conditions
(Thomas et al., 2011; Thomas & Simmons, 2011; Pascoal et al., 2015). A female-biased CHC
polymorphism may exist among T. oceanicus populations, with one female morph
representing the standard T. oceanicus CHC profile while the other morph exhibits a more
complicated profile with multiple additional peaks (Fig. S2). The divergent females in
question may therefore belong to the latter morph but were assigned to the T. commodus
group due to the multitude of peaks. Interestingly, there is a well-studied example of a
female-biased CHC polymorphism among D. melanogaster races. Females exhibit two
distinct morphs for the major cuticular hydrocarbon, one fixed in populations outside Africa
and the Caribbean while the other morph reaches high frequencies only within east Africa
and the Caribbean (Coyne et al., 1994). The adaptive significance of this polymorphism is
still being debated but it has been suggested that the distribution of the polymorphism reflects
59
adaptive divergence to different climates and pleiotropically also contributes to sexual
isolation among populations of D. melanogaster (Coyne et al., 1994; Coyne et al., 1999;
Coyne & Elwyn, 2006; Smadja & Butlin, 2009). Once again, why this polymorphism is sexspecific is not clear.
In this study, both species exhibited sex differences in their CHC profiles which may
reflect differential selection pressures operating on both sexes (Figure 2-7; Figure 2-9).
Female CHCs have been suggested to play a particularly important role in eliciting male
courtship (Jallon, 1984). Indeed, male preferences for CHCs in T. oceanicus have been found
to be stronger than those exerted by females (Thomas & Simmons, 2009, 2010). In addition,
close range mating trials have found males of both Teleogryllus species discriminate upon
close contact against heterospecific females (Chapter 3), implicating CHCs in male mate
choice and species isolation. Overall female CHCs may be under stronger sexual selection
whereas male CHCs might evolve primarily in response to environmental effects. The pattern
of isolation by distance detected for some CHC components may reflect adaptation to an
environmental gradient (e.g. temperature, rainfall etc). As sexual selection is predicted to be
weaker on male vs. female CHCs, due to the enhanced role of CHCs in male mate choice,
one may predict a stronger pattern of isolation by distance for males in their CHC profiles. In
line with this prediction, the principal axis of CHC differentiation (LD1) for both species was
more strongly correlated with geographic distance in males than in females (Figure 2-10;
Figure 2-11). However, among T. oceanicus females CHC differentiation on LD2 was
strongly correlated with geographic distance but not in males. These mixed results do not
provide strong support for male’s exhibiting greater isolation by distance in their CHC
profiles in response to an environmental gradient.
60
Reproductive Character Displacement – Long vs. short range signals
It has long been argued that species interactions can strengthen reproductive barriers
through reinforcement or, if reproductive isolation is already complete, drive character
displacement (Butlin, 1987; Howard 1993; Noor, 1999; Servedio & Noor., 2003). If sibling
species overlap geographically and heterospecific matings are costly, selection is predicted to
enhance assortative mating through signal or preference divergence (or both) (Dobzhansky,
1940, Coyne & Orr, 2004). In the absence of hybridization, divergence in signal elements
may also occur to minimize signal interference (reviewed in Gerhardt, 2013; but see Tobias
et al., 2014). Long range signals may be particularly prone to signal interference as a result of
being broadcast over long distances. In addition, long range signals may be subject to greater
selection to reduce mismatings, as earlier acting barriers have greater potential to limit gene
flow.
In this study close range signals exhibited a stronger pattern of divergence between
allopatric and sympatric populations than long range signals, contrary to the previous
prediction (Figure 2-6; Figure 2-8). The discordance between the signal traits may highlight a
more important role for close range signals such as CHCs in mediating species isolation. A
literature review by Veen et al. (2012) on reproductive isolation in Gryllidae found no
support for an increased importance of long range vs. short range signals in species isolation.
Instead short range mate choice appears equally important for determining mating outcomes.
In areas of high density, males and females may come into contact with conspecific and
heterospecific individuals by chance and the importance of close range signals in sexual
isolation may be more pronounced. In the field, I found males of both species singing in very
close proximity, which suggests individuals may often come into close contact with
heterospecifics inadvertently. Under these situations, close range signals may be equally, if
not more important, than long range signals in mate choice and interspecific sexual isolation.
61
Detecting a shift in a signal trait amongst sympatric versus allopatric populations is
not enough to conclude reproductive character displacement as preferences need to also be
considered (Howard, 1993). To test if CHC divergence was driven by reinforcement, further
work would need to examine if signal divergence is salient to mating partners and enhances
assortative mating. As both species’ CHCs profiles are complex and little is known about
CHC preferences it is difficult to infer if the direction of trait divergence maximizes
dissimilarity with the heterospecific signal. Identifying the compounds that actually
contribute to mate choice, in both males and females, requires careful mate preference
studies. Moreover, CHC divergence in sympatry may reflect environmental adaptation, such
as a response to temperature differences due to desiccation resistance between allopatric and
sympatric populations.
How predictable do we expect reinforcement or reproductive character displacement
to be? The effects of species interactions on the evolution of mate attracting signals will
depend on localized properties of each population, such as the relative proportion of each
species and the degree of ecological overlap (Nosil et al., 2003; Servedio & Noor., 2003;
Taylor et al., 2006). Lemmon (2009) found that in two- and three-species assemblages of
chorus frogs (Pseudacris) populations of the rarer species diverged in different acoustic
signal traits, nonetheless maximizing dissimilarity with the heterospecific signal across the
contact zone. A classic experimental evolution study by Higgie et al., (2000) on Drosophila
serrata and D. birchii demonstrated that species interactions in the lab could drive allopatric
populations to converge on the same cuticular hydrocarbon profiles as sympatric populations
from the field. In our study, the song components which contributed most to differentiating
populations were frequency and intersong interval which both play an important role in
female preference for song in T. oceanicus and T. commodus (Hennig & Weber, 1997;
Brooks et al., 2005; Bailey et al., 2016 unpublished). However, sympatric populations were
62
found to diverge in opposing directions, in some cases increasing song similarity to that of
the sibling species. This suggests that song divergence in sympatry is unlikely to be due to
selection enhancing assortative mating. Instead song diversification in sympatry may reflect
processes such as genetic drift or environmental heterogeneity.
The weak support for reproductive character displacement in calling song does not
mean it can be completely dismissed. The striking absence of T. commodus males amongst
the sympatric laboratory populations limited us to testing if sympatry had an effect on song
among T. oceanicus males only. However T. commodus males may be more likely to exhibit
character displacement as they appear to be in the minority (personal observation) and
therefore under stronger selection to reduce mismatings (Coyne & Orr, 2004; Lemmon, 2009;
Yukilevich, 2012). However, a previous study by Hill et al. (1972), which examined both
song and female preferences over a slightly smaller section of the same population transect
used in this study, found no evidence for accentuated divergence in song or preferences
amongst sympatric populations of either of these Teleogryllus species.
Sympatry can have other outcomes aside from reinforcement (Noor, 1999). One such
example comes from the well-studied Chorthippus parallelus grasshopper hybrid zone where
sympatry has no effect on female mate discrimination (Ritchie et al., 1989). Instead hybrid
males from sympatry have reduced hybrid dysfunction compared to those produced from
allopatric crosses (Ritchie et al., 1992; Virdee & Hewitt, 1992; Virdee & Hewitt, 1994).
Therefore, instead of selection favouring increased mate discrimination it may act to reduce
genetic incompatibilities among heterospecifics in sympatry, thereby reducing the potential
driver of reinforcement. To accurately predict the outcome of sympatry it is therefore
63
important to examine reproductive barriers using individuals from sympatric populations.
Previous studies estimating post-zygotic barriers between the two Teleogryllus study species
have been based on crosses of individuals from allopatric populations (Hogan & Fontana,
1973; Moran et al., in review), therefore the costs of hybridization may differ among
sympatric populations, reducing the potential for reinforcement to occur.
The weak level of population differentiation observed in our study provides no
support for the rapid divergence of song. Calling song and female preference components in
Teleogryllus have been suggested to have a shared genetic basis (i.e. “genetic coupling”
hypothesis) (Hoy et al., 1977), which could facilitate rapid shifts in signal-receiver systems
between populations (Shaw et al., 2011). However the evidence to support pleiotropy or even
linkage disequilibrium amongst preference and signal traits in this species is premature as
only F1 hybrids have been examined (Butlin & Ritchie, 1989). Female preferences for calling
song components in T. commodus have been found to be largely stabilizing (Brooks et al.,
2005; Bentsen et al., 2006) which may constrain song diversification and promote gradual
coevolution across the species ranges. Simmons and colleagues (2001, 2004) examined both
geographic and genotypic variation in song (in particular the proportion of chirp) and
associated female preferences among T. oceanicus populations, and found no covariance
between female preference and song traits, and a mismatch between female preferences and
the mean population trait values. This mismatch between female preference and the male
traits across multiple populations may indicate these traits are not coevolving. However
experimental design could also confound the results (Greenfield et al., 2014). Linkage
disequilibrium will decline by 50% per generation under disassortative mating which may
explain the failure to detect a genetic correlation between signals and preferences among
colonies maintained in the laboratory for multiple generations.
64
Female preferences for male CHC profiles in T. oceanicus have been found to be
largely under disruptive sexual selection (Thomas & Simmons, 2009), but also some peaks
are under stabilizing selection (Simmons et al., 2013). Similarly male preferences for females
CHCs have been found to be under both stabilizing and directional selection gradients
(Thomas & Simmons, 2010). The intensity of male preferences acting on female CHCs is
much stronger than female preferences acting on male CHCs (Thomas & Simmons, 2010). In
addition, the peaks under sexual selection in males and females differed. Overall, differences
in the form of sexual selection between male song (largely stabilizing) and male and female
CHCs (different elements under disruptive, directional, and stabilizing selection) may
contribute to the greater level of population differentiation observed for CHCs vs. song.
Spatial scale is an important factor to consider when examining hybrid or contact
zones, as the pattern of variation can be dependent on the scale of analysis, with clinal
variation on a fine scale and a bimodal distribution on a larger scale (Harrison, 1986; Ross &
Harrison, 2002; Bridle & Butlin, 2002). Hybrid zones can range from extremely narrow
contact zones to complicated patchworks of populations dominated by bimodal parental types
interspersed among admixed individuals (Barton & Hewitt, 1985; Harrison & Rand, 1989;
Harrison, 1993). The broad level of sampling employed in this study, with populations
separated by ca. 100km or more, may have prevented us from detecting fine scale patterns of
species interactions. In addition, phenotypic data may not be sufficiently sensitive to identify
low levels of introgression between the species as phenotypic integrity may mask
considerable levels of gene flow (Poelstra et al., 2014). A more comprehensive test of the
strength of reproductive isolation and the relative importance of reproductive barriers is to
combine phenotypic data and compare the pattern of differentiation with genomic markers
across both allopatric and sympatric populations (Payseur, 2010; Chapter 5).
65
Interaction between sexual and ecological selection
Distinguishing the relative contributions of sexual selection and ecological adaptation
and how they interact to delimit species boundaries is a serious challenge in speciation
research (Safran et al., 2013; Scordato et al., 2014). Many classic examples evoking sexual
selection as an important force for species isolation also implicate a role for ecological
adaptation which may interact to drive divergence in sexual traits; such as male nuptial colour
and opsin divergence in Cichlids (Seehausen et al., 1999; Seehausen, 2006; Arnegard &
Kondrashov, 2004) and body size in sticklebacks (Nagel & Schluter, 1998; Head et al.,
2009). Divergent ecological selection could directly or indirectly alter acoustic behaviour or
chemical signals among populations, such as environmental heterogeneity promoting
differential body sizes and desiccation resistance, habitat structure to optimize signal
transmission and predation risks altering structural or temporal signalling behaviour.
In our study, the clear discordance between song and CHCs in their relationship with
geographic distance (Figure 2-11; Figure 2-12), which was particularly acute among T.
commodus populations, suggests the two traits are under different selection pressures. CHCs
are likely to be under both ecological and sexual selection whereas song is likely to be
primarily influenced by the latter. The interaction between ecological and sexual selection on
signal traits may be antagonistic and constrain signal diversification and sexual isolation
(Sharma et al., 2011). Geographic clines for CHCs have been identified in some Drosophila
species, however the factors that promote and maintain this variation are not fully resolved
(Coyne & Elwyn, 2006). In one intensively studied group D. serrata, on the east coast of
Australia, a cline of CHC compounds increasing in chain length in males, has been found to
be functionally associated with higher temperatures (Frentiu & Chenoweth, 2010). Indeed,
long chained CHC compounds have been suggested to increase desiccation resistance
66
(Rouault et al., 2001; Savarit & Ferveur, 2002). However, in our study, the strong cline in
CHCs among T. commodus males and females, encompassing a similar geographic region as
D. serata, is not clearly associated with an increase in CHC chain length, with no clear
pattern in the loadings (Table S10). This is interesting as one of the main differences between
the two Teleogryllus species in their CHC profiles, is that T. oceanicus individuals have CHC
peaks with a greater retention time, reflecting longer chained CHC components. The higher
proportion of long chained CHC compounds may reflect adaptation to the higher
temperatures in the northern region where T. oceanicus occurs. Overall, the processes
involved in maintaining the CHC cline in T. commodus may be complicated, but further
examination is likely to provide important insights into the ecological factors contributing to
CHC variation and population divergence.
Conclusions
Theory predicts that geographic overlap with heterospecifics can lead to strengthening
of reproductive barriers or hybridization and the erosion of species boundaries. How mate
recognition systems diverge during or after speciation is important in determining which of
the above outcomes occurs. This study looked at broad patterns of geographic variation in
calling song and CHCs for a classic study species for sexual selection, across an extensive
latitudinal transect. I identified the presence of a third putative sibling species T. marini
which has only been previously described once and appears localized to an area of north
eastern Australia. The sample transect therefore encompassed at least two contact zones with
potential for interspecies interactions. Calling song clearly delimited the species boundary in
sympatry with no intermediate individuals suggesting current hybridization is rare or absent.
67
CHC variation exhibited an unexpected pattern of female-specific variation across sympatric
and allopatric populations. Further work is needed to determine whether this pattern is the
outcome of species interactions, such as reproductive character displacement, or represents a
female-biased polymorphism. Song variation among allopatric and sympatric populations
was not strongly supportive of reproductive character displacement. In contrast, the relative
abundance of CHCs was significantly different among both geographic regions. However,
this discrepancy between traits may reflect an environmental association rather than an
increased importance of CHCs in mediating species interactions. Future research would
therefore benefit from testing for associations between CHC diversity and environmental
variables. The location of the contact zone and the range limits of both species are also likely
to be associated with environmental variables and therefore warrants further study.
68
Supplementary
Two potential contributing sources of song variation, temperature and body size, were
examined in the lab recordings. Temperature was found to be weakly correlated with
different song elements in both species (Table 1.). For T. commodus temperature was
positively associated with frequency, chirp trill interval, and negatively with intersong
interval. For T. oceanicus temperature was negatively associated with song duration, trill
pulse duration, and chirp duration.
Body size was also found to associate with different call elements in both species. Amongst
T. commodus individual’s weight had a weak, positive association with the number of trills
and pronotum length showed no relationship with any song parameters. While for T.
oceanicus, weight had a weak, positive relationship with chirp duration and intersong
interval, and pronotum length (PL) was positively associated with frequency (Table 2.).
Amongst T. oceanicus populations there was no significant differences for weight, while
there was for PL (Table 5.). Amongst T. commodus populations, both pronotum length and
weight varied significantly (p<0.001). Although for neither species did these morphological
traits show a relationship with geographic distance (Mantel test p >0.05).
Table S1. Temperature effects on both species call characteristics
Species
T. commodus
T. oceanicus
Song Element
df
r
p
Frequency
Chirp Trill Interval
122
122
0.21
0.19
0.022
0.037
Intersong Interval
122
-0.25
0.006
Song duration
Trill pulse duration
130
130
-0.175
-0.211
0.044
0.015
Chirp duration
130
-0.2
0.022
69
Table S2. Body size effects on call characteristics (residuals corrected for temperature)
Species
T. commodus
T. oceanicus
Trait
Song Element
df
r
p
Weight
Weight
Weight
Trill number
Chirp Duration
Intersong Interval
113
117
117
0.186
0.193
0.202
0.047
0.035
0.028
PL
Frequency
112
0.2
0.033
Figure S1. PCA on field recorded data, plot of PC scores on the first two dimensions. The
two main species are delimitated on PC1. While a third putative species T. marini sampled
from the extreme northern area of Queensland overlaps with T. commodus.
70
Species
T.com
T.com
T.com
T.com
T.com
T.com
T.com
T.com
T.oc
T.oc
T.oc
T.oc
T.oc
T.oc
T.oc
T.oc
Pop
AM
BN
CC
MV
BL
CH
UQ
SV
HB
TS
RH
YP
PL
JC
DV
KH
16
18
18
14
17
16
17
15
17
11
20
17
18
16
16
12
n
607.6
635.12
649.61
643.4
604.2
634.67
586.36
671.14
658.06
716.44
650.47
656.36
567.75
640.92
656.31
444.00
Weight
0.16
0.17
0.17
0.17
0.17
0.17
0.16
0.17
0.17
0.17
0.16
0.17
0.15
0.17
0.17
0.15
PL
0.26
0.27
0.29
0.27
0.27
0.25
0.28
0.30
0.26
0.30
0.26
0.29
0.25
0.24
0.29
0.29
Chirp
Duration
4.51
4.92
5.02
4.83
4.67
4.24
5.02
5.23
5.31
5.82
5.30
5.82
5.24
5.11
6.11
6.07
Chirp
Number
1.44
1.42
1.45
1.42
1.27
1.37
1.38
1.51
1.46
1.67
1.65
1.42
1.39
1.27
1.55
1.43
Total
Duration
8.28
7.84
7.81
7.97
7.06
7.75
8.20
8.29
2.40
2.22
2.34
2.61
2.08
1.90
3.29
1.32
Number
of Trills
71
5047.84
5029.97
4910.21
5022.81
4913.99
4925.99
4949.14
5191.93
3894.00
3925.35
3938.67
3855.88
3896.07
3886.11
3925.85
3999.69
Freq
1.14
1.09
1.10
1.09
0.95
1.07
1.05
1.15
1.18
1.35
1.37
1.12
1.12
1.02
1.24
1.13
Trill
Total
Duration
0.05
0.06
0.06
0.06
0.05
0.06
0.05
0.06
0.02
0.02
0.02
0.02
0.02
0.01
0.02
0.02
Chirp
Trill
Interval
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
Chirp
Pulse
Length
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
Chirp
Interpulse
Interval
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.02
0.02
0.02
0.02
0.02
0.02
0.02
Trill
Pulse
Duration
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
Trill
Interpulse
Interval
2.15
2.22
2.15
2.16
2.19
2.18
2.08
2.10
16.93
22.09
18.24
24.48
20.23
12.51
15.94
29.46
Average
Pulses
per Trill
Table S3. Laboratory recordings showing song parameters means (ms and Hz) for each population of both species (observed data). Population
mean weights (mg) per individual and pronotum length (PL) (mm). The grey coloured rows represent sympatric populations.
Species
T.com
T.com
T.com
T.com
T.com
T.com
T.com
T.com
T.oc
T.oc
T.oc
T.oc
T.oc
T.oc
T.oc
T.oc
Pop
AM
BN
CC
MV
BL
CH
UQ
SV
HB
TS
RH
YP
PL
JC
DV
KH
16
18
18
14
17
16
17
15
17
11
20
17
18
16
16
12
n
18.20
12.67
15.37
7.06
14.66
11.91
16.00
17.91
16.99
24.2
17.14
14.1
14.81
29.66
19.96
18.31
Weight
7.26
5.98
4.51
3.46
6.13
6.73
6.23
6.10
6.73
9.21
6.94
6.14
7.66
5.35
7.45
6.28
PL
16.68
34.48
16.67
25.16
18.45
21.09
22.47
26.83
16.34
20.83
21.57
13.28
13.63
32.85
22.84
23.71
Total
Duration
20.99
17.08
11.87
21.49
14.43
15.39
17.07
17.40
16.18
18.12
17.74
18.57
10.09
24.08
25.84
36.12
Chirp
Duration
18.97
15.52
10.48
19.37
16.57
14.82
17.47
14.37
17
16.38
14.23
14.73
10.46
21.49
25.09
32.74
Chirp
Number
11.79
13.76
15.47
8.45
7.95
14.19
16.93
6.71
8.51
9.56
7.52
15.27
12.52
15.53
18.49
10.54
Chirp
Pulse
Length
72
27.14
20.51
27.80
22.75
13.44
31.3
19.33
19.29
19.35
21.91
20.88
30.78
25.69
30.61
25.18
15.41
Chirp
IPI
2.83
4.07
3.26
3.68
3.7
4.39
3.30
5.08
3.09
2.59
2.99
4.94
4.68
5.73
3.99
3.64
Freq
23.28
29.87
18.02
15.87
14.9
16.88
13.90
30.42
31.26
32.43
60.96
50.59
51.28
30.45
29.73
21.79
Chirp
Trill
Interval
19.10
43.30
22.18
29.75
23.33
25.72
28.16
31.98
19.67
24.03
23.85
16.07
17.16
36.59
26.28
33.72
Trill Total
Duration
17.72
38.10
20.57
22.97
20.96
24.87
25.53
25.06
56.47
43.5
47.54
56.01
38.57
52.5
48.6
26.26
Trill
Number
16.05
14.39
12.27
2.84
12
10.8
10.35
13.03
7.98
9.15
10.95
16.15
14.71
10.66
17.58
12.31
Trill Pulse
Duration
23.24
29.33
32.07
35.34
21.01
35.06
27.26
19.69
14.91
26.91
23.14
37.45
39.45
33.62
40.78
42.81
Trill
IPI
15.22
13.51
11.62
7.23
18.61
11.35
4.44
8.08
62.18
67.97
59.82
61.1
49.66
74.26
63.59
44.43
Average
Pulses
per Trill
Table S4. Laboratory recordings showing coefficient of variation of song parameters for each population of both species. The grey coloured
rows represent sympatric populations.
Table S5. Principal Component loadings from both the field and lab data. The loadings
displayed are those >0.5.
Field Data:
Lab
Data
Song
Variables
Average Pulses
Chirp
Duration
per Trill
Chirp
Interpulse
Chirp
Number
Chirp
Pulse
Interval
Chirp
LengthTrill
Frequency
Interval
Intersong
Total
Duration
Interval
Trill number
Trill Interpulse
Interval
Trill
Pulse
Trill
Total
Duration
Duration
Eigenvalue
Total Variance
Cumulative
percentage of
variance
PC1
PC2
PC3
PC4
PC1
0.74
0.81
0.78
-0.60
-0.90
0.61
0.75
-0.78
0.58
0.71
5.83
44.88
44.88
0.57
0.57
2.34
18.00
62.88
-0.53
0.72
0.64
1.59
12.23
75.10
0.49
-0.46
0.94
7.22
82.32
-0.72
0.71
0.85
0.91
-0.67
0.85
0.77
4.82
37.04
37.05
73
PC2
0.62
0.81
0.72
2.57
19.76
56.80
PC3
PC4
-0.71
0.59
-0.71
2.02
15.50
72.31
0.59
0.58
-0.52
-0.63
1.66
12.73
85.034
Figure S2. Chromatograms of model T. commodus (black) and T. oceanicus (red) CHC
profiles. The method for integrating the species-specific CHC profiles was primarily
developed based on these representative samples. The blue chromatograms indicate samples
which exhibited CHC profiles that differed from the two standard species profiles but were
analysed using the T. commodus method due to the large number of peaks. The blue
chromatograms represent individuals that originated from T. oceanicus allopatry. In A) and
B) the blue chromatogram represents females from the same population (PL) and show
markedly different profiles from the expected T. oceanicus or T. commodus profiles.
74
Figure S3. Discriminant function analysis of CHCs from females that were assigned to the T.
commodus group (due to the larger number of peaks) and analysed using the corresponding T.
commodus method. LD1 revealed a clear distinction among females originating from the
three different geographic regions: T. commodus allopatry, sympatry and T. oceanicus
allopatry.
75
Table S6. DAPC loadings for T. commodus females grouped by three geographic regions (T.
commodus allopatry, sympatry and T. oceanicus allopatry). Cells in bold indicate major
contributing factors.
CHC peaks
X4_Target_Response
X5_Target_Response
X6_Target_Response
X7_Target_Response
X8_Target_Response
X9_Target_Response
X11_Target_Response
X12_Target_Response
X13_Target_Response
X14_Target_Response
X16_Target_Response
X17_Target_Response
X18_Target_Response
X19_Target_Response
X20_Target_Response
X21_Target_Response
X22_Target_Response
X23_Target_Response
X24_Target_Response
X27_Target_Response
X28_Target_Response
X29_Target_Response
X30_Target_Response
X32_Target_Response
X33_Target_Response
X34_Target_Response
X35_Target_Response
X36_Target_Response
X37_Target_Response
X38_Target_Response
X40_Target_Response
X41_Target_Response
X42_Target_Response
X43_Target_Response
X50_Target_Response
LD1
LD2
0.200
0.001
0.138
0.039
0.236
0.249
0.268
0.169
-0.091
-0.064
0.128
0.176
0.013
0.066
0.040
0.083
-0.201
0.059
0.196
-0.007
0.349
-0.184
0.032
-0.154
-0.226
-0.284
-0.114
-0.249
0.016
0.021
0.040
-0.050
0.125
-0.043
0.006
0.044
-0.028
0.036
-0.126
0.044
0.047
0.092
-0.092
0.109
-0.052
0.001
0.040
0.206
0.078
-0.243
-0.189
0.031
-0.154
-0.013
0.277
0.146
-0.034
-0.059
-0.071
-0.091
-0.310
0.091
-0.123
0.107
-0.306
0.028
-0.080
-0.013
0.025
-0.078
76
Table S7. T. oceanicus CHCs: Tukey pairwise population comparisons for LD1, LD2, LD3
for both sexes separately. Only significant (P < 0.05) comparisons are shown. DAPC was run
with population as the prior grouping factor.
Females
Males
Estimate
Std.
Error
t value
Pr(>|t|)
Estimate
Std.
Error
t value
Pr(>|t|)
JC-HB
-1.212
0.353
-3.434
0.017
HB-DV
1.455
0.286
5.084
<0.001
KH-DV
1.068
0.318
3.361
0.022
KH-JC
1.824
0.340
5.371
<0.001
JC-DV
0.903
0.292
3.095
0.0450
KH-HB
-1.843
0.264
-6.982
<0.001
PL-DV
1.971
0.339
5.808
<0.001
KH-JC
-1.291
0.270
-4.781
<0.001
PL-HB
1.515
0.317
4.786
<0.001
PL-KH
1.115
0.250
4.462
<0.001
PL-JC
2.727
RH-KH
-1.695
0.360
7.577
<0.001
RH-DV
2.082
0.286
7.271
<0.001
0.404
-4.192
<0.001
RH-JC
1.178
0.283
4.163
0.0011
RH-PL
-2.598
0.422
-6.164
<0.001
RH-KH
2.469
0.264
9.353
<0.001
TS-KH
-1.174
0.318
-3.696
0.008
RH-PL
1.354
0.264
5.128
<0.001
TS-PL
-2.077
0.339
-6.122
<0.001
TS-DV
1.571
0.277
5.669
<0.001
YP-DV
1.520
0.373
4.072
0.002
TS-KH
1.959
0.254
7.709
<0.001
YP-JC
2.277
0.392
5.804
<0.001
TS-PL
0.844
0.254
3.320
0.0227
YP-RH
2.147
0.449
4.779
<0.001
YP-DV
1.725
0.273
6.311
<0.001
YP-TS
1.627
0.373
4.357
<0.001
YP-KH
2.113
0.250
8.452
<0.001
YP-PL
0.998
0.250
3.991
0.0023
LD1
LD2
LD1
LD2
HB-DV
-1.049
0.332
-3.161
0.039
KH-DV
1.631
0.273
5.968
<0.001
JC-DV
-1.816
0.373
-4.863
<0.001
KH-HB
0.976
0.264
3.696
0.0065
KH-DV
-1.052
0.318
-3.312
0.025
PL-DV
1.802
0.273
6.594
<0.001
PL-DV
-1.711
0.339
-5.041
<0.001
PL-HB
1.147
0.264
4.343
<0.001
RH-DV
-2.086
0.433
-4.818
<0.001
PL-JC
0.992
0.270
3.672
0.0069
TS-DV
-2.809
0.354
-7.946
<0.001
RH-DV
1.246
0.286
4.354
<0.001
TS-HB
-1.760
0.332
-5.304
<0.001
TS-DV
1.129
0.277
4.074
0.0017
TS-KH
-1.757
0.318
-5.529
<0.001
YP-DV
1.292
0.273
4.727
<0.001
TS-PL
-1.099
0.339
-3.238
0.031
YP-DV
-2.362
0.373
-6.325
<0.001
YP-HB
-1.313
0.353
-3.720
0.007
YP-KH
-1.309
0.340
-3.855
0.005
LD3
LD3
HB-DV
1.120
0.332
3.374
0.021
JC-DV
-1.413
0.292
-4.842
<0.001
JC-DV
1.939
0.373
5.193
<0.001
JC-HB
-1.186
0.283
-4.191
0.0010
KH-DV
1.019
0.318
3.208
0.034
KH-JC
1.005
0.270
3.722
0.0059
TS-HB
-1.037
0.332
-3.124
0.044
PL-JC
1.268
0.270
4.697
<0.001
TS-JC
-1.856
0.373
-4.971
<0.001
RH-JC
1.032
0.283
3.647
0.0078
YP-JC
-1.431
0.392
-3.647
0.009
TS-JC
1.545
0.274
5.640
<0.001
YP-JC
0.888
0.270
3.288
0.0254
77
Table S8. T. oceanicus CHC loadings for both males and females for three discriminant
functions based on 16 retained PCs. Loadings in bold indicate main contributing factors.
CHC peaks
Female
LD1
LD2
LD3
Male
LD1
LD2
LD3
X6_Target.Response
-0.124
-0.470
-0.989
-0.454
0.600
0.282
X19_Target.Response
-1.284
-0.581
1.164
0.593
-0.865
-0.148
X36_Target.Response
-0.421
0.424
-0.503
0.731
-0.497
-0.241
X37_Target.Response
0.042
0.152
1.027
0.255
-0.621
-0.492
X38_Target.Response
0.128
-1.591
0.372
0.220
0.132
0.320
X41_Target.Response
0.150
-0.069
0.292
0.092
0.115
-0.485
X42_Target.Response
0.236
0.808
-0.579
-0.835
0.784
0.021
X43_Target.Response
1.014
0.556
0.257
-0.499
0.103
0.492
X44_Target.Response
0.366
0.307
-0.976
0.152
0.300
0.007
X45_Target.Response
0.490
-0.200
-0.249
0.633
0.679
-0.443
X46_Target.Response
-0.237
1.269
-0.120
-1.462
-0.752
0.098
X47_Target.Response
0.542
0.102
-0.236
-0.215
0.046
-0.353
X48_Target.Response
0.083
0.141
-0.265
-0.445
0.258
0.755
X49_Target.Response
-0.018
0.126
0.085
0.285
-0.170
0.339
X50_Target.Response
0.310
-0.115
0.025
2.157
0.826
-0.374
X51_Target.Response
-0.904
0.856
-0.777
-0.417
-0.082
-0.740
X52_Target.Response
-0.099
-1.242
-0.480
-0.419
-0.312
0.182
X53_Target.Response
-0.318
-1.374
0.788
-0.262
-0.389
0.228
X54_Target.Response
0.120
-0.612
0.379
0.225
-0.115
-0.108
X57_Target.Response
0.067
0.581
0.111
0.004
0.361
0.911
% Variation explained
32.99
28.14
13.06
48.15
19.75
12.79
78
Table S9. T. commodus CHCs: Tukey pairwise population comparisons for LD1, LD2, LD3
for both sexes separately. Only significant (P < 0.05) comparisons are shown. DAPC
coordinates from analysis with population as grouping factor.
Females
Males
Estimate
Std.
Error
t value
Pr(>|t|)
LD1
Estimate
Std.
Error
t value
Pr(>|t|)
BL-AM
2.029
0.356
5.695
<0.001
BL-AM
2.661
0.313
8.495
<0.001
CH-AM
3.062
0.347
8.826
SV-AM
1.543
0.344
4.481
<0.001
CH-AM
2.906
0.309
9.406
<0.001
<0.001
MV-AM
1.274
0.309
4.122
0.001
UQ-AM
2.133
0.350
6.096
<0.001
SV-AM
2.989
0.285
10.491
<0.001
BN-BL
-1.670
0.349
-4.792
<0.001
UQ-AM
3.265
0.318
10.272
<0.001
CC-BL
CH-BL
-1.817
0.330
-5.507
<0.001
BN-BL
-1.719
0.334
-5.149
<0.001
1.034
0.306
3.380
0.020
CC-BL
-2.135
0.339
-6.292
<0.001
MV-BL
-1.324
0.325
-4.075
0.002
MV-BL
-1.387
0.320
-4.330
<0.001
CH-BN
2.704
0.339
7.975
<0.001
CH-BN
1.964
0.330
5.954
<0.001
SV-BN
1.184
0.336
3.521
0.013
SV-BN
2.047
0.307
6.658
<0.001
UQ-BN
1.774
0.342
5.188
<0.001
UQ-BN
2.323
0.338
6.868
<0.001
CH-CC
2.850
0.320
8.911
<0.001
CH-CC
2.380
0.335
7.097
<0.001
SV-CC
1.331
0.317
4.198
0.001
SV-CC
2.463
0.313
7.861
<0.001
UQ-CC
1.920
0.323
5.947
<0.001
UQ-CC
2.739
0.344
7.972
<0.001
MV-CH
-2.358
0.315
-7.491
<0.001
MV-CH
-1.632
0.316
-5.162
<0.001
SV-CH
-1.519
0.292
-5.206
<0.001
SV-MV
1.715
0.293
5.859
<0.001
UQ-CH
-0.930
0.298
-3.118
0.044
UQ-MV
1.991
0.325
6.128
<0.001
UQ-MV
1.428
0.318
4.492
<0.001
BL-AM
1.850
0.356
5.193
<0.001
BL-AM
1.005
0.313
3.209
0.034
BN-AM
3.555
0.385
9.229
<0.001
BN-AM
1.208
0.323
3.740
0.006
CC-AM
1.354
0.368
3.676
0.008
CC-AM
1.385
0.329
4.215
0.001
SV-AM
1.620
0.344
4.703
<0.001
CH-AM
1.187
0.309
3.842
0.004
UQ-AM
1.293
0.350
3.697
0.007
MV-AM
1.043
0.309
3.376
0.020
BN-BL
1.705
0.348
4.892
<0.001
SV-AM
1.809
0.285
6.349
<0.001
CH-BL
-0.911
0.306
-2.980
0.064
UQ-AM
-1.337
0.318
-4.207
0.001
MV-BL
-1.851
0.325
-5.698
<0.001
UQ-BL
-2.342
0.329
-7.120
<0.001
CC-BN
-2.200
0.361
-6.097
<0.001
UQ-BN
-2.545
0.338
-7.525
<0.001
CH-BN
-2.616
0.339
-7.717
<0.001
UQ-CC
-2.722
0.344
-7.922
<0.001
MV-BN
-3.556
0.356
-9.979
<0.001
UQ-CH
-2.524
0.325
-7.769
<0.001
SV-BN
-1.935
0.336
-5.754
<0.001
UQ-MV
-2.380
0.325
-7.325
<0.001
UQ-BN
-2.262
0.342
-6.615
<0.001
UQ-SV
-3.146
0.302
-10.413
<0.001
MV-CC
-1.356
0.338
-4.009
0.002
MV-CH
-0.940
0.315
-2.988
0.063
SV-MV
1.621
0.312
5.199
<0.001
UQ-MV
1.295
0.318
4.074
0.002
LD1
LD2
LD2
79
Table S9. Continuation from previous Table.
Estimate
Std.
Error
t value
Estimate
Std.
Error
t value
Pr(>|t|)
BN-AM
1.385
0.385
3.596
0.010
CC-AM
2.147
0.368
5.828
<0.001
BL-AM
1.756
0.313
5.605
<0.001
BN-AM
2.727
0.323
8.444
<0.001
CH-AM
1.751
0.347
5.045
<0.001
CC-AM
1.059
0.329
3.224
0.032
MV-AM
1.593
0.364
4.378
<0.001
UQ-AM
1.309
0.318
4.120
0.002
SV-AM
1.581
UQ-AM
1.265
0.344
4.592
<0.001
BN-BL
0.971
0.334
2.909
0.078
0.350
3.617
0.009
CH-BL
-1.745
0.320
-5.448
<0.001
BN-BL
1.645
0.348
4.722
<0.001
MV-BL
-0.984
0.320
-3.071
0.050
CC-BL
2.408
0.330
7.298
<0.001
SV-BL
-1.204
0.297
-4.050
0.002
CH-BL
2.011
0.306
6.577
<0.001
CC-BN
-1.668
0.348
-4.787
<0.001
MV-BL
1.854
0.325
5.706
<0.001
CH-BN
-2.717
0.330
-8.235
<0.001
SV-BL
1.842
0.303
6.083
<0.001
MV-BN
-1.955
0.330
-5.926
<0.001
UQ-BL
1.526
0.309
4.938
<0.001
SV-BN
-2.175
0.307
-7.074
<0.001
UQ-BN
-1.417
0.338
-4.191
0.001
CH-CC
-1.049
0.335
-3.128
0.042
UQ-CH
1.299
0.325
3.999
0.003
LD3
Pr(>|t|)
LD3
80
Table S10. T. commodus CHC loadings for both males and females for three discriminant
functions based on 30 retained PCs. Loadings in bold indicate main contributing factors.
CHC peaks
Female
LD1
LD2
LD3
Male
LD1
LD2
LD3
X4_Target_Response
0.564
0.563
-1.395
0.317
0.922
-0.804
X5_Target_Response
-0.834
-0.651
1.283
-2.005
1.096
-0.201
X6_Target_Response
-0.971
0.535
0.677
-0.487
-0.129
0.715
X7_Target_Response
0.363
-0.137
0.554
-0.465
0.960
-0.218
X8_Target_Response
0.213
0.620
0.768
0.586
0.038
-0.639
X9_Target_Response
-0.147
0.866
0.577
0.974
0.074
0.338
X11_Target_Response
1.889
-0.937
0.331
-0.334
-0.371
0.094
X12_Target_Response
0.140
-0.116
0.738
0.209
-2.281
-0.218
X13_Target_Response
0.598
-0.538
-1.269
0.778
1.044
1.145
X14_Target_Response
-0.037
-0.555
0.171
1.039
-0.225
0.178
X16_Target_Response
-0.207
1.199
-0.422
0.375
-0.662
0.107
X17_Target_Response
-0.710
-0.081
-1.727
0.064
0.321
-0.358
X18_Target_Response
-0.427
-0.524
-0.189
0.167
-1.723
-0.154
X19_Target_Response
0.467
-0.744
-0.217
-0.428
-0.078
-0.881
X20_Target_Response
0.751
-0.681
-1.296
0.940
1.683
1.356
X21_Target_Response
-0.077
-0.265
-0.357
-0.987
1.699
-0.146
X22_Target_Response
0.620
0.931
0.016
0.524
-0.713
1.156
X23_Target_Response
-1.213
-0.635
-0.418
-0.626
-1.684
0.275
X24_Target_Response
0.506
-0.016
0.780
0.798
-0.149
-0.659
X27_Target_Response
-0.460
0.117
-0.399
-0.064
1.280
0.118
X28_Target_Response
-1.093
-0.116
-0.173
-0.071
-0.009
0.516
X29_Target_Response
-0.327
0.918
0.845
0.459
0.979
-0.082
X30_Target_Response
0.323
0.263
0.745
0.528
-0.292
-0.502
X32_Target_Response
0.745
-0.589
-0.198
-0.721
0.728
-0.377
X33_Target_Response
0.004
-0.207
0.111
-0.305
-0.505
0.381
X34_Target_Response
0.091
0.554
0.765
-0.191
0.492
-0.708
X35_Target_Response
-0.204
0.663
0.429
-0.126
-0.071
0.152
X36_Target_Response
0.155
1.141
0.634
-0.115
0.048
-0.284
X37_Target_Response
0.599
-0.328
-0.623
0.299
0.119
-0.345
X38_Target_Response
-0.271
0.102
-0.016
-0.718
-0.662
-0.182
X40_Target_Response
0.091
0.187
-0.615
0.478
0.250
-0.475
X41_Target_Response
-0.098
0.077
-0.872
-0.290
-0.171
-0.280
X42_Target_Response
0.047
-1.815
0.301
-0.149
-0.411
0.084
X43_Target_Response
-0.713
0.488
0.040
-0.166
0.275
0.163
X50_Target_Response
-0.489
-0.076
-0.074
-0.348
-0.097
1.115
% Variation explained
27.8
24.0
16.0
35.8
21.4
17.2
81
Chapter 3
Sexual Isolation and the Evolution of Sexually Selected Traits
Abstract
Understanding the genetic and behavioural basis for species differences is key to
deciphering the mechanisms driving the establishment of reproductive barriers and hence
speciation. A considerable volume of research has examined the genetics of postzygotic
barriers and provided important generalizations about their evolution. Less attention has been
paid to the genetics of behavioural isolation among parental species and their hybrids. In this
study I examine the role of behavioural barriers in reproductive isolation between two closely
related field crickets: Teleogryllus oceanicus and T. commodus. Most previous research on
sexual communication in these species has focused on long range calling song and patterns of
intraspecific mate choice. However, close range courtship behaviours may play an important
role in the species isolation. I conducted two sets of experiments. First, through close range
mating trials, I tested whether the parental species and their hybrids are sexually isolated.
Heterospecific pairs exhibited reduced rates of courtship behaviour due to male and female
discrimination and the species were symmetrically isolated. Surprisingly, hybrid males and
females were not sexually selected against, even though hybrid females are sterile. If
reproductive isolation is not complete amongst the parental species, backcrossing with hybrid
males could erode the species boundary. Second, I characterized the extent of divergence
amongst the species and their hybrids in key sexual traits; calling song, courtship song and
cuticular hydrocarbons. All three traits exhibited species differences which may mediate
assortative mating and species isolation. Components of song traits exhibited evidence of
sex-linkage and transgressive segregation, while CHCs exhibited only the latter. Different
82
patterns of inheritance suggest the genetic architectures may differ among the sexual traits.
Overall the results suggest that close range behaviours may be as important as long-distance
signals in contributing to interspecific sexual isolation.
Introduction
Multiple reproductive barriers are expected to develop between diverging populations
and incipient species. Isolating mechanisms can be broadly categorized based on the stage at
which they occur; prezygotic barriers prevent the formation of hybrid zygotes and
postzygotic barriers reduce the viability and fecundity of hybrids through developmental or
environmental effects (Coyne & Orr., 2004). Pre-mating and post-mating barriers generally
evolve at the same rate when taxa are allopatric, but pre-mating barriers evolve more rapidly
when taxa are sympatric, which suggests an important role for sexually selected traits in
species isolation (Coyne & Orr, 1989, 1997). As early acting barriers, behavioural barriers
have the potential to strongly curtail gene flow, even in the absence of post-mating barriers
(Seehausen et al., 1997; Arthur & Dyer, 2015).
Behavioural isolation
Closely related species are often most distinguishable primarily due to differences in
their mate recognition systems (Wells & Henry, 1998; Gleason & Ritchie, 1998; Mendelson
& Shaw, 2005; Seehausen, 2006). Mate recognition systems are composed of sexual signals
that are processed by a receiver, eventually resulting in mating and fertilization (Paterson,
1985). Sexual traits and preferences may diverge rapidly between closely related species due
to a combination of processes which can be divided into intraspecific (e.g. divergent sexual
and/or ecological selection and genetic drift within species) and interspecific processes (e.g.
83
reinforcement). If sibling species’ mate recognition systems overlap, extensive hybridization
and gene flow may occur in sympatric populations.
Behavioural isolation is often asymmetric, with species discrimination being much
stronger in one direction of the cross than the other (Coyne & Orr, 2004). This can have
important consequences for the extent and direction of gene flow. Several mutually nonexclusive theories have been proposed to explain asymmetrical behavioural isolation
including founder effects (Kaneshiro, 1980; Tinghitella & Zuk, 2009), runaway sexual
selection (Arnold et al., 1996), reinforcement (Coyne & Orr, 2004; Yukilevich, 2012), male
or female preferences for heterospecific mates (Ryan & Wagner, 1987; Ryan & Rand, 1995;
Svensson et al., 2016) and stronger intrasexual selection in one species (Debelle et al., 2016).
Kaneshiro (1980) proposed that speciation via a founder event may lead to loss of courtship
behaviours in the derived species, resulting in heightened discrmination against derived males
by ancestral females (but see Barton & Charlesworth, (1984)). Independent of the constraints
of population bottlenecks and the loss of courtship elements inherent to Kaneshiros model,
Arnold et al., (1996), proposed that runaway sexual selection could generate asymmetrical
behavioural barriers. Their model predicts behavioural asymmetries to be transient and only
materialize at the earliest stages of species divergence. However, long-diverged species often
show asymmetrical behvioural isolation (Coyne & Orr, 2004, p 226-227). An additional
explanation may be reinforcement; when species come into secondary contact, sexual signals
used for intraspecific mate choice may experience selection to enhance assortative mating to
avoid costly matings (Coyne & Orr, 2004). Divergent selection is expected to be stronger for
the rarer species which may lead to asymmetrical species discrimination (Peterson et al.,
2005; Yukilevich, 2012). However, allopatric species often exhibit asymmetrical behavioural
isolation (Coyne & Orr, 2004). Further work is needed to determine how consistent
84
behavioural asymmetries are, which could provide important insights into the causes of
behavioural isolation.
Hybrids as gatekeepers of gene flow
If reproductive isolation is incomplete between species the fate of the hybrids will
determine the extent and direction of interspecies gene flow (Broyles, 2002; Borge et al.,
2005). Hybrids are generally expected to suffer reduced fitness compared to individuals of
the parental species due to i) postzygotic intrinsic factors – physiological hybrid sterility and
hybrid inviaibility and ii) postzygotic extrinsic factors -disruptive ecological and sexual
selection (Coyne & Orr., 2004, Chapters 6 & 7). Recent theoretical and empirical work
highlight the importance of divergent ecological and sexual selection in speciation (Panhuis
et al., 2001; Ritchie, 2007; Safran et al., 2013; Scordato et al., 2014). However, the role of the
latter in determining the fitness of hybrids has been understudied (Servedio, 2004a).
Hybrids may be unfit due to difficulties in securing mates because they exhibit
intermediate or aberrant courtship behaviours (Chapter 1). Behavioural sterility can be due to
intrinsic or extrinsic postzygotic factors. Coyne & Orr (2004; p 254) suggest a subtle
distinction between the two forms; intrinsic hybrid behavioural sterility involves a
neurological or physiological defect leading to ablated or disrupted behaviours, while in the
extrinsic form, hybrids present intermediate behaviours, which may fall outside the range of
the parental species preferences (Naisbit et al., 2001). In the latter case hybrid fitness may be
frequency dependent, if hybrid females (with intermediate preferences) become locally
common, extensive gene flow among the species may occur (Bridle et al., 2006). Empirical
examples of sexual selection against hybrids include: in cichlid fish the symmetry and extent
of reproductive isolation depends on the phenotypic distinctiveness of hybrids (Selz et al.,
2014), hybrid males of Nasonia wasps (N. vitripennis and N. giraulti) are behaviourally
sterile due to failures to court females properly (Clark et al., 2010), female Chorthippus
85
grasshoppers (C. brunneus and C. jacobsi) discriminate against F1 hybrid song (Bridle et al.,
2006), in Heliconious butterflies (H. cydno and H. melpomene) hybrids of both sexes show a
reduction in mating >50% with the parental species (Naisbit et al., 2001) and in wolf spiders
(Schizocosa ocreata and S. rovneri) both male and female F1 hybrids suffer almost complete
behavioural sterility (Stratton & Uetz, 1986).
Hybrids may alternatively experience increased reproductive success due to
transgressive segregation (Rieseberg et al., 1999; Abbott et al., 2013), with more extreme
phenotypes eliciting a stronger response from the parental species (Rosenthal, 2013). Higher
mating success could even counteract negative hybridization costs such as reduced viability
or fertility (Pfennig, 2007). The expression of novel mating signals and preferences in
hybrids may then have important evolutionary consequences such as adaptive hybridization
and even hybrid speciation (Rieseberg et al., 1999; Seehausen, 2004; Jiggins et al., 2008;
Rosenthal, 2013). However, more experimental work is needed, in particular determining
how behavioural divergence in the parental species contributes to hybrid traits and mating
success (Blows, 1998; Blows & Allan, 1998; Rieseberg et al., 1999; Burke & Arnold, 2001),
and the role of hybrid mate choice in species isolation (Rosenthal, 2013; Schmidt & Pfennig,
2015).
Considerable research has focused on the genetics of physiological sterility (failure to
produce viable gametes) and inviability and given rise to important and consistent
generalizations about the genetics of postzygotic intrinsic barriers, in particular Haldane’s
rule (Haldane, 1922) and the large X effect (Coyne & Orr, 1989; Coyne & Orr, 2004). These
two patterns are so consistent they have been likened to rules (Coyne & Orr, 1989b) and
suggest that sex chromosomes play a key role in the establishment of post-zygotic barriers.
86
By contrast, the genetics of behavioural sterility have been much less studied but are likely to
play an important role in species isolation (Ritchie & Phillips, 1998). Behavioural barriers
may share certain genetic features with classic intrinsic barriers (i.e. physiological sterility
and inviaiblity) such as a disproportionate X effect or reduced (or disrupted) behaviour in the
heterogametic sex in accordance with Haldane’s rule (Davies et al., 1997).
Sex-linkage of genes influencing male signals and female preferences could have
important consequences for sexual trait elaboration and for sexual selection to drive
divergence in incipient species (Kirkpatrick & Hall, 2004). Theory predicts disproportionate
X-linkage of sexually selected traits (especially sex limited traits), due to the immediate
exposure of recessive mutations and sexually antagonistic fitness effects (Rice, 1984;
Charlesworth et al., 1987). If both premating and postmating barriers are sex-linked this
would have important consequences for the potential for reinforcement to complete
speciation, due to reduced recombination between these traits (Noor et al., 2001; Servedio &
Saetre, 2003; Lemmon & Kirkpatrick, 2006). However, empirical results have been mixed
with some studies supporting sex-linkage of courtship behaviours in female heterogametic
species (Z-linkage in Lepidoptera and some birds; Prowell, 1998; Saetre et al., 2003) but
generally not for male heterogamteic species (Reinhold, 1998; Ritchie & Phillips, 1998; Orr,
2001; Qvarnström & Bailey, 2009).
Study system
The two field cricket species T. commodus and T. oceanicus are a classic study system
for sexual communication, in particular acoustic behaviour (Huber & Loher, 1989). Both
species overlap in an extensive area, of what is believed to be secondary contact, on the
eastern coast of Australia and readily hybridize in the laboratory (Hill et al., 1972; Otte &
87
Alexander, 1983; Chapter 5). Research has focused almost extensively on intraspecific
reproductive behaviours and long range calling song has been assumed to be the primary
barrier isolating the species (Hill et al., 1972; Hoy & Paul, 1973; Pollack & Hoy, 1979; Hoy
et al., 1982; Hennig & Weber, 1997; Bailey & Macleod, 2013). However, female
discrimination against heterospecific song does not appear to be sufficiently precise to
prevent heterospecific matings. Bailey & Macleod (2013) reported that when females of
either species (from allopatric populations) were given a choice between conspecific and
heterospecific calling song, upto 25% of the time they chose the heterospecific song. When
given no choice Hill et al., (1972) found females of either species (from both allopatric and
sympatric populations) exhibited high rates of phonotaxis to heterospecific song (T.
commodus females 56 -58%; T. oceanicus females 40 – 55%). Overall this suggests that
calling song alone is not sufficient for reproductive isolation and other barriers must also be
important in maintaining the species boundaries.
Hybrid females, in both directions of the cross, are sterile (Hogan & Fontana, 1973;
Chapter 3), providing a remarkably rare exception to Haldane’s rule (Chapter 4). If both
primary and secondary sexual trait development are disrupted in hybrid females, this
exception to Haldane’s rule may extend to behavioural sterility (Davies et al., 1997). Intrinsic
hybrid behavioural sterility may occur due to a breakdown in the sex determination pathways
in hybrids. This could provide important insights into mechanisms of sex determination in
taxa without sexually dimorphic chromosomes (i.e. XO sex determination systems in
crickets). As hybrid males are fertile they potentially provide a gateway for inter-species gene
flow, therefore their ability to elicit and enact successful matings is of particular interest for
predicting the extent and direction of gene flow.
88
Courtship behaviour
Long range mate attraction and close range courtship behaviour in these species is a
multimodal process, involving at least acoustic and chemical signals (Loher & Rence, 1978;
Balakrishnan & Pollack, 1997; Bailey, 2011; Simmons et al., 2013). In both species, female
phonotaxis towards male calling song results in antennation and physical contact. Individuals
gain information from a partner through cuticular hydrocarbons (CHCs), which are waxy
molecules secreted on the cuticle of male and female insects and play an important role in sex
recognition and mate choice (Balakrishnan & Pollack, 1997; Tregenza & Wedell, 1997;
Smadja & Butlin, 2009; Bailey et al., 2011). In T. oceanicus, CHC profiles are heritable
(Thomas & Simmons, 2008a), sexually dimorphic (Thomas & Simmons, 2008b) and are used
by both males and females to discriminate against unattractive mates (Thomas & Simmons,
2009b; Thomas & Simmons, 2010). Upon physical contact males may produce courtship
song which then elicits females to mount the male. Courtship song has been less well studied
than calling song. In T. oceanicus it may act as an indicator of mate quality (Zuk et al., 2008;
Simmons et al., 2010; Simmons et al., 2013). Once females have mounted the male,
copulation lasts between 5-10 mins during which time males make jerking movements which
are necessary for threading the long spermatophore tube into the female genital tract
(Alexander & Otte, 1967; Loher & Rence, 1978) After mating, females may remove the
spermatophore, which can affect sperm transfer success, as complete sperm transfer takes ca.
40mins in both species (Simmons et al., 2003; Bussière et al., 2006). Behavioural or
mechanical isolation may occur at any stage during this sequence of courtship behaviours.
Most research has focused on intraspecific courtship behaviour leaving a substantial gap in
our knowledge about the potential role of courtship songs and CHCs in interspecific sexual
isolation.
89
Aims
The goal of this study was twofold: to examine the strength and direction of sexual isolation
amongst the two species and their hybrids and to characterize the pattern of sexual trait
divergence.
Sexual Isolation
In the first experiment I examined close range courtship behaviours amongst the parental
species and the hybrids to address three main questions.
1) Are the parental species behaviourally isolated and at which stage during courtship
does this occur?
2) Is behavioural isolation symmetrical, with both males and females of each species
equally discriminating against heterospecifics? If there is an asymmetry in behavioural
isolation amongst the parental species this would provide a testable prediction for directional
gene flow between the species.
3) Is courtship behaviour disrupted among hybrid males and females?
Sexual Trait Divergence
In the second experiment I characterized the patterns of divergence between the two
parental species and the reciprocal hybrids for three sexual traits: calling song, courtship song
and cuticular hydrocarbons (CHCs). The aim of this section was twofold:
1) The pattern of trait inheritance and expression in F1 hybrids, combined with the
behavioural trials data, can help identify the traits which promote behavioural isolation. If
reciprocal hybrids exhibit intermediate sexual phenotypes, females and males of the parental
90
species may discriminate equally against both hybrid types. Alternatively if hybrids exhibit
dominant (phenotypes more similar to one of the parental species than the other) or
transgressive (transgressive segregation when the mean trait value exceeds that of the
parental lines either in a positive or negative direction) sexual traits, asymmetrical mate
choice or complete behavioural sterility may be more likely.
2) The pattern of trait inheritance in hybrid crosses can also inform about the
underlying genetic architecture of sexually selected traits which is important for many models
of speciation (Shaw, 1996, 2000; Ritchie, 2000; Shaw & Parsons, 2002; Henry et al., 2002).
Since only F1 progeny were produced the main focus was on testing for dominance and sexlinkage (or maternal effects). If F1 hybrid males and females exhibit intermediate sexual
traits, it suggests these traits are based primarily on alleles acting additively in hybrids.
However, intermediate phenotypes do not imply that individual loci show no dominance, if
the species are fixed for dominant loci acting in opposite directions their effects will tend to
cancel out in the hybrids (Falconer & Mackay, 1996). Comparisons between reciprocal
hybrid males (heterogametic sex) can be used to test if traits are sex-linked, as they share the
same autosomes but differ in their X chromosomes, inheriting a single copy from the
maternal species (Reinhold, 1998). Early hybridization studies in these Teleogryllus species
suggested that most elements of hybrid calling song exhibit intermediate patterns, although
some, namely inter-trill interval and phrase reputation rate, may be sex-linked (Bentley &
Hoy, 1972). However, the inheritance pattern among hybrids for courtship song or CHC
profiles is not known.
91
Methods
Maintenance and rearing
The crickets used in this study originated from wild caught females from two
allopatric Australian populations (T. commodus – near Moss Vale, NSW and T. oceanicus
near Townsville, QLD). Colonies were bred in the lab for three generations before the
experiment began. Stock crickets were housed in 16-L plastic boxes of ca. 80 individuals in a
25 ⁰C temperature-controlled room on a 12:12 light:dark cycle. They were provided twice
weekly with ad libitum Burgess Excel “Junior and Dwarf” rabbit food and cotton wool pads
for drinking water and supplied with cardboard egg cartons for shelter.
Behavioural trials
No-choice mating experiments were conducted over two generations to investigate
patterns of sexual isolation between the species. To determine the strength and direction of
premating barriers, the first generation (F1) comprised heterospecific and conspecific pairs.
To test whether hybrids are sexually selected against, which could prevent inter-species gene
flow, reciprocal hybrid males and females were paired with individuals of each parental
species (backcrosses - BC1) and their courtship behaviours quantified.
Close range mating trials were conducted between 0900 and 1300 hours in a dim lit
room, illuminated by red light and temperature controlled (23⁰-25.5⁰). Virgin adult males and
females 10-20 days past adult eclosion were placed on opposite sides of a partition in a
mating arena (22x27cm) with a sand base. Pairs were given one minute acclimatization
period to adjust to the new environment before the partition was removed and the trial started.
Each trial lasted for ten minutes and if the pair made no physical contact within that period
92
they were retested the next day. If no physical contact was observed during the retrial, the
pair was removed from the analysis. The final dataset consisted of individuals that were only
tested once. Four courtship behaviours were scored as present or absent for each ten minute
trial thereby providing a proportional measure of courtship behaviour per cross type. The
amount of behaviours exhibited during a close range mating trial (count data) was also
examined and the results were qualitatively the same so I focus only on the proportional data
here. The behaviours scored tended to occur sequentially and were as follows; courtship song
(“song”), mounting, mating and spermatophore transfer (“transfer”). “Mounting” was defined
as when a female got up on a male, whereas “mating” was assigned when males initiated
jerking movements, necessary for transferring the spermatophore (Loher & Rence, 1978). I
distinguished these two behaviours as they are likely to reflect different processes, which
could independently contribute to sexual isolation. If a spermatophore was transferred to a
female, during the ten minute trial it was scored as successful. The pair were also maintained
together in the arena and observed every twenty minutes for the next hour to estimate the
duration of spermatophore retention (Simmons et al., 2003).
Table 3-1 Courtship trials: cross types and sample sizes for F1 crosses and backcrosses
The two letter codes indicate species identity; CC = T. commodus; OO = T. oceanicus, OC =
hybrid offspring from a cross with a female T. oceanicus and a male T. commodus, while CO
represents offspring from the reverse cross.
Crosses
n
Crosses
♀x♂
F1 Cross
Heterospecific
Conspecifc:
Total
OO x CC
CC x OO
OO x OO
CC x CC
n
♀x♂
30
30
30
30
120
Backcrosses
Hybrid male
backcross
Hybrid female
backcross
Total
93
OO x OC
CC x OC
OO x CO
CC x CO
OC x CC
CO x CC
OC x OO
CO x OO
20
20
20
20
20
20
20
20
160
Quantifying reproductive barriers
To quantify the relative strength of each mating behaviour to overall reproductive
isolation I calculated relative reproductive isolation (RRI) indices using a method devised by
Coyne & Orr, (1989) and adapted by Ramsey et al., (2003). This method compares the
proportion of pairs which exhibited a response (song, mounting, mating, spermatophore
transfer) in heterospecific crosses to that of conspecific crosses (similar to the approach in
Veen et al., (2012)).
The general form is:
𝑹𝑰 = 𝟏 −
𝐏𝐫𝐨𝐩𝐨𝐫𝐭𝐢𝐨𝐧 𝐨𝐟 𝐛𝐞𝐡𝐚𝐯𝐢𝐨𝐮𝐫 𝐞𝐱𝐩𝐫𝐞𝐬𝐬𝐞𝐝 𝐚𝐦𝐨𝐧𝐠 𝒉𝒆𝒕𝒆𝒓𝒐𝒔𝒑𝒆𝒄𝒊𝒇𝒊𝒄 𝒑𝒂𝒊𝒓𝒔
𝐏𝐫𝐨𝐩𝐨𝐫𝐭𝐢𝐨𝐧 𝐨𝐟 𝐛𝐞𝐡𝐚𝐯𝐢𝐨𝐮𝐫 𝐞𝐱𝐩𝐫𝐞𝐬𝐬𝐞𝐝 𝐚𝐦𝐨𝐧𝐠 𝒄𝒐𝒏𝒔𝒑𝒆𝒄𝒊𝒇𝒊𝒄 𝒑𝒂𝒊𝒓𝒔
(1)
RI quantifies the relative strength of a barrier, where a value of 0 indicates no barrier
and 1 represents a complete barrier. As earlier acting barriers have a greater potential to
prevent successful matings, the absolute contribution (AC) represents the contribution of a
barrier while accounting for the reduction due to earlier acting barriers (equations 1 – 4 in
Ramsey et al., (2003)).
AC1 = RI1
AC2 = RI2 (1- AC1)
AC3 = RI3 [1- (AC1 + AC2)]
And more generally:
𝑨𝑪𝒏 = 𝑹𝑰𝒏 (𝟏 − ∑𝒏−𝟏
𝒊=𝟏 𝑨𝑪𝒊 ) (2)
Cross types were classified by two letter codes. The first letter indicates the maternal
species and the second the paternal species (CC = T. commodus; OO = T. oceanicus, OC =
hybrid offspring from a female T. oceanicus and male T. commodus cross, while CO
represents offspring from the reverse cross) (Table 3-1). The experimental design was similar
94
across both generations. However, in the first generation male and female crickets had been
isolated for ten days prior to the courtship trials while in the second generation due to the
difficulty of isolating larger numbers of crickets, males and females were instead maintained
in sex segregated boxes of ca. 50 individuals prior to the trials. Therefore direct comparisons
across the two generations should be made with due caution.
Pattern of trait divergence
Parental species and reciprocal hybrid individuals were phenotyped for calling song,
courtship song and CHCs. The CHC profiles and calling song recordings were based on
individuals that were not used in the close range mating trials as close physical contact could
contaminate the target male’s CHC profile. In contrast, courtship song was recorded from
males used in the close range mating trials as its production is dependent upon close physical
contact. Both calling and courtship songs were recorded using a Seinheisser ME 66
microphone under a dim red light, in a temperature controlled room (22⁰-26⁰). The males
used were ca. 10-20 days post-adult eclosion. Calling songs were recorded from males
isolated in 118-mL plastic containers. Approximately one minute of song was recorded per
individual and five song phrases analysed to obtain an individual’s average. All songs were
analysed using Sony Sound Forge (7.0). Both spectral and temporal call elements were
measured and the same set of call variables were measured for both song types (Figure 3-1,
Figure 3-2 (below)).
95
Table 3-2 Sexual trait inheritance: Cross types and sample sizes
Song is a male-limited trait while cuticular hydrocarbons (CHCs) were extracted from both
males and females.
Trait
CC
CO
OC
OO
Calling Song
♀
-
♂
19
♀
-
♂
12
♀
-
♂
13
♀
-
♂
19
Courtship Song
-
16
8
-
13
10
10
8
-18
-
CHCs
-12
14
14
16
12
-
-
Figure 3-1. Calling song oscillograms: parental species and the reciprocal hybrids
Two oscillograms are shown for each hybrid type, to highlight the high variability among
individuals (adapted from Bailey & Macleod, 2013).
96
Figure 3-2 Courtship song oscillograms: parental species and the reciprocal hybrids
Two oscillograms per cross type are shown to illustrate the high degree of variability within
and between parental species and the hybrids (high standard errors in Table 3-6).
CHCs were extracted from virgin males and females from the parental species and
reciprocal hybrids ca. 10-20 days past adult eclosion. Individual crickets were isolated in
small (118-mL) plastic containers for ca. 10 days, provisioned with Burgess Excel “Junior
and Dwarf” rabbit food and water, before being anesthetized by chilling and then placed in
small glass extract vials (5ml) and stored at -20°C. Prior to CHC extractions, cricket samples
were thawed at room temperature, for ca. 10 minutes. 4ml of hexane was then pipetted into
each vial and left for 5 minutes before removing the cricket with clean forceps. 100µl of
extract was pipetted into 0.3ml fixed insert vials (Chromacol LTD, Item # 11573680) and left
to evaporate under a fume hood for ca. twelve hours. Samples were subsequently
97
reconstituted using 100µl of hexane containing 10 ng µL-1 dodecane as an internal standard.
Gas chromatography was performed using an Agilent 7890A GC coupled with an Agilent
5975C mass spectrometer, and an HP-5ms column (30m x 0.25mm x 0.25um). 2ul of extract
was injected into a multimode inlet operating in pulsed splitless mode, at a temperature of
250C. Hydrogen was used as the carrier gas at a column flow rate of 1ml/min. The oven
temperature profile was as follows: 70C for 1 min, ramping at 20C/min to 250C, then 4C/min
to 320C with a 5 minute hold. The MS transfer line was set at 280C. Mass spectra were
obtained at a scan range of 40-500 m/z. Data analysis was performed using Agilent MSD
Chemstation E.02.02.1431.
98
Figure 3-3 Typical chromatogram of the two parental species
Example of species-typical CHC profiles. Unfortunately chromatograms for hybrid profiles
were unavailable.
99
Statistical analysis
Courtship behaviour trial data were examined using Generalized linear models (GLM)
with a binomial distribution (quasibinomial was used for both generations of song and all
traits in the second generation due to overdispersion of the data). Explanatory variables
included: the focal individual’s species identity (i.e. identity of the male species for courtship
“Song” or female species for the other behaviours) and cross type (e.g. conspecific or
heterospecific) which were fitted as fixed effects. Temperature and the weight of individuals
were included as covariates. For the second generation “mating type” which refers to the
mating pair’s identity (e.g. OO x CC), was fitted as a fixed effect,. The decision to include or
remove a variable from the model was made based on comparison of the model fit using
ANOVAs and chi squared distributions (α < 0.05). To determine if groups of interest
differed, Tukey pairwise comparisons were implemented. All statistical analyses were
performed in R (Version 3.1.3).
Due to the multivariate nature of each of the sexual traits, divergence patterns were
examined with principal component analysis (PCA) using the R package FactoMiner (Lê et
al., 2008). To control for temperature differences between song recordings, temperaturecorrected residuals were used in the PCA. Prior to CHC analysis the data were standardized
by dividing the abundance of each peak by the internal standard (10 ng µL-1 dodecane) and
normalized by a log10 transformation. Histograms of the individual principal component
scores (PC1 vs. PC2) were used to visually examine the pattern of divergence. One way
ANOVAs and post-hoc pairwise comparisons (glht function) on the scores were used to test
the degree of divergence between the main groups of interest (species, sex, or species*sex)
using the R multcomp package (Hothorn et al., 2008).
100
X-linkage
The influence of X chromosomes on interspecific differences for male sexual traits
(the heterogametic sex) was estimated based on the mean trait differences between F1
reciprocal hybrid crosses divided by the difference between the two parental lines (Reinhold,
1998; Oh et al., 2012).
𝐼𝑥 =
𝐶𝐴𝑥𝐵 −𝐶𝐵𝑥𝐴
𝐶𝐴 −𝐶𝐵
(3)
Where C gives the trait mean in the male of the parental lines A and B and the
reciprocal hybrids (𝐶𝐴𝑥𝐵 , 𝐶𝐵𝑥𝐴 ). The normalized index of Ix allows for comparisons of the
degree of X-linkage for different traits (Reinhold, 1998). An Ix of 0 indicates no X-linkage of
the trait in question, while an Ix of 0.5 reflects half of the phenotypic difference between the
parental lines being caused by X-linked genes (or maternal effects), while a value of 1 is
consistent with all of the differences being due to the maternal species. I considered only
cases where the reciprocal hybrid types were significantly different (Tukey pairwise
comparisons) as providing support for sex-linkage. In addition to examining the multivariate
trait means, I also tested for sex-linkage of the univariate trait trill interval (“Trill
1.2.interval”) as this component was previously suggested to be sex-linked (Bentley & Hoy,
1972).
Transgressive segregation
F1 hybrid phenotypes are expected to be intermediate to that of the parental species if
most gene interactions are additive. The extent of departure from intermediacy due to
directional dominance (heterosis) can be estimated by the difference between the mean F1
101
trait value (𝑀𝐹1 ) and the mid-parent mean value (𝑀𝑃̅ ) (Falconer & Mackay, 1996; p 255).
The estimate of heterosis can be positive or negative. The term heterosis is often used in
context of increased hybrid fitness, but here I use it solely to refer to deviations from the
midparent value in hybrids, irrespective of hybrid fitness.
The mid-parent value is:
𝑀𝑃̅ =
1
2
(𝑀𝑃1 + 𝑀𝑃2 )
The amount of heterosis is:
𝐻𝐹1 = (𝑀𝐹1 − 𝑀𝑃̅ ) (4)
In line with Rieseberg et al., (1999) traits were defined as transgressive if the
character means lay outside the means of both parental species (in a negative or positive
direction). Traits were classified as dominant if they exhibited a significant difference (Tukey
pairwise contrasts) from one of the parental species but were undistinguishable from the
other. Otherwise the traits were classified as intermediate.
102
Results
Interspecific sexual isolation is symmetrical
A large proportion of males started singing when paired with a conspecific female
(OO - 0.56, CC - 0.72), but only a small proportion of pairs successfully transferred
spermatophores (OO - 0.12, CC - 0.19) during the ten minute trial period. Heterospecific
pairs had a lower tendency to exhibit courtship behaviours compared to conspecific pairs
(Figure 3-4; Table 3-4). The strongest difference between heterospecific and conspecific pairs
was for courtship song, in particular T. oceanicus males paired with heterospecific females
exhibited a marked reduction in the tendency to produce courtship song (Tukey comparisons:
OO x OO vs. OO x CC – p = 0.005; CC x OO vs. CC x CC – p = 0.435) (Table 3-3;Table
3-4). Females of both species exhibited a similar tendency to mount or mate with
heterospecific males at a lower rate than with conspecific partners. Overall both species
exhibited similar responses, indicated by a non-significant interaction between cross type and
male or female identity (Table 3-4). Behavioural isolation was almost complete between the
parental species with a total RRI for CC x OO crosses = 0.993, and OO x CC = 1.
103
Figure 3-4 Close range courtship trial results: conspecific and heterospecific pairs
Proportion of trials in which courtship behaviours were observed. Symbols indicate the focal
individual species identity (CC = T. commodus, OO = T.oceanicus). The broken line indicates
that “Song” is a male sex-limited trait.
Table 3-3 Reproductive isolation index
Components of reproductive isolation (Eq. 1) and the absolute contribution (AC) (Eq. 2). AC
measures the contribution of each reproductive barrier while accounting for the reduction due
to earlier acting barriers. The proportion of pairs exhibiting each behaviour are compared
between conspecific and heterospecific pairs (CC = T. commodus; OO - T.oceanicus). Song
is expressed only by males and therefore the conspecific reference is the paternal species,
while the other behaviours are female controlled, therefore the reference is the maternal
species.
Barrier
Song
Mount
Mate
Transfer
Total
contribution
Components of
Absolute contribution
reproductive isolation (RI)
to reproductive isolation (AC)
OO x CC
(♀ x ♂)
CC x OO
(♀ x ♂)
OO x CC
(♀ x ♂)
CC x OO
(♀ x ♂)
0.224
0.838
0.811
0.716
0.761
0.867
1
1
0.224
0.650
0.102
0.017
0.993
0.761
0.207
0.032
0
1
104
Table 3-4 GLM results courtship trials amongst conspecific and heterospecific pairs
Generalized linear model results examining the frequency (binomial distribution) of four
courtship behaviours (Song, Mount, Mate, Transfer) amongst conspecific and heterospecific
pairs (Cross type). After model selection the final model included significant or nearly
significant effects which are highlighted in bold. F-tests were used for song as quasibinomial
models were fitted, while chi2 tests were fitted for the other binomial traits. To test for “cross
type” differences, Tukey mean contrasts (glht) were used and p-values are shown. Significant
comparisons are highlighted by asterisks (p >0.01=*, p >0.001=**, p <0.001=***). Full
pairwise model results provided in Supplementary section (Table S2.)
Trait
Model
selection
df
F
p
Cross type
Pairwise
Comparisons
p
<0.0001***
0.002**
CC x OO - CC x CC
OO x CC - CC x CC
OO x OO - CC x CC
OO x CC - CC x OO
OO x OO - CC x OO
OO x OO - OO x CC
0.450
0.001***
0.547
0.012*
0.997
0.006**
CC x OO - CC x CC
OO x CC - CC x CC
OO x OO - CC x CC
OO x CC - CC x OO
OO x OO - CC x OO
OO x OO - OO x CC
0.133
0.857
1.000
0.768
0.135
0.836
Song
Cross type
Male identity
1,124 11.503
1,123 10.332
Cross type
Male weight
df
Chi2
1,124 90.647
1,123 87.257
Cross type
Male weight
1,124 80.919
1,123 77.102
<0.001**
0.051
CC x OO - CC x CC
OO x CC - CC x CC
OO x OO - CC x CC
OO x CC - CC x OO
OO x OO - CC x OO
OO x OO - OO x CC
0.108
1.000
0.991
1.000
0.178
1.000
Cross type
Male weight
1,124 66.315
1,123 61.267
0.004**
0.025*
CC x OO - CC x CC
OO x CC - CC x CC
OO x OO - CC x CC
OO x CC - CC x OO
OO x OO - CC x OO
OO x OO - OO x CC
0.316
1.000
0.999
1.000
0.409
1.000
Mount
p
<0.001***
0.066
Mate
Transfer
105
Hybrids are not behaviourally sterile
The backcrosses with hybrid males and females all exhibited relatively high rates of
courtship behaviour (Figure 3-5). A large proportion of males produced courtship song,
irrespective of whether they were a hybrid male (ranging from 0.5 - 0.8) or a male of either
parental species (0.65 - 0.9). Hybrid females mounted (0.45 – 0.7) and mated (0.2 – 0.35) at a
relatively high rate. None of the factors, including mating type, predicted the frequency of
courtship behaviours (Table 3-5). Pairwise comparisons amongst the parental species (CC &
OO) and reciprocal hybrid pairs (OC & CO) revealed no significant differences (all p > 0.05).
This may partly reflect a higher variance in the courtship behaviours amongst hybrids. The
proportion of pairs exhibiting courtship behaviours were higher than those observed among
the conspecific crosses in the first generation of crosses (Figure 3-4; Figure 3-5). This may
reflect an increased propensity for hybrids to mate (i.e. hybrid vigour). However,
experimental differences in the rearing of the crickets between the two generations prevent a
formal comparison. Overall the results indicate that hybrid males and females are both
capable of eliciting and enacting courtship behaviour, even though hybrid females are sterile
(Chapter 4).
106
Figure 3-5 Close range courtship trial results: backcrosses with hybrid males and
females
Proportion of trials in which courtship behaviours were observed amongst backcross pairs.
Symbols indicate the female species identity and the X-axis indicates the male’s identity (CC
= T. commodus; OO = T. oceanicus, OC = hybrid offspring from female T. oceanicus and
male T. commodus crosses, CO= reverse cross). “Song” is distinguished by dotted lines as it
is a male sex limited trait while the other behaviours are principally under female control and
therefore reflect female identity.
Table 3-5 GLM results courtship trials amongst backcross pairs
Generalized linear model results examining the frequency (binomial distribution) of the four
courtship behaviours (Song, Mount, Mate, Transfer) amongst mating types. All behaviours
were found to be overdispersed and quasibinomial models were fitted and F-tests
implemented. None of the factors examined were significant.
Trait
Model
df
F
p
Song
Mating type
7, 151
1.385
0.215
Mount
Mating type
7, 151
0.980
0.448
Mate
Mating type
7, 151
0.396
0.904
Transfer
Mating type
7, 151
1.26
0.274
107
Patterns of trait divergence
Calling song
Multivariate analysis on the scores from the first four principal components
(eigenvalues > 1), which account for 81.5% of the total variation, revealed significant
differences in calling song among the four different cross types (MANOVA: Wilks λ=0.029,
F3,59 =35.077, p < 0.0001). The first principal component (PC1) was dominated by trill and
frequency components while the second component (PC2) contrasted song length with
number of pulses per trill (Figure 3-6; Table S3). PC1, which accounts for 42.02% of the
variation, clearly separated the parental species (Tukey comparisons: OO vs. CC, p <0.001;
CO vs. OC, p = 0.836) while PC2 (20.3% of the variation) predominantly differentiated the
reciprocal hybrids (OO vs. CC, p = 0.220; CO vs. OC, p = 0.009) (Figure 3-6, Table S6).
Hybrid males from both cross directions were intermediate between the parental
species for PC1 (Figure 3-6) and distinguishable from the parental species (Tukey
comparisons: CC vs. both hybrid types, p < 0.001; OO vs. both hybrid types, p < 0.001)
(Table S6). For PC2, hybrids with a T. oceanicus X chromosome (OC) exhibited more
extreme trait values than either of the parental species (Tukey comparisons: OC vs. CC, p =
0.004; OC vs. OO, p < 0.001) or even to the hybrids from the other cross direction (OC vs.
CO, p = 0.009) (Table 3-6).
Applying the Ix index (Reinhold, 1998), to estimate the influence of X-linked genes on
the parental species differences, I found no support for sex-linkage of PC1 song components
(IX = -0.05). For PC2, there was a large difference between the reciprocal hybrids, CO
hybrids were most similar to their maternal species, however OC hybrids were most
divergent from their maternal species (IX = -1.75). For the univariate song trait, inter trill
interval, I found no evidence of sex-linkage (IX = -0.084). Instead inter trill interval was
108
almost perfectly intermediate between the range of both parental species (HF1: OC -0.006;
CO -0.063) (Fig. S1).
Figure 3-6 Calling song inheritance pattern
Calling song, PC1 and PC2 histograms for the two parental species (CC – T. commodus; OO
– T.oceanicus) and the reciprocal hybrids (CO – male hybrids with a T. commodus X
chromosome; OC – male hybrids with a T. oceanicus X chromosome). The red dotted line
indicates the group mean.
Courtship song
Multivariate analysis on the scores from the first four principal components
(eigenvalues > 1), which account for 76.23% of the total variation, revealed significant
differences in courtship song among the four different cross types (MANOVA: using “cross
type” as a fixed effect, Wilks λ=0.147, F3,41 =6.684, p < 0.0001). PC1 was dominated by song
length elements and pulse durations, while PC2 contrasted frequency and number of trills
against chirp elements (Figure 3-2, Table.S4). PC1, which accounts for 31.96% of the total
variance, was significantly different between the parental species (Tukey comparisons; CC
109
vs. OO, p = 0.001; Figure 3-7, Table S6) but most of the variation in PC1 was due to
differences between both hybrid types (Tukey comparison: CO vs. OC, p = 0.006) (Figure
3-7; Table S6). In particular, hybrid males with a T. commodus X chromosome (CO)
exhibited a mean trait value more extreme than either of the parental species (𝐻𝐹1 : 2.403),
while OC males were intermediate (Figure 3-7, Table 3-6). The greatest group difference was
between CO males and the paternal species T. oceanicus (Tukey comparisons; CO vs. OO: p
< 0.001) (Table S6). Applying the Ix index, revealed strong support for sex-linkage (IX = 1.16)
for song elements associated with PC1 (Table 3-6).
For PC2, the parental species were distinguishable (Tukey comparisons; CC vs. OO, p
= 0.004; Table S6) and the reciprocal hybrids exhibited an intermediate pattern between that
of both parental species (𝐻𝐹1 : OC -0.274, CO 0.024; Figure 3-7). The reciprocal hybrid
crosses were not significantly different (Tukey comparisons; CO vs. OC: p = 0.972) (Table
S6) and the IX index was low indicating that the parental species differences are not due to Xlinked genes (IX = -0.159) (Table 3-6).
110
Figure 3-7 Courtship song inheritance pattern
Courtship song, PC1 and PC2 histograms for the two parental species and the reciprocal
hybrids. The red dotted line indicates the group mean.
Overall both song types exhibited clear species differences and are divergent between
the hybrids which supports the potential role for these traits in species discrimination.
Multivariate song components exhibited predominantly intermediate patterns in the hybrids,
however there is also an indication of transgressive segregation and sex-linkage for some
song elements (PC2 calling song, PC1 courtship song).
Cuticular hydrocarbons (CHCs)
Multivariate analysis on the scores from the first twelve principal components
(eigenvalues > 1), which accounted for 75.43% of the total variation in the CHC data,
revealed significant differences amongst the sexes (MANOVA: Wilks λ=0.376, F1,104 =
12.82, p < 0.0001), cross type (Wilks λ=0.048, F3,102 = 13.496, p < 0.0001) and cross type
subdivided by sex (Wilks λ= 0.003, F7,98 = 9.849, p < 0.0001). Most of the variation in the
111
CHC data was accounted for by the parental species differences on PC1 (25.68%) and by sex
(13.16%) on PC2 (Figure 3-8). Males and females of the parental species were clearly
distinguishable from each other by PC1 (Tukey comparisons: OO_F vs. CC_F, p < 0.001;
OO_M vs. CC_M, p < 0.001) (Table 3-6, Table S7). The extent of sexual dimorphism
differed between the parental species, being higher in T.oceanicus. Male and female T.
oceanicus were distinguishable on PC1 (Tukey comparisons: OO_M vs. OO_F, p = 0.007)
but not PC2 (p = 0.656) (Table S7), while T. commodus males and females (CC_M vs.
CC_F) were indistinguishable from each other for either PC1 (p = 0.999) or PC2 (0.503)
(Figure 3-8, Table S7).
Hybrids did not exhibit an intermediate pattern for PC1, as expected for a polygenic
additive trait, instead hybrid CHC profiles exhibited dominance, sharing a strong similarity
with the parental species T. commodus (Figure 3-8). Tukey pairwise comparisons revealed
that both male and female hybrids were significantly different from T. oceanicus individuals
(Tukey comparisons: p < 0.01; Table S7) but indistinguishable from T. commodus individuals
(p > 0.05). Applying the Ix index for the hybrid males (the heterogametic sex), to estimate the
influence of X-linked genes on the parental species differences for PC1, revealed no strong
support for sex-linkage of CHC components associated with PC1 (IX = -0.088) (Table 3-6).
For PC2 individuals exhibited large sex differences, even within the same hybrid
crosses (Figure 3-8, Table S7). Hybrids appeared to exhibit extreme mean profiles outside the
range of the parental species (Figure 3-8, Table 3-6). However, the hybrid variances were
large and the differences between hybrids and the parental species were not signficant for
either hybrid males or females (Table S7).
112
Figure 3-8 CHCs inheritance pattern
PC1 and PC2 histograms for the two parental species and the reciprocal hybrids. “F”
indicates females and “M” males. The red dotted line indicates the group mean.
113
Table 3-6 Summary data sexual trait inheritance pattern
Summary data for PC1 and PC2 from the three sexual traits among parental species and the
reciprocal hybrids; trait means ± SE, estimate of heterosis (HF1) and estimate of sex-linkage
IX (described in methods section).
Trait
Calling Song
Species
ID
PC1
(mean ± SE)
CC
CO
OC
OO
2.995 ± 0.217
-0.269 ± 0.231
0.026 ± 0.155
-2.843 ± 0.234
HF1
IX
PC2
(mean ± SE)
0.093 ± 0.346
-0.157 ± 0.420
-1.734 ± 0.424
0.994 ± 0.288
-0.345
-0.05
HF1
-0.701
-2.278
-0.05
Courtship Song
CC
CO
OC
OO
0.319 ± 0.339
2.236 ± 0.994
-0.248 ± 0.438
-1.615 ± 0.333
-1.75
-1.637 ± 0.321
1.165 ± 0.994
0.797 ± 0.322
0.807 ± 0.388
2.884
0.4
1.58
1.212
1.162
CHCs
CC_F
CC_M
CO_F
CO_M
OC_F
OC_M
OO_F
OO_M
-2.266 ± 0.313
-2.516 ± 0.182
0.159 ± 0.771
-2.897 ± 0.195
-0.922 ± 0.344
-2.09 ± 0.536
3.422 ± 0.936
6.674 ± 0.716
-0.088
114
-0.159
-0.407 ± 0.566
1.412 ± 0.220
-2.182 ± 0.668
1.217 ± 0.747
-3.051 ± 0.552
2.809 ± 0.505
-0.964 ± 0.57
0.411 ± 0.138
-0.419
-4.976
-1.5
-2.09
IX
-1.5
0.306
-2.366
1.898
1.59
Discussion
The genetic and behavioural basis to species differences remains an elusive but
important area of speciation research (Coyne & Orr, 2004). The pattern of inheritance and
expression of primary and secondary sexual traits in hybrids can have important
consequences for the extent and direction of gene flow (Arnold et al., 2012). In this study I
identified species differences in key sexual phenotypes and multiple behavioural barriers
during close range courtship which may play an important role in curtailing gene flow
between the species. I discuss the role each of these traits may play in reproductive isolation,
their pattern of divergence amongst the species and hybrids, and some of the constraints of
this study.
Symmetry of behavioural isolation
In this study females of the parental species exhibited a similar tendency to
discriminate against heterospecific partners. How consistent is symetrical species
discrimination and can it inform us about the evolution of species differences leading to
behavioural isolation? Behavioural isolation is often asymmetric, having been observed in a
number of taxa; Drosophila (Watanabe & Kawanishi, 1979; Kaneshiro, 1980; Yukilevich,
2012; Arthur & Dyer, 2015), damselflies (Svensson et al., 2007; Sánchez-Guillén et al., 2012;
Svensson et al., 2016), parasitic wasps (Bordenstein et al., 2000), crickets (Veen et al., 2012),
salamanders (Arnold et al., 1996) and snakes (Shine et al., 2002). Among the true crickets
(Gryllidae) a literature review of reproductive barriers by Veen et al., (2012) highlighted that
19% (3 out of 16 cases) of species pairs exhibited asymmetrical prezygotic barriers while
31% displayed asymmetrical postzygotic barriers. Similarly a review by Lowry et al., (2008),
115
involving 19 species pairs of plant taxa, found that asymmetries in prezygotic barriers
occurred three times less than for postzygotic barriers. Prezygotic barriers may be expected to
exhibit less consistent patterns due to the complexity of traits involved, whose development,
expression and fitness is often dependent on the environment (Grant & Grant, 1993; Pfennig,
2007), in contrast to intrinsic genetic incompatibilities which may be more functionally
constrained (Turelli & Moyle, 2007). Further work is needed to compare taxonomic groups to
determine whether asymmetries in prezygotic barriers may be more common in certain
groups than others.
Gender differences in species discrimination
Gender differences in mating behaviour may indicate divergent selection pressures
between the sexes (Svensson et al., 2007). Females are generally expected to be the more
discriminating sex due to predicted higher fitness costs for mating with incompatible mates
(Anderson, 1994). The results highlight an important distinction between the species in the
strength of male and female discrimination. Male T. oceanicus discriminated against
heterospecific females upon contact, indicated by the females’ failure to elicit courtship song,
while male T. commodus did not exhibit as strong a reduction in singing behaviour when
paired with a heterospecific female (Figure 3-4). Instead, the main barrier preventing T.
commodus males mating with heterospecific females is due to females declining to mount
(Table 3-3).
A growing body of research indicates that sex roles during mate choice can be highly
dynamic due to differences in the relative costs of mating (reviewed in Gwynne, 1991;
Bonduriansky, 2001). In these cricket species males do not provide post copulatory gifts or
parental care which have been reported to drive increased male choosiness (Gwynne, 1981,
116
1985; Gwynne & Simmons, 1990; Edward & Chapman, 2011). Instead high investment in
costly courtship behaviours, which expend energy and attract predators, may have increased
the costs of mate attraction, encouraging male choosiness in T. oceanicus. Shorter durations
of courtship song were observed to be sufficient for T. oceanicus males to elicit a response
from conspecific females in comparison to T. commodus pairs. Differences in the acoustic
behaviour of the two species may reflect differing selective pressures, such as higher fitness
costs of courtship song (Hack, 1998) or alternative mating strategies.
T. oceanicus males are likely to discriminate against heterospecific females based on
their divergent CHC profiles. The increased sexual dimorphism in CHC profiles amongst T.
oceanicus (Figure 3-8) may indicate an important role for this trait in mate choice (Thomas &
Simmons, 2009b). Indeed, T. oceanicus males have been shown to assess female
attractiveness based on CHCs and the strength of male selection on female CHCs is an order
of magnitude stronger than female selection on male CHCs (Thomas & Simmons, 2009b). In
addition, T. oceanicus males can assess sperm competition, by detecting CHCs left on a
female by other males, and adjust their sperm allocation accordingly (Thomas & Simmons,
2009a).
The relative reproductive isolation index (RRI) suggested behavioural isolation was
almost complete among the parental species (Total RRI = T. commodus x T. oceanicus =
0.993, and T. oceanicus x T. commodus = 1). This highlights the importance of close range
mating behaviours in reducing the potential for interspecies gene flow. However, the low rate
of successful conspecific matings, in which a spermatophore was transferred (T. oceanicus
12%; T. commodus 19%), may reflect the short duration for the mating trials (10mins) and
therefore overestimate the contribution of behavioural barriers to reproductive isolation. In
117
contrast, the use of no choice mate trials and male and female virgins, with no previous
mating experience, may overestimate the propensity for heterospecific matings (Dougherty &
Shuker, 2015). The optimal experimental design to estimate the strength of mate choice
depends upon the social and ecological conditions under which mate choice is exerted in
natural populations. Little is known about mating behaviour in natural populations of these
Teleogryllus species. Social effects are likely to be important, as both species are highly
polyandrous (Simmons & Beveridge, 2010a) and prior mating experience can influence
patterns of mate choice (Bailey & Zuk, 2008; Bailey, 2011; Bailey & Macleod, 2013).
Hybrid behavioural sterility
If hybrids are viable and fertile their mating behaviour, in particular which parental
species they mate with, will determine the strength and direction of gene flow. In this study
both hybrid males and females elicited and enacted courtship behaviours at a relatively high
rate (Figure 3-5), even though hybrid females are sterile (Chapter 4). The genetic factors
which disrupt hybrid female fertility do not appear to negatively affect secondary sexual trait
development or expression (Davies et al., 1997). Hybrid males and females exhibited the
same propensity to mate with partners of either parental species (Figure 3-5). The frequency
of hybrid courtship behaviours was higher compared to conspecific pairs in the first
experiment. However caution is needed in interpreting this, as differences in the experimental
design (see methods) limit the ability to determine if this reflects hybrid vigour in mating
success.
118
Behavioural isolation and sex chromosomes
Genes underlying sexually selected traits are predicted to accumulate on sex
chromosomes (Charlesworth et al., 1987) but empirical results supporting this have been
mixed (Reinhold, 1998; Qvarnström & Bailey, 2009). Multivariate analysis of the sexual
signals revealed patterns of inheritance consistent with sex-linkage for three out of six
components (PCs) studied (IX > 0.5; Calling song – PC2; Courtship song - PC1, CHCs –
PC2). All three cases exhibited phenotypic means outside the range of the parental species,
which suggests transgressive segregation in addition to sex-linkage. For courtship song and
CHC components the hybrid crosses diverged in the direction of their maternal species, as
expected for X-linked traits. However, for calling song components associated with PC1, OC
individuals were most dissimilar to their maternal species T. oceanicus. This makes it
difficult to distinguish the effects of X-linkage from nonsymmetric heterosis (Reinhold,
1998). The differences observed amongst the reciprocal hybrids could be due to sex-linkage
or maternal effects. To properly distinguish these confounding effects, one needs to examine
trait segregation over multiple generations (Butlin & Ritchie, 1989).
Dominance and transgressive segregation
Hybridization can lead to intermediate, dominant or even extreme phenotypes in
hybrids which lie outside the range of those in both parental species (transgressive
segregation) (Rieseberg et al., 1999; Lynch & Walsh, 1999). In agreement with expectations
of song being a polygenic and additive trait, calling song and courtship song components
were predominantly intermediate among hybrids. However, segregation of some song
components indicated transgressive segregation: PC2 in calling song (OC hybrids: Fig. 3-6)
and PC1 for courtship song (CO hybrids: Fig. 3-7). In contrast, the CHCs primarily exhibited
dominance and transgressive segregation. This apparent distinction between the two signal
119
modalities may reflect differences in genetic architectures. Previous studies have suggested
many genes of small effect predominate acoustic communication systems while a few genes
of major effect underlie chemical communication (Ritchie & Phillips, 1998; Shaw et al.,
2011; Ellison et al., 2011)
Transgressive segregation may be frequent, resulting in novel mating signals and
preferences, with important evolutionary consequences such as adaptive hybridization and
even hybrid speciation (Rieseberg et al., 1999; Seehausen, 2004; Jiggins et al., 2008; Abbott
et al., 2013; Rosenthal, 2013). A review by Rieseberg et al., (1999) examining trait
segregation among plant and animal hybrids revealed that transgressive segregation occurred
in 78% of animals studied (45 of 58 cases) and 97% for plants (110 of 113 studies). Focusing
just on wild outbred populations, which are more informative for understanding natural
patterns of hybridization, revealed that extreme phenotypes occurred in 23% of animal taxa
and 36% for plants, suggesting transgressive segregation may be an important but
underappreciated component in hybrid zone dynamics (Arnold et al., 1999; Abbott et al.,
2013).
What factors can predict the occurrence of transgressive segregation? Transgressive
hybrid phenotypes are largely due to complementary gene action (in each of the parental
species if multiple loci have opposing effects on the trait when recombined in hybrid
individuals they can produce complementary effects leading to extreme hybrid phenotypes)
and sometimes overdominance and epistasis (Rieseberg et al., 2003). The genetic distance
between species has been found to predict the occurrence of transgressive segregation as the
number of quantitative trait loci fixed for different alleles in different species should increase
with the time since species divergence (Stelkens & Seehausen, 2009). The history of
selection acting on the trait is also likely to determine the occurrence of transgressive
segregation. Traits with a history of directional selection may be less likely to exhibit
120
transgressive segregation than those with a history of stabilizing selection, as directional
selection will lead to the fixation of alleles whose effects are in opposing directions, which in
hybrids will cancel out producing intermediate trait patterns (Rieseberg et al., 1999).
Female preferences for calling song in T. commodus have been found to exert
stabilizing selection on chirp and trill elements (Brooks et al., 2005; Bentsen et al., 2006).
Sexual selection on courtship song in T. oceanicus is directional (but nonlinear), favouring
males with the highest sound content (Simmons et al., 2013). In contrast sexual selection in
T. commodus exerts stabilizing effects on both chirp and trill elements of the courtship song
(Hall et al., 2010). Female preferences for male CHC profiles in T. oceanicus have been
found to be largely under disruptive sexual selection ( Thomas & Simmons, 2009b) but also
some peaks are under stabilizing selection (Simmons et al., 2013). Future studies would
benefit from comparing patterns of trait segregation and the form of selection on the trait in
the parental species (Rieseberg et al., 1999).
Conclusions
Close range behavioural barriers are likely to play a key role in species isolation. In
this study heterospecific pairs exhibited reduced rates of courtship behaviour indicating a
potentially important role for behavioural isolation in maintaining the species integrity.
Behavioural isolation was almost complete between the species. Surprisingly, there was no
evidence for sexual selection against hybrids. Instead, hybrids appeared to enact and elicit
relatively high rates of courtship which would facilitate gene flow. However, the apparent
absence of hybrids among sympatric populations (Chapter 5), suggests pre-mating barriers
may be sufficiently strong to prevent hybridization. Even though components of the signal
traits exhibited a greater similarity to one or the other of the parental species, in particular the
CHC profiles, there was no corresponding asymmetry in mate choice in the behavioural
121
trials. The patterns of inheritance for song traits (i.e. primarily intermediate patterns) differed
to that of the CHC profiles (exhibited dominance) which may indicate different underlying
genetic architectures. As the crickets used in this study originated from allopatric populations
the ability to predict the strength and form of sexual isolation in sympatric populations is
constrained. As is common in speciation research it is difficult to determine whether these
behavioural barriers may have preceded the species secondary contact. However, differences
between allopatric populations support a role for intraspecific processes, such as sexual
selection in driving the divergence of these sexual traits. Overall the results support the view
that close range courtship behaviours may be just as important as long range signals in mate
choice and reproductive isolation.
122
Supplementary
Table S1. Cross types and the number of pairs (n) shown for both sets of courtship trials. The
percentage of pairs to exhibit a response for each of the four behaviours are shown.
Cross Type
F1 Cross
Heterospecific
Conspecifc:
Total
Backcrosses
Hybrid male
backcross
Hybrid female
backcross
Total
Percentage of pairs exhibiting a response
Song
Mount
Mate
Transfer
Crosses ♂ n
x♀
OO x CC
CC x OO
OO x OO
CC x CC
30
30
30
30
120
13.3
53.3
55.6
71.8
3.3
3.3
20.6
28.1
0
3.3
17.6
28.1
0
3.3
11.8
18.8
OC x OO
OC x CC
CO x OO
CO x CC
CC x OC
CC x CO
OO x OC
OO x CO
20
20
20
20
20
20
20
20
160
80
50
80
70
89.5
80
75
65
35
40
60
55
52.6
45
70
55
35
30
45
30
21
30
35
35
15
10
20
0
5.3
10
5
5
123
Table S2. Pairwise comparisons of the mating types for the four courtship behaviours, using
Tukey mean contrasts and based on the fit of a binomial GLM containing “Mating type” as
an explanatory variable and male weight if significant.
Song
Pairwise
comparison
CC x OO - CC x CC
OO x CC - CC x CC
OO x OO - CC x CC
OO x CC - CC x OO
OO x OO - CC x OO
OO x OO - OO x CC
Estimate
-0.805
-2.810
-0.702
-2.005
0.103
2.108
CC x OO - CC x CC
OO x CC - CC x CC
OO x OO - CC x CC
OO x CC - CC x OO
OO x OO - CC x OO
OO x OO - OO x CC
Estimate
-2.452
-17.263
-0.182
-14.811
2.270
17.081
CC x OO - CC x CC
OO x CC - CC x CC
OO x OO - CC x CC
OO x CC - CC x OO
OO x OO - CC x OO
OO x OO - OO x CC
Estimate
-2.441
-0.957
0.064
1.483
2.504
1.021
CC x OO - CC x CC
OO x CC - CC x CC
OO x OO - CC x CC
OO x CC - CC x OO
OO x OO - CC x OO
OO x OO - OO x CC
Estimate
-1.888
-16.852
-0.108
-14.964
1.780
16.744
Std. Error
0.546
0.676
0.532
0.661
0.511
0.649
Mate
Std. Error
1.146
1765.431
0.667
1765.431
1.188
1765.431
Mount
Error
value
-1.474
-4.154
-1.320
-3.036
0.201
3.249
Pr(>|z|)
0.450
0.001
0.547
0.012
0.997
0.006
Error
value
-2.140
-0.010
-0.273
-0.008
1.912
0.010
Pr(>|z|)
0.108
1.000
0.991
1.000
0.178
1.000
Error
Std. Error
value
1.145
-2.132
1.232
-0.777
0.658
0.097
1.567
0.946
1.180
2.123
1.245
0.820
Transfer
Error
Std. Error
value
1.171
-1.612
1790.023
-0.009
0.763
-0.141
1790.024
-0.008
1.226
1.452
1790.023
0.009
124
Pr(>|z|)
0.133
0.857
1.000
0.768
0.135
0.836
Pr(>|z|)
0.316
1.000
0.999
1.000
0.409
1.000
0.867
0.864
0.823
0.652
0.481
0.419
0.379
-0.260
-0.390
-0.607
-0.646
-0.655
-0.689
-0.907
PC1
5.883
42.018
42.018
Average.Pulses.per.Trill
Intersong.Interval
Chirp.Number
Chirp.Duration
Total.Duration
Trill.Total.Duration
Chirp.Interpulse.Interval
Trill.Interpulse.Interval
Chirp.Pulse.Length
Trill.Pulse.Duration
Chirp.Trill.Interval
Trill.Number
Frequency
Variance
% of var.
Cumulative % of var.
correlation
PC1
Trill.1.2.interval
percentage of variance is also shown.
Average.Pulses.per.Trill
Chirp.Number
Chirp.Trill.Interval
Trill.Interpulse.Interval
Chirp.Duration
Chirp.Interpulse.Interval
Trill.Number
Trill.Total.Duration
PC2
Total.Duration
125
62.299
2.839
20.281
PC2
-0.287
0.380
0.382
0.474
0.540
0.574
0.623
0.717
0.802
correlation
Trill.Total.Duration
Chirp.Interpulse.Interval
Total.Duration
Average.Pulses.per.Trill
Chirp.Trill.Interval
Chirp.Number
PC3
Chirp.Duration
73.095
1.511
10.796
PC3
-0.446
-0.362
-0.261
-0.261
0.418
0.613
0.639
correlation
Table S3. Calling song components significantly correlated with the principal components (p < 0.05) on the first three PCs. Eigenvalues and the
0.839
0.813
0.653
0.526
0.506
0.501
0.303
-0.512
-0.575
-0.676
3.835
31.964
31.963
Total.Duration
Chirp.Interpulse.Interval
Trill.Number
Chirp.Trill.Interval
Average.Pulses.per.Trill
Chirp.Number
Frequency
Trill.Pulse.Duration
Chirp.Pulse.Length
Eigenvalue
% of var.
Cumulative %
Chirp.Number
Chirp.Duration
Average.Pulses.per.Trill
Chirp.Interpulse.Interval
male_weight
Trill.Pulse.Duration
Trill.Total.Duration
Total.Duration
Trill.Number
Frequency
correlation PC2
Trill.Total.Duration
PC1
the percentage of variance is also shown.
126
50.931
18.968
2.276
-0.693
-0.538
-0.386
0.297
0.381
0.393
0.402
0.416
0.567
0.598
Frequency
Chirp.Interpulse.Interval
Chirp.Number
Trill.Pulse.Duration
Trill.Interpulse.Interval
Chirp.Duration
correlation PC3
66.84
1.909
15.909
0.322
0.341
0.498
0.499
0.709
0.768
correlation
Table S4. Courtship song components significantly correlated with the principal components (p < 0.05) on the first three PCs. Eigenvalues and
Table S5. CHCs significantly correlated with the principal components (p < 0.05)
PC1
C35
C21
C33
C20
C27.
C23
C25.
C32
C37
C13
C42.O11
C26.
C28.
C7
C14
C12
C22
C15
C8
C41.O9
C39.O6
C10
C24
C38.O5
O1
C3
C40.O7
C5
C11
C4
O10
O8
C9
C19
O17
O28
O20
C46.O19
O27
O25
O26
Cumulative %
correlation
0.894
0.891
0.882
0.861
0.845
0.813
0.805
0.801
0.790
0.739
0.719
0.670
0.670
0.644
0.608
0.603
0.578
0.573
0.569
0.500
0.480
0.460
0.454
0.411
0.378
0.353
0.345
0.311
0.294
0.272
0.254
0.238
-0.240
-0.281
-0.420
-0.473
-0.509
-0.516
-0.576
-0.582
-0.628
PC1
25.681
PC2
O21
O23
C43.012
C6
O18
O24
C46.O19
O20
O17
C10
O26
O27
O25
C22
C8
C3
O8
O29
O28
C9
C12
C7
C4
C40.O7
C14
C28.
O22
C13
C42.O11
C5
C31.O2
C38.O5
C19
C36.O4
C44.O13
C2
correlation
0.683
0.680
0.663
0.598
0.584
0.558
0.550
0.549
0.528
0.520
0.497
0.485
0.483
0.461
0.444
0.425
0.418
0.407
0.403
0.399
0.394
0.381
0.366
0.355
0.336
0.331
0.300
0.213
0.207
0.200
-0.222
-0.233
-0.261
-0.284
-0.507
-0.614
PC2
38.845
127
PC3
O1
C11
C24
O16
C31.O2
O10
C39.O6
C36.O4
C41.O9
O22
C26.
C40.O7
O29
O27
O17
O20
C23
C38.O5
O15
C15
O21
C25.
O26
O28
C45.O14
C7
C28.
C8
C14
C22
C4
C10
C3
C5
correlation
0.701
0.636
0.608
0.557
0.547
0.535
0.522
0.515
0.463
0.443
0.334
0.330
0.305
0.291
0.283
0.272
0.269
0.266
0.262
0.247
0.239
0.235
0.219
0.219
0.202
-0.222
-0.249
-0.297
-0.305
-0.348
-0.398
-0.403
-0.411
-0.510
PC3
48.795
Table S6. Calling Song and Courtship song Analysis: pairwise comparisons for post-hoc
Tukey test on PC scores from first four PCs. Comparisons between reciprocal hybrid males
are highlighted in grey. Significant comparisons are in bold.
Calling Song
PC1
OC - CC
CO - CC
OO - CC
CO - OC
OO - OC
OO - CO
PC2
OC - CC
CO - CC
OO - CC
CO - OC
OO - OC
OO - CO
PC3
OC - CC
CO - CC
OO - CC
CO - OC
OO - OC
OO - CO
PC4
OC - CC
CO - CC
OO - CC
CO - OC
OO - OC
OO - CO
Courtship Song
Estimate
-2.969
-3.264
-5.839
-0.295
-2.870
-2.574
-1.826
0.064
0.902
1.891
2.728
0.837
Std.
Error
0.317
0.324
0.285
0.352
0.317
0.324
0.515
0.527
0.464
0.572
0.515
0.527
t value
Pr(>|t|)
-1.431
1.249
-2.308
2.68
-0.877
3.557
Std.
Error
0.664
0.664
0.573
0.767
0.689
0.689
-2.155
1.88
-4.029
3.494
-1.272
5.16
0.152
0.25
0.001
0.006
0.583
<0.001
0.004
0.999
0.220
0.009
0.001
0.392
PC2
OC - CC
CO - CC
OO - CC
CO - OC
OO - OC
OO - CO
0.658
0.955
1.862
0.297
1.205
-0.908
0.594
0.594
0.512
0.686
0.617
0.617
1.107
1.607
3.635
0.433
1.954
-1.472
0.685
0.384
0.004
0.972
0.22
0.46
PC3
OC - CC
CO - CC
OO - CC
CO - OC
OO - OC
OO - CO
-0.102
2.119
1.468
2.221
1.57
0.651
0.476
0.476
0.41
0.549
0.494
0.494
-0.214
4.457
3.581
4.045
3.181
1.319
0.996
<0.001
0.005
0.001
0.014
0.554
0.19
0.469
0.406
0.977
0.126
0.586
-0.065
0.395
-0.46
0.469
0.404
0.542
0.487
0.487
0.268
1.448
-0.12
0.812
-0.945
0.993
0.475
0.999
0.847
0.779
t value
Pr(>|t|)
PC1
-9.378
-10.064
-20.458
-0.838
-9.063
-7.937
<0.001
<0.001
<0.001
0.835
<0.001
<0.001
OC - CC
CO - CC
OO - CC
CO - OC
OO - OC
OO - CO
-3.549
0.122
1.944
3.304
5.302
1.589
1.422
0.853
0.441
-0.569
-0.981
-0.412
0.415
0.425
0.374
0.461
0.415
0.425
3.429
2.008
1.180
-1.234
-2.366
-0.970
0.006
0.196
0.641
0.607
0.095
0.766
0.068
0.403
0.169
0.998
-0.119
0.166
-0.187
0.098
0.285
0.413
0.363
0.448
0.403
0.413
-0.287
0.458
-0.416
0.244
0.691
0.992
0.968
0.975
0.995
0.900
PC4
OC - CC
CO - CC
OO - CC
CO - OC
OO - OC
OO - CO
128
Estimate
Table S7. CHCs: pairwise comparisons for post-hoc Tukey test on PC scores from the first
four PCs. Comparisons between reciprocal hybrid males and females are highlighted in grey.
Cuticular Hydrocarbons (CHCs)
PC1
PC2
CC_M-CC_F
-0.250
Std.
Error
1.019
CO_F-CC_F
2.425
0.943
2.571
0.178
-1.775
0.853
-2.081
0.432
CO_M-CC_F
-0.631
0.943
-0.669
0.998
1.625
0.853
1.905
0.549
OC_F-CC_F
1.344
0.975
1.378
0.864
-2.644
0.882
-2.999
0.064
OC_M-CC_F
0.176
0.898
0.196
1.000
3.216
0.812
3.960
<0.01
OO_F-CC_F
5.688
0.918
6.195
<0.001
-0.557
0.830
-0.671
0.998
OO_M-CC_F
8.940
0.975
9.167
<0.001
0.818
0.882
0.928
0.983
CO_F-CC_M
2.675
0.943
2.836
0.097
-3.594
0.853
-4.215
<0.01
CO_M-CC_M
-0.381
0.943
-0.403
1.000
-0.195
0.853
-0.229
1.000
OC_F-CC_M
1.594
0.975
1.635
0.726
-4.464
0.882
-5.062
<0.001
OC_M-CC_M
0.426
0.898
0.475
1.000
1.397
0.812
1.720
0.672
OO_F-CC_M
5.939
0.918
6.468
<0.001
-2.377
0.830
-2.863
0.091
OO_M-CC_M
9.190
0.975
9.423
<0.001
-1.002
0.882
-1.136
0.947
CO_M-CO_F
-3.056
0.861
-3.549
0.013
3.399
0.778
4.367
<0.001
OC_F-CO_F
-1.081
0.896
-1.206
0.928
-0.870
0.810
-1.073
0.961
OC_M-CO_F
-2.249
0.812
-2.770
0.114
4.991
0.734
6.801
<0.001
OO_F-CO_F
3.264
0.834
3.915
0.004
1.218
0.754
1.615
0.738
OO_M-CO_F
6.515
0.896
7.271
<0.001
2.593
0.810
3.200
0.037
OC_F-CO_M
1.975
0.896
2.204
0.358
-4.269
0.810
-5.269
<0.001
OC_M-CO_M
0.807
0.812
0.994
0.974
1.592
0.734
2.169
0.378
OO_F-CO_M
6.319
0.834
7.581
<0.001
-2.182
0.754
-2.895
0.084
OO_M-CO_M
9.571
0.896
10.681
<0.001
-0.807
0.810
-0.996
0.974
OC_M-OC_F
-1.168
0.849
-1.376
0.865
5.861
0.768
7.636
<0.001
OO_F-OC_F
4.344
0.870
4.995
<0.001
2.087
0.787
2.654
0.148
OO_M-OC_F
7.596
0.930
8.169
<0.001
3.462
0.841
4.118
<0.01
OO_F-OC_M
5.512
0.783
7.043
<0.001
-3.774
0.708
-5.333
<0.001
OO_M-OC_M
8.764
0.849
10.324
<0.001
-2.399
0.768
-3.125
0.046
OO_M-OO_F
3.252
0.870
3.738
0.007
1.375
0.787
1.749
0.654
Estimate
t value
Pr(>|t|)
1.820
Std.
Error
0.921
1.975
0.502
t value
Pr(>|t|)
Estimate
-0.246
1.000
129
Figure S1. Trill.1.2. histogram for the two parental species (CC – T. commodus; OO –
T.oceanicus) and the reciprocal hybrids (CO – male hybrids with a T. commodus X
chromosome; OC – male hybrids with a T. oceanicus X chromosome). The red dotted line
indicates the group mean.
130
Chapter 4
A Rare Exception to Haldane’s Rule: Are X Chromosomes Key to
Hybrid Incompatibilities?
.
Abstract
The prevalence of Haldane’s rule suggests that sex chromosomes commonly play a
key role in reproductive barriers and speciation. However, the majority of research on
Haldane’s rule has been conducted in species with conventional sex determination systems
(XY and ZW) and exceptions to the rule have been understudied. Here I test the role of Xlinked incompatibilities in a rare exception to Haldane’s rule for female sterility in field
cricket sister species (Teleogryllus oceanicus and T. commodus). Both have an XO sex
determination system. Using three generations of crosses, I introgressed X chromosomes
from each species onto a mixed genomic background and tested the fertility and viability of
all cross types. I predicted that females with two different species X chromosomes would
suffer reduced fertility and viability compared to females with two parental X chromosomes.
However, there was no strong support for X-linked incompatibilities. The results preclude XX incompatibilities and instead support an interchromosomal epistatic basis to hybrid female
sterility. I discuss the broader implications of these findings, principally whether deviations
from Haldane’s rule might be more prevalent in species without dimorphic sex chromosomes.
131
Introduction
Haldane’s rule is one of very few generalizations in evolutionary biology. It predicts
that in crosses between closely related species, if either sex of the offspring suffers
disproportionate fitness costs, such as reduced fertility or viability, it will be the
heterogametic sex (Haldane, 1922). It is a widespread phenomenon, observed across a broad
range of taxa, irrespective of whether males or females are heterogametic (e.g. mammals,
birds, reptiles, amphibians, fish, insects, and the plant genus Silene (Coyne & Orr, 2004;
Brothers & Delph, 2010; Schilthuizen et al., 2011). The pervasiveness of the rule indicates
that sex chromosomes might commonly play a key role in the establishment of postzygotic
reproductive barriers and by extension, speciation (Presgraves, 2008; Qvarnström & Bailey,
2009). However, the majority of research on Haldane’s rule has been conducted in species
with conventional sex determination systems (e.g. XY and ZW). Exceptions to the rule,
although rare, do occur but have been understudied and can be important for illuminating the
rule (Turelli & Orr, 1995; Laurie, 1997; Malone & Michalak, 2008; Watson & Demuth,
2012). Atypical sex determination systems and exceptions to the rule provide unique
opportunities to test the generality of proposed genetic explanations. Here, I test the
importance of X chromosome incompatibilities in a rare deviation from Haldane’s rule for
female sterility, which occurs in both cross directions, in an XO sex determination system.
The general consensus from published research is that Haldane’s rule results from a
composite of evolutionary processes (Coyne & Orr, 2004). This is unsurprising considering
that fertility and viability largely represent distinct functional pathways (Orr, 1993b; Wu &
Davis, 1993). Three of the most consistent genetic theories proposed to explain the ubiquity
132
of Haldane’s rule (which are not mutually exclusive) are the dominance theory, faster male
theory and the faster X theory (Coyne & Orr, 2004).
The dominance theory (Muller, 1942; Orr, 1993a; Turelli & Orr, 1995) proposes that
the heterogametic sex suffers disproportionate fitness effects because all X (or Z)-linked loci
involved in incompatible interactions with other loci are expressed. In contrast, the
homogametic sex will only be affected by dominant or co-dominant incompatibilities as
recessive X-linked incompatibility loci will be masked by the other X chromosome.
Therefore, a key assumption of the dominance theory is that X-linked incompatibility loci
contributing to the manifestation of Haldane’s rule should be predominantly recessive. The
dominance theory appears to be the most common cause underlying Haldane’s rule, as it has
the most empirical support and can explain both sterility and inviability irrespective of which
sex is heterogametic (Davies & Pomiankowski, 1995; Coyne & Orr, 2004).
The faster male theory (Wu & Davis, 1993) suggests that hybrid sterility is more
prevalent in heterogametic males due to sex differences in the rate of evolution of sterility
loci arising from stronger sexual selection in males. In addition, spermatogenesis has been
suggested to be especially prone to hybrid dysfunction (Wu & Davis, 1993; Presgraves, 2008;
Malone & Michalak, 2008). There is good empirical support for the faster male theory from
introgression experiments in mosquitoes (Presgraves & Orr 1998) and Drosophila (Coyne &
Orr, 2004; Masly & Presgraves, 2007), and gene expression studies in Drosophila (Michalak
& Noor, 2003; Ranz et al. 2004). However, the faster male theory fails to explain Haldane’s
rule in female heterogametic taxa, despite the fact that many groups such as Lepidoptera obey
Haldane’s rule for sterility (Presgraves, 2002).
The faster X theory copes with this because it argues that X chromosomes
disproportionately accumulate hybrid incompatibilities, as recessive loci that increase fitness
133
in the heterogametic sex would accumulate more readily on the X chromosome as they are
immediately exposed to selection (Charlesworth et al. 1987). Such a pattern could partly
reflect ascertainment bias from underestimating autosomal effects in backcross designs (Wu
& Davis 1993; Hollocher & Wu 1996), although genome-wide introgression studies in
Drosophila controlling for this potential bias have identified a higher density of hybrid male
sterility factors on the X chromosome compared to the autosomes (Masly & Presgraves,
2007). The faster X theory favours the occurrence of Haldane’s rule in both male and female
heterogametic species but has the weakest empirical support out of the three main theories.
Overall, these prominent genetic models all predict that X-linked incompatibilities play a
central role in Haldane’s rule.
Unusual sex determination systems and taxa that disobey Haldane’s rule provide
important opportunities to test the generality of these genetic models, to identify less well
recognized processes, and to disentangle their relative contributions to Haldane’s rule
(Malone & Michalak, 2008; Koevoets & Beukeboom, 2009; Schilthuizen et al., 2011).
Traditionally, species with XO systems have been understudied, and the species pairs which
have been examined have been found to conform to Haldane’s rule (Virdee & Hewitt, 1992;
Baird & Yen, 2000; Baird, 2002; Woodruff et al., 2010; Kozlowska et al., 2012). Recently,
Caenorhabditis nematodes (XO sex determination system) have emerged as a compelling
system for studying postzygotic reproductive barriers. Hybridization studies have revealed
that some of the species pairs exhibit Haldane’s rule (Baird, 2002; Dey et al., 2014; Bundus
et al., 2015). However, the diversity of reproductive modes, with many of the Caenorhabditis
species pairs examined involving gonochoristic (male/female) and androdioecious (male/
hermaphrodite) partners (exception Dey et al. (2014)) may make them difficult to compare to
dioecious taxa. Although the three main genetic models should still apply in XO taxa, the
134
absence of dimorphic sex chromosomes may reduce the likelihood that Haldane’s rule will
manifest (Johnson, 2010). An obvious distinction is the absence of Y chromosomes, which
have been found to play an important role in male sterility in some species of Drosophila but
not others (Coyne 1985; Turelli & Orr 2000). Additionally, the potential for meiotic drive or
genomic conflict, which have been argued to contribute to Haldane’s rule for sterility, may be
reduced in taxa with monomorphic sex chromosomes (Coyne et al., 1991; Frank, 1991; Tao
et al., 2001; Johnson, 2010; McDermott & Noor, 2010; Meiklejohn & Tao, 2010).
Like most Orthopterans the two closely related Australian field cricket species:
Teleogryllus oceanicus and T. commodus, have an XO sex determination system, yet they
provide an intriguing rare exception to Haldane’s rule for sterility and inviability (Hogan &
Fontana, 1973). As males of this species are heterogametic (XO - they inherit a single X
chromosome from their mother) and females are homogametic (XX - they inherit an X from
each parent) Haldane’s rule predicts that hybrid males should suffer disproportionate negative
fitness effects. However, early studies reported that reciprocal F1 hybrid females experienced
disproportionate sterility and inviablity compared to hybrid males (Hogan & Fontana, 1973).
Reasons for this exception to Haldane’s rule are not clear. Both T. oceanicus and T.
commodus share the same diploid number of chromosomes (2n = 26 + XO, XX), but differ in
the frequency of chiasmata and structural rearrangements, especially on the X (Fontana &
Hogan, 1969; Hogan & Fontana, 1973). As a result of these differences, one possibility is that
X-X interactions during alignment and crossing over may be disrupted, resulting in meiotic
dysfunction and thus hybrid female sterility. Hogan & Fontana (1973) reported that hybrid
females had degenerate ovaries and laid few eggs, suggesting a combination of
incompatibilities targeting both somatic and germ line cells in the female reproductive
system.
135
In this experiment I tested whether interactions between X chromosomes might
explain female sterility and inviability in T. commodus and T. oceanicus. X chromosomes
were introgressed from either species onto the background of the other over three generations
of crosses, and the fertility and viability of the different cross types were tested. I predicted
that females which inherited two different X chromosomes on a controlled autosomal
background would be less viable and suffer reduced fertility compared to females with two
pure parental species X chromosomes.
Methods
Maintenance and Rearing
Laboratory populations from the offspring of ca. 35 wild caught females from each of
two allopatric Australian populations were established (T. commodus – near Moss Vale,
NSW and T. oceanicus near Townsville, QLD). Colonies were bred in the lab for at least
three generations before the experiment began. Stock crickets were housed in 16-L plastic
boxes of ca. 80 individuals in a 25 ⁰C temperature-controlled room on a 12:12 light:dark
cycle. They were provided twice weekly with Burgess Excel “Junior and Dwarf” rabbit food
and cotton wool pads for drinking water and supplied with cardboard egg cartons for shelter.
136
Cross Design
The experimental design was similar across the three generations of crosses (Figure
4-1). Penultimate instar juveniles were separated into single-sex boxes to ensure virginity.
For crosses, virgin adult males and females ca. 10-20 days past eclosion were paired together
in smaller boxes (7 x 5 cm). Approximately 20 pairs per cross type were used (Figure 4-2).
Females oviposited in moist cotton pads; these egg pads were collected every three to four
days and mating pairs were kept together for a ten day period. Eggs were counted by
examining the egg pads with a magnifying glass. The collected egg pads were monitored
every 3-4 days, to prevent desiccation and to check for hatchlings. Newly hatched offspring
were provisioned with food ad libitum and cardboard shelter. Egg pads were retained for 2-3
weeks and the final hatchling count was conducted ca. 3 weeks after the final egg pad was
removed. Once the hatchlings reached the penultimate instar juvenile stage (ca. 2 months) sex
ratios were estimated.
In the first generation crosses (F1), which comprised heterospecific and conspecific
pairs, I tested whether the species obey Haldane’s rule for inviability and whether
unidirectional or bidirectional incompatibilities exist between them. The cross types were
classified by two letter codes, indicating the female offspring sex chromosome type. The first
letter indicates the maternal species identity and the second the paternal species identity (C =
T. commodus; O = T. oceanicus) (Figure 4-1).
In the second generation (BC1), reciprocal F1 hybrid females and males were
backcrossed to both parental species to test if the species obey Haldane’s rule for sterility and
if X-linked incompatibilities contribute to offspring inviability. The key comparisons were
between backcross types in which female offspring on average share the same autosomal
137
background (~75:25% species combination) but differ in their complement of X
chromosomes (Figure 4-1B). I predicted that cross types in which females inherited two
different species Xs would exhibit reduced numbers of hatchlings and a higher proportion of
males due to X-linked incompatibilities, compared to females which inherited two of the
same species Xs.
In the third generation (BC2), female offspring from BC1 were backcrossed to their
maternal species to test directly if X-linked incompatibilities contribute to female sterility.
The key comparisons were again between groups which on average share the same autosomal
background (~87.5:12.5% species combination expected) but differ in their sex chromosome
complement; either inheriting two pure species X chromosomes or one pure and the other an
inter-species recombinant X (Figure 4-1C).
138
Figure 4-1 Schematic of the cross design
[A] F1 Reciprocal Hybrids: Reciprocal inter-species crosses [B] Backcross 1 (BC1):
Reciprocal F1 hybrid females (i) and males (ii) backcrossed to both parental species. Female
hybrid crosses are highlighted in grey as I did not expect any offspring. Striped X
chromosomes represent inter-species X recombinants. Arrows indicate the key comparisons,
in which females either share or differ in their X chromosome complement. [C] Backcross 2
(BC2): BC1 females backcrossed to their maternal species. The arrows indicate group
comparisons. X in parentheses indicates an inter-species recombinant X. Control crosses, of
139
pure species pairs, were also carried out for the F1 and BC2 generations but are omitted for
clarity.
Statistical analysis
Binomial tests were applied to test whether sex ratios differed from the predicted
mean of 0.5 within each cross type and also to test whether the sex ratios differ between the
main groups of interest. Generalized linear models (GLM) were fitted to test whether the X
chromosome complement of females predicted their fertility as would be expected if X-linked
incompatibilities make a significant contribution to female fertility. All statistical analyses
were performed in R (Version 3.1.3).
The analyses focused on two types of data that reflect different processes: the
proportion of pairs exhibiting any response (a binary measure) among different cross types,
and the magnitude of any responses (a continuous measure) among cross types. For example,
the response variables included (i) the proportion of pairs that produced eggs, (ii) the
proportion that produced offspring, (iii) egg numbers, (iv) offspring numbers, and (v)
hatchling success rate (offspring/eggs). In each case, the main predictor of interest was
female offspring XX type which was fitted as a fixed effect. Female weight was fitted as a
covariate. The decision to include or remove variables from models was made based on
comparison of the model fit using ANOVAs and chi squared distributions (or F test for quasi
likelihood models). Models were compared using the Akaike information criterion (AIC),
and models with the lowest AIC were considered the best fit.
The count data were heavily overdispersed (theta > 20), so I examined if quasibinomial, quasi-poisson and negative binomial regression models fitted better using the
“MASS” package (Venables & Ripley, 2002). In some cases the models were still
140
overdispersed, so zero adjusted models were fitted. These type of models account for the
excess of zeros and allow for distinctions between two different biological processes; whether
females laid eggs and if they did, how many hatched. There are two types of zero adjusted
models which differ in the treatment of zeros: zero inflated and zero altered (Zuur et al.,
2012). Zeros in egg counts can be treated as arising from a single process, either females laid
eggs or did not lay eggs, and therefore I used zero altered models for egg counts (specifically
zero altered negative binomial (ZANB) models fitted best). However, an offspring count of
zero could occur when females lay no eggs, or when females laid eggs but none hatched,
therefore I used zero inflated models for offspring counts (specifically zero inflated negative
binomial (ZINB) models fitted best). Zero inflated models assume there are two processes
generating the zeros in the data and models these two processes separately, a poisson GLM
for the count data and a binomial GLM for the occurrence of zeros. The package “pscl” was
used to fit zero adjusted models (Zeileis et al. 2008). To test for differences between the
groups of interest Tukey pairwise comparisons were fitted with the multcomp package (glht
function; Hothorn et al., 2008).
141
Results
F1 generation
Asymmetric production of F1 hybrids
Reproductive success was strongly asymmetric. Crosses between T. commodus females and
T. oceanicus males (CO) had lower fertility compared to the reverse cross (OC). (Fig. 2A,
Table 1). Nearly all females laid eggs, but the number of eggs was markedly lower for CO
crosses (mean ± SE: CO, 84 ± 27.75) compared to the reciprocal cross (OC, 239.56 ± 34.28)
(Negative binomial GLM: Z3,80 =-3.226, P = 0.007; Table 1). There was an excess of zeros
among CO pairs, as only 41% of CO pairs produced offspring compared to 70% for OC
crosses (ZINB binomial: Z11,73 =2.426, P = 0.053). Females from the CO group also produced
fewer offspring (mean ± SE, 55.73 ± 21.8) than the OC cross (155.04 ± 29.77), although this
was non-significant (Table 1).
The asymmetry in reproductive success may be due to maternal effects or sperm-egg
incompatibilities. If X-cytoplasmic interactions contribute to the asymmetry in F1 production,
we predicted hybrid females would suffer disproportionate inviability compared to males as
they inherit an X on a foreign species’ cytoplasmic background. However, the absence of
sex-specific inviability indicates this is not the case (Fig. 2Aiii). In line with T. commodus
females performing poorly when crossed to a heterospecific, they also had reduced fertility
when paired with a conspecific partner in the F1 generation (Fig. 2Ai-ii; Table S1). They
produced both fewer eggs (Parental CC vs. Parental OO: negative binomial GLM, Z3,80 =
2.374, P = 0.082) and fewer offspring (ZINB negbin, Z3,80 = -3.325, P < 0.001). However,
this species difference was not observed in the BC2 generation (Fig. 2Ci - ii, Table S1).
142
Hatching
Success
Offspring
Female XX
Quasi-binomial
Female XX
0.585
<0.001
Binomial
-
0.022
0.029
0.002
Pr(>|X |)
2
negbin
-
Female weight
ZINB
-
Negative binomial
Eggs
Model
Components
Female XX
Model &
Predictors
Response
Variable
indicate statistical significance at α < 0.05.
143
CO - OC
OC - CO
OC - CO
OC - CO
Main
comparisons
3,82
11,73
11,73
1,79
3,80
Df
-1.021
1.441
-0.482
-1.014
Estimate
0.438
0.594
0.354
0.314
Std.
Error
-2.333
2.426
-1.362
-3.226
Z value
0.090
0.053
0.455
0.007
Pr(>|Z|)
fitted for either the negative binomial (negbin) or binomial component. Main comparisons based on Tukey pairwise contrasts. P values in bold
fixed effects examined using likelihood ratio tests (X2), by comparing a null model with only the intercept fitted to a model with the predictor
employs two components, a negative binomial count model (negbin) and the logit model (binomial) for predicting excess zeros. Significance for
were the X chromosome composition of female offspring (“Female XX”) and female weight. The Zero inflated negative binomial model (ZINB)
Results from generalized linear models examining egg number, offspring number and hatching success in F1 crosses. Main predictors fitted
Table 4-1 F1 crosses: GLM results
No evidence of Haldane’s rule for inviability
All four F1 cross types, two intra-specific (parental crosses) and two inter-specific crosses,
had a higher proportion of males than the expected 0.5 sex ratio (Binomial exact test: P
<0.001) (Fig. 2Aiii). Importantly, there was no differential viability between males and
females in the hybrid crosses compared to the parental crosses (Parental CC vs. CO: X2 =
0.418, df = 1, P = 0.518; Parental OO vs. OC: X2 = 0.02, df = 1, P = 0.888). Therefore, there
is no evidence for Haldane’s rule for inviability within these species.
BC1 Generation
Reciprocal exceptions to Haldane’s rule for sterility
T. oceanicus and T. commodus provide a reciprocal exception to Haldane’s rule as nearly all
hybrid females were sterile in both directions of the cross (only a single BC1 offspring was
produced from 80 backcrosses), while all four hybrid male backcross types were fertile (Fig.
2Bii). We predicted that hybrid male backcrosses which produced female offspring with
heterospecific X chromosomes would exhibit reduced fertility (BC1: OO vs. OC or CC vs.
CO) due to X-X interactions. We found no support for this hypothesis in either the proportion
of pairs exhibiting a response or in the strength of response (i.e. number of eggs or offspring
per pair) (Table 2). Contrary to the prediction that heterospecific X-X interactions would
reduce fertility, CO pairs (T. commodus females paired with male hybrids carrying a T.
oceanicus X chromosome) produced more eggs (mean ± SE: CO 242.4 ± 27.36) than the
comparison CC pairs (T. commodus females paired with male hybrids carrying a T.
commodus X chromosome) (mean ± SE: 103.25 ± 22.29) (ZANB negbin: Z9,70 = 3.72, P
<0.001). However, the number of offspring was not significantly different between these two
groups (mean ± SE: CO 40.55 ± 8.37 vs. CC 22.5 ± 6.89) (ZINB negbin: Z9,70 = -0.861, P =
144
0.389). In the other group comparison, there was no difference between OC and OO pairs in
either the number of eggs or offspring (Table 2). The hatching success rate also did not differ
amongst the groups of interest (Table 2). Overall, we detected no support for X-X
interactions affecting fertility.
Figure 4-2 Results from all three generations of crosses
A) F1, B) BC1 and C) BC2 showing, for each cross type, i) numbers of eggs, ii) numbers of
offspring and iii) proportion of male offspring (n, number of pairs per cross type. n equals
twenty for backcrosses with hybrid females). The X axis is labelled based on the female
offspring XX type, the first letter reflects the maternal species X and the second letter the
paternal species X (C=T. commodus; O=T. oceanicus). In BC1 and BC2, (H) indicates
potential inter-species recombination on the X. Significant comparisons are highlighted by
brackets (P>0.01=*, P>0.001=**, P<0.001=***). In the last row, error bars indicate 95%
confidence intervals (binomial test) for the observed proportions.
145
Hatching
Success
Offspring
ZANB
Female XX
Eggs
Female XX
Quasi-binomial
Female XX
ZINB
Female XX
Female XX
Model &
Predictors
Response
Variable
-
negbin
Binomial
negbin
Binomial
negbin
Binomial
negbin
Binomial
Component
Model
0.189
0.015
0.197
-
< 0.0001
0.47
-
Pr(>X2)
values in bold indicate statistical significance at α < 0.05.
146
OO - OC
CC - CO
OO - OC
OO - OC
CC - CO
CC - CO
OO - OC
OO - OC
CC - CO
CC - CO
Main comparisons
3,67
9,70
9,70
9,70
9,70
9,70
9,70
9,70
9,70
Df
0.087
-0.166
0.085
-0.326
0.261
-1.545
0.039
0.876
0.799
0.747
Estimate
0.323
0.354
0.293
0.717
0.303
0.782
0.228
0.934
0.215
1.268
Std.
Error
0.268
-0.469
0.291
-0.455
0.861
-1.975
0.172
0.937
3.721
0.589
Z value
0.993
0.966
0.989
0.959
0.783
0.155
0.998
0.733
<0.001
0.915
Pr(>|Z|)
the predictor fitted for either the negative binomial (negbin) or binomial component. Main comparisons based on Tukey pairwise contrasts. P
Significance for fixed effects examined using likelihood ratio tests (X2), by comparing a null model with only the intercept fitted to a model with
model employs two components, a negative binomial count model (negbin) and the logit model (binomial) for predicting excess zeros.
negative binomial distribution (negbin) while all the zero data is modelled together (binomial). The zero inflated negative binomial (ZINB)
model zeros. The zero altered negative binomial (ZANB) model employs two components, the positive (i.e. non-zero) data follows a truncated
were female offspring XX type (“Female XX”) (female weight was not significant). Zero altered and zero inflated models differ in how they
Results from Generalized Linear Models examining egg number, offspring number and hatching success in BC1 crosses. Main predictors fitted
Table 4-2 BC1 crosses: GLM results
No X effect on viability
Under a scenario in which X-linked incompatibilities disproportionately affect viability, we
predicted an excess of males due to female inviability in groups in which females inherited
two different species Xs. Again, contrary to this prediction, there was a lower proportion of
females in the OO group than the expected mean of 0.5 (Binomial exact test, P <0.001), and
this was significantly lower than the comparison group OC (OO vs. OC groups: X2 = 5.358,
df=1, P =0.021) (Fig. 2Biii). Comparing the CC vs. CO cross types, there was no sex ratio
bias (X2 = 2.326, df=1, P =0.127). Overall, females that inherited two different species X
chromosomes did not exhibit reduced viability.
BC2 Generation
X-X interactions do not cause female sterility
We predicted that females with a mixed species complement of X chromosomes would suffer
reduced fertility compared to females with conspecific X chromosomes. There was no
difference between the CC vs. (H)C groups in either the number of eggs produced (ZANB
negbin: Z13,128 = -0.418, P = 0.992) or the number of offspring (ZINB negbin: Z13,128 = 0.417,
P = 0.991; Fig. 2Cii, Table S1). In line with our prediction, there was a marginal difference in
fertility between OO vs. (H)O groups. OO females appeared to produce more eggs (mean ±
SE: OO, 92.5 ± 22.9 vs. (H)O, 33.13 ± 14.2), although this was not significant (ZANB
negbin: Z13,128 = -1.593, P = 0.434, Table 3). However, OO pairs produced more offspring
than the corresponding (H)O group (mean ± SE: OO, 28.68 ± 10 vs. (H)O, 6.92 ± 3.34)
(ZINB negbin: Z13,128 = 2.957, P = 0.017; Table 3), which was consistent with our prediction
that females with a mixed species complement of X chromosomes will suffer reduced
147
fertility. Although the proportion of parental OO pairs (control crosses) that produced eggs
was surprisingly low (0.56) (Table S1), all parental pairs that produced eggs resulted in
hatchlings, compared to a range of only 19% - 63% for the backcrosses.
Limited role for X chromosomes in inviability
Sex ratio data showed a higher proportion of females in the (H)O group compared to the OO
group (Binomial test; X2 =4.059, df =1, P =0.044) indicating that (H)O males may suffer
disproportionate inviability (Fig. 2Ciii). In this cross, males potentially inherit an interspecies
recombinant X, which is hemizygous and could therefore expose them to an elevated
likelihood of epistatic incompatibilities involving recessive X substitutions (e.g. X-autosomal
incompatibilities). Comparisons between CC & (H)C revealed no significant sex ratio
difference (Binomial test; X2 =0.772, df=1, P =0.38). Both parental species crosses showed a
reduction of females from the expected mean of 0.5, particularly in the parental CC crosses
(Fig. 2Ciii).
148
ZANB
Female XX
ZINB
Female XX
Quasi-Binomial
Eggs
Offspring
Hatching
Success
Female XX
Female weight
Female XX
Model &
Predictors
Response
Variable
-
negbin
Binomial
negbin
Binomial
negbin
Binomial
negbin
Binomial
Component
Model
< 0.001
0.004
0.011
<0.0001
0.125
0.011
Pr(>|X2|)
149
OO vs. (H)O
CC vs. (H)C
CC vs. (H)C
CC vs. (H)C
OO vs. (H)O
OO vs. (H)O
CC vs. (H)C
CC vs. (H)C
OO vs. (H)O
OO vs. (H)O
Main
comparisons
5, 85
1, 84
5, 85
13,128
13,128
13,128
13,128
13,128
13,128
13,128
13,128
Df
0.125
0.253
0.344
0.342
1.479
-0.579
0.168
-0.135
0.662
1.168
Estimate
0.553
0.737
0.824
0.817
0.500
0.629
0.402
0.519
0.416
0.648
Std.
Error
0.225
0.344
0.417
0.418
2.957
-0.920
0.418
-0.259
1.593
1.801
Z value
1.000
0.999
0.991
0.991
0.017
0.856
0.992
0.999
0.434
0.301
Pr(>|Z|)
excess zeros. Significance for fixed effects examined using likelihood ratio tests (X2). Main comparisons based on Tukey pairwise contrasts.
binomial (ZINB) model employs two components, a negative binomial count model (negbin) and the logit model (binomial) for predicting
follows a truncated negative binomial distribution (negbin) while all the zero data is modelled together (binomial). The zero inflated negative
(“Female XX”) and female weight. The zero altered negative binomial (ZANB) model employs two components, the positive (i.e. non-zero) data
GLM examining egg number, offspring number and hatching success in BC2 crosses. Main predictors fitted were female offspring XX type
Table 4-3 BC2 crosses: GLM results
Discussion
Two important empirical findings in evolutionary biology, Haldane’s rule and the
large X effect, are so consistent that they have been likened to “rules” (Coyne & Orr, 1989;
Coyne & Orr, 2004). Both suggest that X chromosomes play a key role in the establishment
of post-zygotic barriers between species (Coyne & Orr, 1989; Masly & Presgraves, 2007;
Presgraves, 2010). However, most research on the genetic basis of reproductive isolation has
focused on male sterility and on male heterogametic species, as opposed to female fertility
(though see Orr & Coyne, 1989; Hollocher & Wu, 1996; Watson & Demuth, 2012; Suzuki &
Nachman, 2015). Rare cases in which homogametic females suffer disproportionate effects of
hybridization provide an important opportunity to investigate the genetic basis of female
sterility and processes that may counter Haldane’s rule. Crosses between T. oceanicus and T.
commodus provide one such remarkably rare exception to Haldane’s rule – female hybrids
were almost uniformly sterile in this experiment, out of 80 backcrosses with reciprocal hybrid
females only a single offspring hatched. A considerable number of hybrid females, amongst
the different cross types, produced eggs which indicates that not all ovaries are degenerate
(Figure 4-2Bi). This suggests a complex genetic basis for hybrid female sterility, in which
certain hybrid genic combinations may occasionally result in fertile hybrid females in natural
populations (Virdee & Hewitt, 1994).
Contrary to a previous report, which found a 1:1 sex ratio for pure-species crosses (Hogan &
Fontana, 1973), we found a male biased sex ratio for both intraspecific and interspecific
crosses. This discrepancy between the studies could have arisen due to population
differences. The previous cytogenetic (Fontana & Hogan, 1969) and hybridisation work
(Hogan & Fontana, 1973) was conducted on laboratory populations of T. oceanicus collected
150
from Ayr, northern Queensland (ca. 90km from where we sampled our study population in
Townsville), and T. commodus from Melbourne, southern Victoria (ca. 750km from where
we sampled our study population in Moss Vale, New South Wales). In general, populations
within a species can show a high degree of variation for genetic incompatibilities (Cutter,
2012) with other species, including X-chromosome inversions, endosymbiont strains or
infection rates (e.g. Wolbachia (Telschow et al. 2005)) that alter sex ratios. However, the
latter mechanisms usually result in female bias. In addition, differences in environmental
conditions, such as temperature, or differential fertilization of nullo-X sperm may alter sex
ratios (Wade et al., 1999; Bundus et al., 2015).
Could the same incompatibilities that cause F1 female sterility explain the result that
BC2 X(O) females produce fewer offspring than OO females? The reduced fertility for the
BC2 X(O) crosses supports a role of X chromosome incompatibilities. However, this cannot
explain the pattern of hybrid female sterility in both cross directions as there was no
detectable difference between the X(C) and CC females. I hypothesized that reciprocal hybrid
female sterility had a shared basis, namely due to chromosomal rather than genic interactions,
in particular X-X interactions leading to meiotic dysfunction. Overall the results do not
support a shared basis to female sterility, instead female sterility may have a different basis in
both cross directions.
Asymmetrical reproductive isolation
Asymmetrical genetic incompatibilities are a common observation among animal and
plant hybridizations (Turelli & Moyle, 2007). They are believed to principally arise from
negative epistasis between autosomal or sex-linked loci and uniparentally inherited maternal
151
factors (e.g. mitochondrial DNA, cytoplasmic background) (Turelli & Orr, 2000; Turelli &
Moyle, 2007; but see Bundus et al., 2015). In this study, there was a clear asymmetry in
genetic compatibility; T. commodus females mated to T. oceanicus males produced far fewer
eggs and offspring than the reciprocal cross (Figure 4-2A). In other words, hybridisation was
more successful when the mother was T. oceanicus. This unidirectional incompatibility
appears to manifest at a very early stage, as egg laying was disrupted.
Maternal effects may disproportionately affect female viability if incompatibility loci
are sex linked, as hybrid females inherit one of their X chromosomes on a different species’
cytoplasmic background. However, there was no sex-specific inviability in comparisons
between the hybrid and parental species crosses (Figure 4-2Aiii). Instead, sperm-egg
incompatibilities or autosomal-cytoplasmic interactions, rather than X cytoplasmic
interactions, may be responsible for the asymmetrical reduction in fertility. If species differ in
the degree of sperm competitiveness, asymmetric gametic isolation may occur (MartínCoello et al., 2009). Females of both Teleogryllus species mate multiply in natural
populations, and paternity is highly skewed, more so in T. oceanicus than T. commodus
(Simmons & Beveridge, 2010a). Heterospecific crosses with T. oceanicus males may
therefore be predicted to have higher mating success compared to the reciprocal cross.
However, this was not the case; heterospecific crosses with T. oceanicus males had reduced
fertility compared to the reverse cross.
X-linked incompatibilities
What is the genetic cause of the deviation from Haldane’s rule for sterility in
Teleogryllus, and can it inform us more broadly about hybrid incompatibilities? Maternal
effects have often been implicated in deviations from Haldane’s rule for inviability
152
(Sawamura et al., 1993; Sawamura, 1996) but not sterility (Orr & Irving, 2001). Early
developmental stages are predicted to be especially sensitive to maternal effects (Mousseau,
1991), however little is known about maternal effects on adult reproductive traits. Disruption
to early developmental stages could influence later reproductive output. However I do not
believe this can explain the hybrid female sterility in this study system as maternal effects
often exhibit asymmetrical effects and would not be expected to influence both directions of
the cross equally (Turelli & Moyle, 2007). Also, if maternal effects were playing a role in
female sterility one would expect backcrosses with hybrid males to be more compatible with
their maternal species, which was not the case.
Laurie (1997) highlighted two factors that might promote exceptions to Haldane’s
rule with respect to female hybrid sterility, and that affect both directions of a cross equally:
X-X incompatibilities and dominant X-autosomal interactions. Both depend on X
interactions, but the results yielded negligible support for the former. Only one of the
comparisons was consistent with X-linked incompatibilities reducing female fertility; a
higher number of offspring produced from OO vs. (H)O groups in BC2 (Figure 4-2ii, Table
4-3). In contrast, among the BC1 crosses the CO pairs produced more eggs on average than
CC pairs (Figure 4-2B, Table 4-2). This pattern also refutes the prediction. If X-X
incompatibilities were primarily responsible for the sterility of F1 hybrid females, I expected
to observe a clear reduction in fertility for crosses in which females inherited two different X
chromosomes. Instead, the results are more consistent with an epistatic origin of the
incompatibilities. This could be autosomal-autosomal or could still involve the X
chromosome if these were X-A interactions. The results do not unambiguously distinguish
these, but the fact there are some large differences between similar genotypes that differ in
the source of the X and A chromosomes rather than the proportion of interspecies material
153
(e.g. CC versus OO in BC1, Figure 4-2C) implies that specific X-A interactions may
contribute to the lower female fertility.
The lack of a large X effect on female sterility may not be surprising as theory
predicts a disproportionate accumulation of male but not female fertility loci on the X
chromosome in male heterogametic species (Charlesworth et al. 1987). The loci underlying
female fertility may be just as likely to accumulate on the autosomes as on the X (Masly &
Presgraves, 2007), so X-linked loci that affect male fertility would need to have pleiotropic
effects in hybrid females to produce a large X effect on female fertility (Coyne & Orr, 1989;
Presgraves, 2008). Introgression studies examining the large X effect in Drosophila have
provided mixed results; some support the view that male and female sterility loci are
qualitatively different (Wu & Davis, 1993; Coyne & Orr, 2004), while others have detected X
effects on both male and female sterility (Orr, 1987; Orr & Coyne, 1989). In this study I did
not test the effect of X introgression on the fertility of both sexes, but the absence of evidence
for a large X effect in females supports the view that X chromosomes do not play a
pronounced role in female sterility.
XO sex determination systems
As exceptions to Haldane’s rule are extremely rare, particularly in both directions of a
cross, could deviations for female sterility be caused by a peculiarity of XO sex
determination systems? While the main genetic models underlying Haldane’s rule should
apply to XO systems, the absence of dimorphic sex chromosomes may relax the operation of
some less well recognized processes that could contribute to Haldane’s rule (e.g. meiotic
drive, Y-incompatibilities). Previous hybridization studies in XO taxa suggest they generally
154
obey Haldane’s rule (Ohmachi & Masaki, 1964; Mantovani & Scali, 1992; Virdee & Hewitt,
1992; Baird & Yen, 2000; Baird, 2002; Woodruff et al., 2010; Kozlowska et al., 2012). Some
Caenorhabditis species pairs exhibit Haldane’s rule through sexual transformation where
genetic males appear phenotypically female (Baird, 2002), while others exhibit male
inviability and sterility, even in both cross directions (Woodruff et al., 2010; Kozlowska et
al., 2012).
However, only two previous reciprocal exceptions to Haldane’s rule have been
described, one for inviability in an XO species (Abe et al., 2005) (Spence, 1990) and the
other for male sterility in a female heterogametic species (Malone & Michalak, 2008). The
latter exception can be explained under existing theory and has been experimentally shown to
be due to faster male evolution (Malone & Michalak, 2008), which would not explain the
exception to Haldane’s rule in our study system. The former case occurs in the Heteropteran
pondskater Limnoporous spp which has an XO sex determination system (Spence 1990).
Spence (1990) found that in crosses between Limnoporus notablis and L. dissortis, F1 hybrid
females suffer disproportionate inviability compared to male hybrids. Applying a backcross
design similar to that used in this study, Spence (1990) tested whether the presence of two
different species X chromosomes contributed to hybrid inviability. Contrary to the results in
this study, he detected a large X effect on female inviability. The absence of a large X effect
for female sterility in this study probably reflects differences between the genetic pathways
underlying sterility and inviability (Wu & Davis, 1993). Considering that XO species
represent a relatively small fraction of the species examined in hybridization studies, yet
exhibit two remarkably rare exceptions to Haldane’s rule (Limnoporous spp – female
inviability; Teleogryllus spp – female sterility), future research would benefit from
investigating why Haldane’s rule might be less prevalent in systems which lack dimorphic
sex chromosomes.
155
Conclusions
T. commodus and T. oceanicus provide a rare exception to Haldane’s rule for sterility,
but not viability. There was a clear asymmetry in fertility in reciprocal F1 crosses, indicating
that maternal effects (e.g. autosomal-cytoplasmic interactions) or sperm-egg incompatibilities
may play an important role in reproductive barriers between these species. Asymmetrical
genetic compatibility can provide a basis for differential introgression of genes from one
species into the background of another, but the direction will also depend on mating
behaviour and sex biased dispersal. Unexpectedly, there was negligible support for X-linked
incompatibilities contributing to hybrid female sterility. The absence of X-X incompatibilities
is surprising given the size of the X chromosomes in these species. When in single copies in
males these represent approximately 20% of the diploid male genome, and when in two
copies in females represent approximately 30% of the diploid female genome (K Klappert;
unpublished data/pers comm). Introgressing such large chromosomes onto a predominantly
heterospecific genomic background might be expected to lead to a number of disruptive genic
and chromosomal interactions, particularly considering that the species are known to differ in
structural rearrangements on the X chromosome (Fontana & Hogan, 1973). Even though no
large X effect was detected in this study it does not rule out the potential for X-linked
incompatibilities. However, the low fitness amongst the backcross groups, irrespective of
their XX identity, suggests that partially dominant autosomal loci may supersede X-linked
interactions in disrupting female fertility. Whether this rare exception to Haldane’s rule
represents a more general pattern of deviation from the rule in systems without dimorphic sex
chromosomes (e.g. XO systems, haplodiploid) remains to be determined.
156
Supplementary
Table S1. Fertility results from three generations of hybridization crosses. The cross types
are labelled based on the species X identity, C= T.commodus, O= T.oceanicus, while hybrid
individuals have a prefix either F1 or BC1 and the species identity of their X chromosomes.
Female hybrids are labelled with two letters indicating the inheritance pattern of their X
chromosome, with the first letter representing the maternal species and the second the
paternal species identity. (H) indicates an inter-species recombinant X. The cross type,
number of pairs (n), the offspring identity (X chromosome complement) are shown. The
corresponding measurements are the proportion of pairs which exhibited a response (eggs,
offspring), the mean and standard deviation (SD) for the number of eggs, number of offspring
and the % hatching success (offspring/eggs). For the proportion of offspring and the mean
number of offspring only pairs that produced eggs are considered
Cross Type
♀x♂
F1 Cross
n
Offspring
identity
♀-♂
Eggs
Offspring
% Hatching
Success
Prop
Mean
SD
Prop
Mean
SD
Mean
SD
0.88
1
1
1
84.00
239.56
363.12
177.80
138.75
178.13
157.20
148.42
0.36
0.7
0.88
0.9
55.72
155.04
251.00
104.500
102.34
154.66
140.09
89.14
0.25
0.48
0.64
0.55
0.35
0.37
0.27
0.3
0.4
0.45
0.4
0.45
3.45
15.3
12.55
10.4
7.47
39.4
28.8
26.87
0
0
0.13
0
0
0
0.13
0
0
0
0.35
0
0
-
0.01
-
0.9
0.95
0.79
0.9
103.25
242.4
260.47
308.8
99.68
122.37
199.13
186.35
0.61
0.85
0.87
0.83
22.5
40.55
70.73
74.06
29.24
37.45
55.44
62.79
0.13
0.15
0.2
0.22
0.14
0.11
0.15
0.17
0.53
0.57
0.54
0.79
0.56
0.94
48.67
43.7
33.13
92.5
108.19
143.82
78.85
63.86
69.59
112.21
127.83
129.67
18.8
23.5
59
63.2
1
1
7.31
6.59
6.92
28.68
82.44
64.75
20.52
16.82
12.03
43.60
56.48
54.46
0.06
0.05
0.1
0.14
0.28
0.19
0.16
0.1
0.12
0.19
0.44
0.52
F1
CxO
25
CO - O
OxC
27
OC - C
Controls: O x O 17
OO - O
Controls: C x C 20
CC - C
BC1 Cross
BC1
Hybrid females x parental species
F1(OC) x C
20 (H)C - (H)
F1(CO) x C
20 (H)C - (H)
F1(OC) x O
20 (H)O - (H)
F1(CO) x O
20 (H)O - (H)
Parental species x hybrid males
C x F1 (C)
20
CC - C
C x F1 (O)
20
CO - C
O x F1 (C)
19
OC - O
O x F1 (O)
20
OO - O
BC2 Cross
BC2
BC1 (CC) x C
30
CC - C
BC1 (CO) x C
30
(H)C - (H)
BC1 (OC) x O
24 (H)O - (H)
BC1 (OO) x O
24
OO - O
Controls: O x O 16
OO - O
Controls: C x C 17
CC - C
157
Chapter 5
Genomic Discordance: Unusual Patterns of Differentiation at
Autosomal vs. X-linked Loci
Abstract
X chromosomes are expected to evolve faster than autosomes and to play a key role in
the establishment of reproductive barriers. The asymmetry in inheritance and expression of X
chromosomes means evolutionary forces may act differently on them than on autosomes.
Comparing the pattern of diversity for X-linked and autosomal markers provides a method to
disentangle the relative importance of evolutionary processes that shape genome variation.
Here I present the first detailed genomic study of two closely related field cricket species:
Teleogryllus commodus and T. oceanicus, which are a classic study system for sexual
selection. Using RADseq I examined the geographic distribution of genetic variation across a
broad area encompassing both allopatric and sympatric populations. There was no evidence
of recent hybridization or introgression, which suggests the species are currently
reproductively isolated. I tested if X-linked loci disproportionately contribute to delimiting
the species boundary and structuring populations, compared to autosomal loci. For both
marker types the species were strongly differentiated, being slightly higher for X-linked loci
(mean FST values = 0.435) compared to autosomal (mean FST values = 0.407). However,
contrary to theoretical predictions X-linked markers showed reduced intraspecific population
differentiation compared to autosomes. The discordance was particularly pronounced within
T. oceanicus. Selective sweeps and/or a combination of sex-biased processes, such as femalebiased dispersal and a polygynous mating system, may explain the surprising discordance. In
158
addition, I identified a third putative species T. marini, which has only previously been
described once, based on its distinct calling song and genitalia. The genomic results
definitively confirm its distinct species identity.
Introduction
Genomic landscapes: Merging intraspecific and interspecific patterns
Understanding the processes that shape the organization of biological diversity is a
central aim of evolutionary biology. The geographic distribution of genetic variation, within
and between species, provides a rich source of information on a species evolutionary history
and the nature of species boundaries (Hewitt, 2001; Sousa & Hey, 2013; Lexer et al., 2013;
Gompert et al., 2014). Distinguishing the relative roles of selection and neutral processes in
shaping the distribution of genotypic variation remains an important challenge (Laurent et al.,
2016). The presence of strong population genetic structure can indicate spatially variable
selection (Moura et al., 2014; Saenz-Agudelo et al., 2015; Machado et al., 2016) or
demographic processes such as sex-biased (Keinan et al., 2009; Chatfield et al., 2010;
Darling et al., 2014) or limited dispersal ability (Bailey et al., 2007). Spatial isolation within
species and hybridization between them can blur species boundaries (Abbott et al., 2013;
Keller et al., 2013; Payseur & Rieseberg, 2016). Population genomics offers an important
opportunity to disentangle the myriad forces which influence population divergence and
ultimately speciation (Luikart et al., 2003; Sousa & Hey, 2013; Seehausen et al., 2014).
Relative role of selection and demography
Discordance among marker types (i.e. autosomal, sex-linked, mitochondrial) can
elucidate evolutionary processes that shape genome variation (Shaw, 2002; Keinan et al.,
159
2009; Lavretsky et al., 2015). Comparisons between autosomal and sex chromosomes are of
particular interest as the latter have been suggested to play a key role in the establishment of
reproductive barriers and, by extension, speciation (Saetre et al. 2003; Presgraves, 2008;
Qvarnström & Bailey, 2009). As X chromosomes have an effective population size (Ne) that
is only three quarters that of autosomes (assuming an equal proportion of males and females),
demography and selection are expected to impact them differently (Charlesworth et al., 1987;
Vicoso & Charlesworth, 2006; Ellegren, 2009; Mank et al., 2010). Under neutral expectations
(population of constant size with only drift and mutation occurring), X chromosomes are
predicted to exhibit relatively lower population diversity and increased differentiation
between populations due to the smaller Ne, which will increase drift, causing alleles to be
fixed quicker (Ellegren, 2009). Sex-biased dispersal can alter the relative effective population
sizes of these genomic regions, with female-biased dispersal tending to equalize the level of
divergence among X-linked vs. autosomal loci.
Furthermore, selection is expected to be stronger on X (or Z in female heterogametic
species) chromosomes as recessive mutations are immediately exposed to selection in the
heterogametic sex (Charlesworth et al., 1987). Hitchhiking effects and selective sweeps may
also be more pronounced on the X chromosome due to the smaller Ne and reduced
recombination rate, leading to a higher turnover of alleles (Kauer et al., 2002). Indeed,
numerous studies have identified that X chromosomes exhibit higher rates of substitutions
(Mank et al, 2007), greater population differentiation (Keinan et al., 2009; Amato et al., 2009;
Lucotte et al., 2016; Machado et al., 2016) and a disproportionate accumulation of both preand postzygotic barriers compared to autosomes (Saetre et al., 2003; Presgraves, 2008;
Qvarnström & Bailey, 2009; Janoušek et al., 2012; Lavretsky et al., 2015; Maroja et al.,
2015). Sex chromosomes may therefore play a pronounced role in reproductive isolation and
speciation.
160
Introgressive hybridization
Hybridization has long been recognized as a widespread process in plants but viewed
as rare in animals (Arnold, 1997; Dowling & Secor, 1997). However, rates of hybridization
have most likely been underestimated due to the difficulty in detection (Mallet, 2005),
because many F1 hybrids or backcross individuals may be morphologically indistinguishable
from the parental species. Increasingly, detailed genetic studies are highlighting that
introgressive hybridization is widespread and may play an important role in animal
diversification (Mallet, 2005; Abbott et al., 2013; Harrison & Larson, 2016). If the boundary
between species is largely porous, much of the evolution of reproductive isolation and
adaptive divergence may occur in the presence of gene flow (Nadeau et al., 2012; Keller et
al., 2013). Hybridization can lead to the erosion or strengthening of reproductive barriers
(Virdee & Hewitt, 1992; Virdee & Hewitt, 1994), the extinction of local species (Seehausen
et al., 1997), the formation of stable hybrid zones ( Barton & Hewitt, 1985; Harrison, 1993)
or even hybrid speciation (Mavárez et al., 2006; Keller et al., 2013; Schumer et al., 2014).
The fitness of hybrids will largely determine which evolutionary outcome occurs.
Hybridization is often viewed to have negative fitness effects which serve to reinforce
pre-existing reproductive barriers through the exaggeration of assortative mating
(Dobzhansky, 1940; Coyne & Orr, 2004). However, not all hybrids fare worse than pure
species individuals (discussed in Chapter 3). Introgressive hybridization can contribute
genetic variation faster than mutational processes, and some of these variants may have
adaptive or novel functions, having been tested previously on a different genetic background
(Mallet, 2005; Rieseberg, 2009; Hedrick, 2013; Martin et al., 2014). Regions of the genome
that are neutral or confer positive fitness effects are likely to introgress freely across the
species boundary, whereas regions enriched for genes involved in reproductive isolation
(such as is predicted for sex chromosomes), will be resistant, resulting in heterogeneous
161
genomic divergence (Barton & Hewitt, 1985; Nosil et al., 2007; Nosil et al., 2009; Nosil &
Feder, 2012; The Heliconius Genome Consortium et al., 2012). Extensive introgression has
been shown to occur well beyond species contact zones (Kronforst et al., 2013; Poelstra et al.,
2014). Examining patterns of introgression and differentiation allows for identification of the
regions of the genome that contribute to maintaining species integrity (Payseur, 2010;
Gompert et al., 2012; Parchman et al., 2013).
Study system
I evaluated the geographic distribution of genetic variation among two closely related
field cricket species, Teleogryllus commodus and T. oceanicus, across a large geographic area
using RADseq. I examined the pattern of population differentiation and species divergence at
both autosomal and X-linked loci. These species are a classic system for sexual selection
studies and their hybrids have been the subject of influential studies on the inheritance of
acoustic signals and female preferences (Hoy & Paul, 1973; Hoy et al., 1977). The species
are partially interfertile and overlap across a broad area of sympatry on the eastern coast of
Australia (Hill et al., 1972; Otte & Alexander, 1983). However, little is known about their
evolutionary history or how the species interact in the area of overlap. In particular, it is not
known whether introgressive hybridization occurs and if sympatry is due to secondary
contact or parapatric divergence. The extent of introgressive hybridization will depend on the
fitness of hybrids and demographic conditions such as population densities of both species
(Payseur, 2010; Payseur & Rieseberg, 2016). Laboratory studies have found both pre-mating
and post-mating barriers between the species (Chapters 3 & 4). In particular, hybrid females
are sterile in both cross directions which provides a rare exception to Haldane’s rule (Moran
et al., in review). However, hybrid males have been shown to successfully mate with females
162
of both species (Chapter 3). Therefore, F1 hybrids are expected to be rare in sympatry but
backcrossing and introgression may be common.
Hybridization and introgression is often associated with human disturbed areas
(Dowling & Secor, 1997; Mallet, 2005). The recent colonization and radical expansion of
agricultural pastures on the eastern coast of Australia may have influenced the southern
species T. commodus to expand northwards (Cairns et al., 2010), bringing the two species
into contact, with the potential for hybridization and introgession. Alternatively, secondary
contact could also have occurred following an older postglacial expansion (Hewitt, 2000;
Hewitt, 2001). Another possibility is that the species may have diverged in parapatry, with
historical and potentially ongoing high levels of gene flow. The broad area of overlap and
multiple contact zones provide a good system to test the concordance of patterns of
introgression, which can help distinguish parapatric speciation from a secondary contact
zone. In addition, if X chromosomes play an important role in reproductive isolation they
should show consistent patterns of differentiation and reduced introgression across multiple
sympatric populations (Payseur, 2010). Here, I present the first fine-scaled genetic analysis of
these two closely related species. In particular, there were three main aims of this study.
Aims
1. Determine whether the species hybridize and test the extent of introgression
If there is recent hybridization and introgression between the species, then genomic
markers should exhibit different ancestries which could be detected using Bayesian clustering
analysis (Raj et al., 2014) and genomic clines analysis (Gompert & Buerkle, 2009).
Alternatively, if the species are reproductively isolated, then sympatric individuals will be
assigned with a high probability to genetic clusters corresponding to the species groups.
Simulation studies have shown that when secondary contact arises due to a recent range
expansion, introgression is likely to be asymmetrical, with genes from the local species
163
introgressing into the invading species genome (Currat et al., 2008). Therefore, a pattern of
asymmetrical introgression would be consistent with a recent range expansion.
2. Test whether X-linked loci exhibit greater species divergence and population
differentiation than autosomal markers
If introgressive hybridization is detected in sympatry, I predict X-linked loci will
show relatively reduced rates of introgression and greater Fst values compared to autosomal
markers, due to a disproportionate role in reproductive isolation (accounting for differences
in Ne) (Payseur, 2010). In addition, if X chromosomes diverge faster than autosomes, due to
selection (i.e. Faster-X theory) and/or drift (due to smaller Ne) I predict increased population
differentiation at X-linked loci. If no introgressive hybridization is detected X-linked loci
may show similar levels of species differentiation, in sympatric and allopatric populations, as
reproductive isolation may have been attained prior to secondary contact.
Under the dominance theory of Haldane’s rule, species with larger X chromosomes
are more susceptible to X-linked incompatibilities (Turelli & Begun, 1997). The X
chromosome in these species accounts for a substantial portion of their genome: when in
single copy, in males, it represents approximately 20% of the diploid male genome, and when
present in two copies, in females, it represents approximately 30% of the diploid female
genome (K Klappert; unpublished data/pers. comm.). Therefore, if genetic incompatibilities
play an important role in constraining interspecies gene flow, I predict X chromosomes to
contribute disproportionately. Alternatively, if genetic incompatibilities are scattered
throughout the genome we may expect no relative difference between autosomal and Xlinked markers in patterns of introgression or population differentiation (above neutral
expectations).
164
3. Test whether geographic distance predicts patterns of population structure
Isolation by distance (IBD), where genetic distance explains patterns of genetic
divergence, is a common observation in population genetic studies and often arises due to
genetic drift and limited dispersal (Wright, 1943, 1946). The extent to which individuals or
populations deviate from the expected association between geographic and genetic distances
allows for inferences to be made on processes other than IBD that structure population
genetic variation, such as migration, admixture or selection (Wang et al., 2012; Fields et al.,
2015; Knowles et al., 2016). Demography is likely to influence the genome as a whole, while
selection will target specific regions. If demographic processes, such as population
bottlenecks or range expansions, have largely shaped the distribution of genetic variation in
these species, I predict largely concordant geographic patterns of population differentiation
among autosomal and X-linked markers. Range expansions generally lead to a loss of genetic
diversity along the axis of expansion due to recurrent bottleneck effects (Excoffier et al.,
2009). If sympatry reflects a recent range expansion we may expect a pattern of IBD and a
decrease in genetic diversity towards the area of overlap for both autosomal and X-linked
markers.
Overall this analysis aims to bridge an important gap in our knowledge on the
interaction between T. commodus and T. oceanicus in sympatry and more broadly to
determine whether X chromosomes play a pronounced role in population divergence and
delimiting species boundaries.
165
Material & Methods
Sample Collection
Individual crickets were sampled from 16 sites, from March 1st – April 2nd 2013,
encompassing a broad latitudinal transect (~2,500 km across eastern Australia), including
both allopatric and sympatric populations (Figure 5-1). Areas of sympatry between the two
species were located based on published studies (Hill et al., 1972; Otte & Alexander, 1983).
For each population ca. 30 crickets were sampled, except for the populations “KH” and “UQ”
where only 14 and 15 individuals were collected (sample sizes for each population provided
in Table S1).
Genomic Library Preparation
Genomic DNA from cricket heads (n=480, Table S1) was individually extracted and
RNAse treated with the DNeasy Blood and Tissue kit (Qiagen), following the manufacturer’s
instructions for animal tissue. The extracted DNA was subsequently quantified and quality
checked using Nanodrop and Qubit. RAD library preparation was carried out following Baird
et al. (2008), with some modifications. Briefly 250 ng of each DNA sample was digested
with SbfI (New England BioLabs), followed by ligation of barcoded P1 adapters (8 nM).
Each cricket was individually barcoded. Fragments were then sonically sheared to size range
300–700 bp. P2 adapters were ligated to sheared ends. Libraries were PCR amplified and
sequenced on 3 lanes of Illumina HiSeq 2000 and 2500 respectively. This protocol provides
two sets of reads at each RAD site: read 1 extends either side of the restriction site (~ 100bp),
while read 2 sequences are more loosely distributed extending up to ~700 bp from the
restriction site (Davey et al., 2013).
166
Figure 5-1 DNA sample collection sites
Sample collection sites (n=16) across eastern Australia and the distributional range of both
species. T. oceanicus is found across the north (represented by light grey area), while T.
commodus (dark grey areas) is mainly restricted to the south-eastern coast, but has also been
documented in the south west. Both species overlap in an area ~400km long on the mideastern coast (wavy white lines). The orange squares indicate allopatric populations of T.
oceanicus, the green diamonds represent sympatric populations of both species, and the blue
circles indicate allopatric populations of T. commodus. The red dot in the north indicates the
site where I encountered and sampled the putative closely-related species T. marini (Otte &
Alexander, 1983).
167
De novo assembly
The sequences were analysed following the RADmapper approach
[https://github.com/tcezard/RADmapper], based on the Stacks RAD pipeline (Catchen et al.,
2013). Individual samples were demultiplexed with process_radtags from Stacks, allowing
one mismatch in the restriction enzyme recognition site. Read 1 sequences from each sample
were individually clustered using ustacks (parameters –M 2 –N 4). The resulting stacks were
merged across samples per species separately and for all samples together using cstacks with
default parameters (Catchen et al., 2013).
The effect of assembly protocol and variant filtering is an important consideration for
comparing divergent non-model species using RADseq (Lexer et al., 2013; Nadeau et al.,
2014). Three different de novo assembly methods were implemented to determine the most
appropriate method to obtain a maximum number of single nucleotide polymorphisms
(SNPs) and to ensure there was no ascertainment bias. The three different assemblies to
create consensus references were as follows:
i) combined species assembly –individuals were subsampled from all populations and
pooled together;
ii) T. commodus assembly – pooled allopatric T. commodus individuals;
iii) T. oceanicus assembly – pooled allopatric T. oceanicus individuals.
The final merged stacks were filtered to remove potential erroneous stacks by only
retaining the ones supported by a minimum number of individuals (combined species
assembly, n = 200; T. commodus assembly, n = 100; T. oceanicus assembly, n = 50).
Read 1 of all individual samples were mapped back against the consensus sequences
using BWA v0.7.9a (aln/samse algorithm) (Li & Durbin, 2009). The mapping statistics were
generated using samtools flagstat and concatenated. To extend the usable sequence for SNP
calling, the read 2 sequences for each read 1 stack were assembled following the methods of
168
Peng et al. (2012). The resulting read 2 assembly was then merged with the read 1 stack
where possible, otherwise the sequences were concatenated. Variants were called using
SAMtools/BCFtools (v0.1.18) (Li et al., 2009; Li, 2011) and SNPs outputted in Variant
Called Format (VCF) files.
To assess if the different de novo assemblies influenced the downstream analysis I
compared the number of SNPs obtained and the overlap between them. Briefly, the three
different assemblies resulted in a similar number of SNPs and the structuring of population
genetic variation was broadly the same, with no strong sign of species bias. More detailed
comparisons of the number of SNPs returned from the three different assemblies is provided
in the supplementary section (Fig. S1). Therefore, to address the pattern of interspecific
genotypic variation I used the combined species assembly, whereas intraspecific analyses
were done using the species-specific assemblies, in order to maximize the number of
potentially informative markers.
SNP quality control and genetic diversity estimates
The Variant Called Format (VCF) files were filtered using VCFtools (Danecek et al.,
2011) and the following filtering steps were performed: all indels were removed, only sites
with a minor allele frequency greater than or equal to 0.05 were kept, and all individual
genotypes with a quality score less than 20 were recoded as missing (--min GQ 20 –
VCFtools command). Next, all loci that were not present in at least 80% of the individuals
were excluded using the --max-missing VCFtools command. Two individuals from the HB
population were dropped due to low mapping coverage (< 50 % mapped reads). For
population genetic analysis the quality filtered VCF files which contained only biallelic SNPs
were converted to required file types using PGDSpider (Lischer & Excoffier, 2012). In some
of the analyses the number of SNPs varied due to subsetting of the data which changes the
169
proportion of missing data. In addition, in some analyses, individuals with high amounts of
missing data were dropped (>40% missing genotypes). The number of SNPs and individuals
used in each analysis is provided in the results.
Candidate X-linked loci
Candidate X-linked markers were identified based on sex differences in both SNP
genotypes and read coverage (custom script written by Tim Cezard & Judith Risse). As
females contain an additional X chromosome compared to males, X-linked loci should have
on average twice the read depth in females and should not be heterozygote in males. To filter
based on SNP genotypes, the following rules were applied; all male samples must be
homozygous at the locus and at least one female sample is heterozygous. To filter based on
coverage, first the overall coverage was normalised by the total number of reads in each
sample and then male vs. female samples were compared (student t-test), and candidates with
p-values < 0.05 were selected (where coverage is lower in males compared to females). In
addition, the mean fold change in coverage was calculated between male and female samples,
and candidate SNPs within the range of 1.8 - 2.2 fold change were selected. The candidate Xlinked markers were then used in further analysis, for comparisons with the autosomal
markers (SNPs retained after removing candidate X-linked markers).
Population structure and the species boundary
To test for introgressive hybridization and the pattern of genotypic variation within
and between species I used a combination of approaches. The ancestries of individuals were
examined using FastStructure (Raj et al., 2014) which applies a Bayesian method to assign
individuals to genetic clusters and estimates an admixture proportion for each individual.
170
First, I ran the analysis on the total dataset (combined assembly) and then subset the data into
species-specific groups to determine how genetic variation is distributed within each species.
The analysis was run for K = 1-10 groups using simple priors (applies a flat beta-prior over
population specific allele frequencies for each SNP) (Raj et al., 2014). The most likely
number of groups (K) was assessed with the “chooseK.py” script from FastStructure, which
identifies the range of K values that provide the lowest model complexity while maximizing
the marginal likelihood and the minimum number of components needed to explain structure
in the data (Raj et al., 2014). To detect fine-scale genetic structuring of populations the
analysis was rerun using logistic priors (for each SNP applies a logistic normal distribution to
estimate the population specific allele frequency) (Raj et al., 2014), based on the previously
obtained most likely value of K. Multiple replicates (30 - 50) were run and the average from
the top 25 most likely replicates was used, and the admixture proportions were plotted using
the “distruct.py” script.
Patterns of introgression were also tested using a genomic clines approach,
implemented in the R package Introgress (Gompert & Buerkle, 2009, 2010). Introgress
estimates the proportion of species admixture (i.e. the hybrid index), and the level of
interspecific heterozygosity. The hybrid indexes are similar to those from Bayesian admixture
proportions (e.g. Structure like methods (Pritchard et al., 2000)), with the distinction that
parental populations are defined a priori and allele frequencies are estimated. To calculate the
hybrid index, a subset of SNP loci which exhibited large allele frequency differences (>80%)
between allopatric individuals of both species were selected. Individuals from allopatric
populations sampled furthest from the contact zone (AM for T. commodus and DV for T.
oceanicus; Figure 5-1) were used as they can be assumed to represent pure species
individuals. Their use guarded against the potential for interspecies gene flow outside the area
of sympatry masking any signal of hybridization and introgression. Using allopatric
171
individuals as a reference can also help to distinguish incomplete lineage sorting from
introgression (Eaton et al., 2015).
The hybrid index is the proportion of alleles inherited from each of the parental
species, ranging from 0 to 1 (0 = pure T. commodus and 1 = pure T. oceanicus) (Gompert &
Buerkle, 2009). It takes into account uncertainty in the genotypes, such as when loci do not
exhibit fixed species differences (Gompert & Buerkle, 2010). Interspecific heterozygosity is
the proportion of genotypes that are heterozygous for parental alleles (0 = all homozygous, 1
= all heterozygous). If sympatric individuals contain alleles, which are absent (or present at
very low frequencies) in allopatric individuals of the same species but present in allopatric
individuals of the other species, it would be consistent with hybridization and introgression of
genes across the species boundary.
Intraspecific population genetic structure
Genetic summary statistics were estimated for each population, including: observed
heterozygosity (Hobs), expected heterozygosity (Hs), and the inbreeding coefficient (Fis),
using the R packages Adegenet (Jombart et al., 2011; Jombart, 2008) and Hierfstat (Goudet,
2005). Pairwise population FST values were calculated using Arlequin (Excoffier & Lischer,
2010), which estimates the variance in allele frequencies among versus within populations
following the method of Weir & Cockerham, (1984). Significance was tested by 10,000
permutations. To examine population differentiation at X-linked loci, the analysis was run on
datasets containing females only, to avoid the influence of male hemizygosity biasing our
estimates. To test the robustness of the results, for comparing autosomal and X-linked
markers, I also ran the analysis on datasets containing both sexes and females only. The
results were qualitatively the same (Table S4 & S5). A Mantel test was used to test if
172
pairwise FST values for autosomal and X-linked loci were correlated, using the R package
ade4 (Dray & Dufour, 2007). Significance was based on 1,000 permutations.
To summarize the major components of genetic variation among the species a
combination of approaches were used. Principal component analysis (PCA) was applied
using the R packages SNPRelate (Zheng et al., 2012) and Adegenet (Jombart et al., 2010) ,
and individual scores plotted to visualize the distribution of genetic diversity across
populations. The genetic relationships among populations of the three putative species were
also visualized using hierarchical clustering trees for pairwise FST values and bootstrapped
neighbour joining trees based on Nei’s genetic distance in the R package Poppr (version
2.20) (Kamvar et al., 2014; Kamvar et al., 2015).
Geographic distribution of genetic variation
To test if population genetic structure exhibited isolation-by-distance (IBD), the
correlation between pairwise population FST values and Euclidean geographic distances
(estimated from each site’s latitude and longitude coordinates obtained from Google maps)
was examined with a Mantel test. This analysis was conducted separately for each species,
based on the species-specific data subsets. Mantel tests were implemented using the R
package ade4 (Dray & Dufour, 2007) and significance was tested using 10,000 permutations.
In addition IBD was tested using Mantel tests on a matrix of Nei’s genetic distances and a
matrix of Euclidean geographic distances between populations (Fig. S4). A correlation
between geographical distance and genetic differentiation can occur under a wide array of
biological processes which may not involve isolation by distance (Meirmans, 2012). To
distinguish a continuous cline of genetic variation from distant and differentiated populations,
the relationship between geographic distances vs. genetic differentiation was plotted using the
173
kernel density function kde2d from the R package MASS (Venables & Ripley, 2002). This
analysis was applied to both autosomal and X-linked markers separately.
To test whether the species differ in their relationship with geographic distance and
genetic divergence, I estimated and compared the frequency distribution of 1,000 slopes and
intercepts at autosomal and X-linked loci using the R package boot (Canty & Ripley, 2008;
following the methods of Baselga, 2010). The probability of one species having a higher
slope or intercept than the other, at autosomal or X-linked loci, was estimated by the
proportion of the bootstraps for which one species had higher parameters than the other. For
p-values I used the probability of obtaining the opposite result by chance.
174
Results
Population genetic structuring and the species boundary
The FastStructure analysis on the total dataset (16 populations), based on the
combined species assembly, revealed four genetic groups, corresponding to the three putative
species and a northern and southern latitudinal gradient among T. commodus populations
(Figure 5-2.A). There was no evidence for recent species admixture among either T.
commodus and T. oceanicus or T. oceanicus and T. marini. Individuals in sympatric areas had
admixture proportions of either > 0.9 or < 0.01, indicating pure species individuals. An
admixture proportion of ca. 0.5 would indicate F1 hybrids, while >0.25 and <0.75 would be
classified as recent generation hybrids. Similarly, the data subset containing only sympatric
individuals (Figure 5-2B) suggested 2 – 3 most likely groups and no species admixture, even
though it included a larger number of SNPs. Sympatric individuals were clearly assigned to
either T. commodus or T. oceanicus species groups which allowed us to subset the data into
species-specific groups, to further examine intraspecific population genetic structuring.
Intraspecific analysis revealed that for both species the most likely number of groups ranged
between 2 – 3 (Figure 5-2C). T. commodus populations exhibited a clear latitudinal gradient
of genetic variation distinguishing southern from northern populations. In contrast there was
no clear geographic structuring among T. oceanicus populations, with individuals from
multiple populations sharing similar admixture proportions (one exception being the DV
population; Figure 5-2C).
175
Figure 5-2 FastStructure analysis
Plots of individual assignments to the inferred genetic clusters for the most probable K-values
for different data subsets. Populations are arranged in geographic order. A) Combined species
assembly for all populations (n = 478, SNPs = 36,566, K = 4). B) Subset of sympatric
populations (n = 127, SNPs = 60,841, K = 2-3). C) Fine scaled analysis of intraspecific
allopatric and sympatric population genetic structure based on species-specific data subsets in
which individuals were assigned to either species group based on the initial analysis (T.
commodus: n = 269, SNPs = 38,121; T. oceanicus: n = 195, SNPs = 41,928).
The genomic clines analysis also suggested that recent hybridization and introgression
was rare or absent with evidence of strong differentiation between both species (Figure 5-3).
The analysis was based on 2,102 SNPs which exhibited greater than 80% allelic frequency
differences between allopatric populations of both species (selected from 33,183 SNPs). Of
176
these 2,102 SNPs, 754 exhibited almost fixed species differences (> 0.99 frequency
difference). There was a sharp transition in the hybrid index among sympatric individuals,
ranging from: 0–0.07 to 0.99–1, in which 0 = pure T. commodus and 1 = pure T. oceanicus.
Interestingly, there was more variation in the hybrid indices among T. commodus individuals,
which is the species more likely to have recently colonized into this area (Cairns et al., 2010).
Figure 5-3 Genomic clines analysis
Each row represents an individual from sympatry (n=127), while the columns correspond to
individual SNP markers (2,102). Individuals are ordered based on the proportion of ancestry
(hybrid index) ranging from 0 = pure T. commodus to 1 = pure T. oceanicus. The coloured
cells denote an individual’s genotype at a given marker; dark green (P1/P1), green (P1/P2)
and light green (P2/P2). White cells reflect missing genotypes. B) The hybrid index plotted
against interspecific heterozygosity for individuals from sympatric populations. The hybrid
index is measured as the proportion of alleles with T. commodus or T. oceanicus ancestry (01).
177
Principal component analysis
Genotypic variation among the species was clearly differentiated by the first principal
component, which explained 18% of the variation in the data (Figure 5-4A). Individuals of
the putative species T. marini clustered separately from the two other species (red circles
clustered between the two larger groups). PC2 distinguished the southern and northern
populations of T. commodus, but accounted for only 1.18% of the variation. Sympatric
individuals (coloured in green) were clearly discriminated between the two main clusters with
no intermediate individuals. This is consistent with the results from the Bayesian clustering
and genomic cline approaches, suggesting that recent introgressive hybridization is absent. In
separate, species-specific analysis, PC1 and PC2 accounted for a much lower amount of
genotypic variation (1-3%). Among T. commodus populations, PC1 distinguished southern
from northern populations (Figure 5-4B). There was less structuring of T. oceanicus
populations, with considerable overlap among allopatric (orange) and sympatric (green)
populations (Figure 5-4C).
178
Figure 5-4 Principal component analysis
PCA for three different data sets. A) Both species encompassing 16 populations (combined
assembly: n = 470, SNPs =39,238 (LD-pruned SNPs 23,393)). The LD-pruned SNPs reflects
the SNPs retained after removing loci in high linkage disequilibrium. The three clusters
correspond to the putative species: T. commodus, T. oceanicus and T. marini. B) T.
commodus individuals from 11 populations, encompassing 8 allopatric and three sympatric
populations (T. commodus assembly; n = 265, SNPs = 40,728 (LD-pruned SNPs 26,742). C)
T. oceanicus individuals from 7 populations (T. oceanicus assembly; n = 191, SNPs = 44,282
(LD-pruned SNPs 27,350). Each point represents an individual cricket and colours
correspond to sampling sites.
179
Species differentiation: Autosomal versus X-linked markers
There was strong differentiation amongst T. commodus and T. oceanicus at autosomal
loci with pairwise population FST values between 0.39 – 0.44 (all comparisons highly
significant; p < 0.001; 10,000 permutations) (Fig. S2; Table S3). Interestingly, the third
putative species T. marini had generally higher levels of FST when compared with T.
oceanicus (mean FST - 0.407) than with T. commodus (mean 0.369) (Figure 5-5; Fig. S2;
Table S3). For X-linked loci, there was greater differentiation for two out of three species
comparisons compared to at autosomal loci: T. commodus vs. T. oceanicus (mean FST 0.435) and T. oceanicus vs. T. marini (mean FST - 0.484) (Table S3). In contrast, T. marini
had lower levels of FST for X-linked loci when compared with T. commodus (mean FST 0.119; Figure 5-5, Table S3). This is interesting, as T. commodus is geographically more
distant from T. marini, whereas T. oceanicus overlaps in sympatry.
Figure 5-5 Hierarchical clustering trees
Pairwise Fst at autosomal and X-linked markers for the 16 populations encompassing the
three putative species. Colours are consistent with Figure 5-1. Neighbour joining trees based
on Nei’s genetic distances with bootstraps are provided in the supplementary (Fig. S5).
180
Reduced population differentiation at X-linked loci
Contrary to our prediction that X-linked loci would show more pronounced
population genetic structuring, populations were more weakly differentiated at X-linked vs.
autosomal loci within both species (Table 5-1, Figure 5-5; Figure 5-6). The discordance
between autosomal and X-linked markers was particularly great among T. oceanicus
populations, with nearly all population pairs showing reduced FST values at X-linked loci
(mean FST: autosomes 0.018 vs. X-linked loci 0.005; Table 5-1). Indeed, there was a negative,
but not significant, correlation between autosomal and X-linked markers in the level of
genetic differentiation among T. oceanicus populations (Mantel test: r -0.178, p = 0.792). For
autosomal loci, nearly all T. oceanicus populations were differentiated (albeit weakly) (range
of FST values 0.001 – 0.057; p < 0.05). In contrast, there was very little population structuring
at X-linked loci (Table 5-1). Moreover, there was a clear reduction in heterozygosity across
all T. oceanicus populations at X-linked loci (Table 5-2, analysis on females only).
Geographic distance was a weak predictor of population differentiation at autosomal
loci for T. oceanicus (Mantel test: FST vs. geographic distance; r = 0.564, p = 0.021), but not
for X-linked loci (Mantel test: r = -0.038, p = 0.527) (Figure 5-6, Table 5-3). Overall there
was a striking reduction in differentiation among T. oceanicus populations for X-linked vs.
autosomal loci (Figure 5-6). Although there was a smaller number of individuals perpopulation for X-linked loci (Table 5-2), which could increase the sampling variances
(Holsinger & Weir, 2009), the results were robust when tested on a dataset including males
and females (and females only) for both marker types (Table S4, S5).
In contrast, T. commodus populations exhibited only a slightly reduced pattern of
differentiation at X-linked vs. autosomal loci (mean pairwise FST values: Autosomes – 0.027
vs. X – 0.023) (Table 5-1). The level of genetic differentiation among T. commodus
populations was positively correlated at autosomal and X-linked markers (Mantel test: r =
181
0.401, p = 0.016). Population structure was pronounced for both marker types (pairwise FST
values ranging from 0.001–0.105 for autosomal loci and -0.001–0.071 for X-linked markers)
and most comparisons were significant at p < 0.05 (Table 5-1). Population differentiation
showed a sharp increase with geographic distance for autosomal loci (Mantel test: r = 0.921,
p < 0.001), but was much weaker for X-linked loci (r = 0.352, p = 0.029) (Figure 5-6).
The relationship between geographic distance and genetic differentiation (FST)
differed between the two species. For autosomal loci the slope was higher in T. commodus
(5.73 x 10-5) than in T. oceanicus (2.45 x 10-5) (significance tested using 1,000 bootstraps: p
< 0.001). Similarly, at X-linked loci, the slope was higher in T. commodus (1.62 x 10-5) than
in T. oceanicus (-8.88 x 10-7) (p = 0.002). Intercepts were higher at autosomal loci in T.
oceanicus (4.14 x 10-3) than T. commodus (-9.5 x 10-3) (p = 0.003), but were not significantly
different at X-linked loci. The results were consistent whether genetic differentiation was
measured using FST or Nei’s genetic distance (Fig S4). The greater geographic range
encompassed by T. commodus may have partly contributed to the species difference for
autosomal loci, but is unlikely to explain the discrepancy between the species at X-linked
markers.
182
Table 5-1 Pairwise population FST: Autosomal vs. X-linked loci
Autosomal markers below the diagonal and X-linked markers above the diagonal. For Xlinked markers only females were included (FST was also estimated with males included and
resulted in a similar pattern (Fig. S2, Table S4, S5). Significance was tested with 10,000
permutations and significant comparisons highlighted in bold. Populations are arranged in
geographic order with AM the most southern population and DV the most northern.
183
Table 5-2 Population genetic summary statistics for autosomal and X-linked loci
For X-linked loci only females were included to avoid bias introduced by male hemizygosity.
Ho is the observed heterozygosity, Hs the within population gene diversity (or sometimes
referred to as expected heterozygosity), and FIS the inbreeding coefficient which can range
from -1 to 1 (low inbreeding - high inbreeding respectively), and which estimates the mean
reduction of heterozygosity within an individual due to assortative mating within a subpopulation. The statistics were estimated using Adegenet (Jombart, 2008; Jombart & Ahmed,
2011) and Hierfstat (Goudet, 2005) which follow the formulas laid out in Nei, (1987). “Area”
of sampling indicates allopatry (“allo”) or sympatry (“sym”).
Pop
Area
AM
Allo
BN
Allo
CC
Allo
MV
Allo
BL
Allo
CH
Allo
UQ
Allo
SV
Allo
HB
Sym
TS
Sym
RH
Sym
Mean FST
HB
Sym
TS
Sym
RH
Sym
YP
Sym
PL
Allo
JC
Allo
DV
Allo
KH
Allo
Mean FST
Species
T.com
T.com
T.com
T.com
T.com
T.com
T.com
T.com
T.com
T.com
T.com
T.oc
T.oc
T.oc
T.oc
T.oc
T.oc
T.oc
T.oc
Autosomal loci (♀ & ♂)
Samples
Ho
Hs
FIS
♀&♂
30
34
34
30
33
30
15
34
5
10
10
0.027
23
24
23
32
31
34
24
14
0.018
0.212
0.221
0.208
0.211
0.198
0.204
0.204
0.202
0.198
0.195
0.191
0.230
0.228
0.240
0.242
0.230
0.219
0.224
0.228
0.231
0.229
0.220
0.084
0.049
0.139
0.132
0.136
0.074
0.085
0.121
0.095
0.126
0.104
0.230
0.221
0.214
0.218
0.221
0.220
0.216
0.168
0.241
0.243
0.239
0.235
0.247
0.248
0.249
0.232
0.055
0.095
0.107
0.080
0.112
0.120
0.129
0.192
184
Sex-linked loci (♀ only)
Samples
Ho
Hs
FIS
♀
15
16
17
16
17
11
10
17
3
6
4
0.023
13
12
11
12
17
12
17
8
0.005
0.225
0.229
0.213
0.225
0.193
0.201
0.203
0.199
0.205
0.19
0.203
0.227
0.227
0.223
0.229
0.208
0.212
0.216
0.214
0.212
0.218
0.204
0.01
-0.008
0.038
0.028
0.071
0.043
0.048
0.06
-0.019
0.092
-0.005
0.116
0.117
0.118
0.109
0.107
0.111
0.117
0.101
0.122
0.120
0.129
0.115
0.114
0.118
0.119
0.121
0.051
0.015
0.073
0.055
0.055
0.052
0.009
0.126
Figure 5-6 Genetic differentiation and geographic distance
Relationship between population genetic differentiation (mean FST) and geographic distance
(Euclidean) at autosomal and X-linked markers for both species. Comparisons between the
species for the slopes and intercepts are in the main text.
Table 5-3 Mantel test results
The strength and significance of the correlation between autosomal and X-linked pairwise FST
values and also the geographic relationship. Significance tested using 10,000 permutations.
Species
Marker
type
Samples SNPs
T. commodus
Autosomal 265
X-linked
132
T. oceanicus
Autosomal 200
X-linked
102
Autosomal vs. X-linked
Geographic association
40260
758
r = 0.401 , p=0.016*
r = 0.92 , p < 0.001***
r = 0.352 p = 0.029*
44770
619
r = -0.178 , p = 0.792
r = 0.564 , p = 0.021*
r = -0.038 , p = 0.527
185
Discussion
Geographic areas where interfertile species overlap provide important testing grounds
for determining how porous species boundaries are and to identify the main barriers to gene
flow. I examined the relative role of autosomal and X-linked loci in population divergence
and delimiting the species boundary, for two closely related field cricket species T.
commodus and T. oceanicus.
Discordance among markers: Reduced population differentiation at X-linked loci
Theory predicts reduced genetic diversity on X chromosomes and increased
population differentiation relative to autosomes, due to differences in selection and drift on
these regions (Charlesworth et al., 1987; Vicoso & Charlesworth, 2006; Mank et al., 2010).
However, contrary to such predictions, I found reduced population differentiation at X-linked
vs. autosomal loci (Table 5-1; Figure 5-5; Figure 5-6). Deviations from expected patterns of
diversity on sex chromosomes, beyond that expected due to differences in Ne (Ramachandran
et al., 2004), are usually interpreted as resulting from selection or demographic effects, such
as sex-biased dispersal (Keinan et al., 2009). Numerous studies in Drosophila have reported
higher levels of genetic differentiation at X-linked vs. autosomal loci, both between
populations and between semi-species (Ford & Aquadro, 1996; Charlesworth, 1998; some
exceptions, Machado et al., 2016; Sedghifar et al., 2016). X-linked loci may exhibit higher
population differentiation than autosomal loci due to a reduction of diversity, driven by
bottlenecks, selective sweeps or background selection (Keinan et al., 2009). As X-linked loci
occur in regions of low recombination, linkage disequilibrium is likely to be strong and Xlinked diversity may be reduced within populations due to selection for locally beneficial
186
alleles (selective sweeps) or against deleterious mutations (background selection)
(Charlesworth et al., 1997). Increased population differentiation may occur if populations
become fixed for different variants (Ford & Aquadro, 1996). Alternatively, a selective sweep
may spread across populations which would reduce overall population differentiation (Hall,
2004; Nolte et al., 2013). The general signature of a selective sweep is reduced diversity
across a genetic region, whose size will depend on the strength of selection and the rate of
recombination. T. oceanicus populations had reduced diversity at X-linked loci (lower
observed and expected heterozygosity, Table 5-2), which could suggest a selective sweep or
background selection across the set of populations, leading to reduced differentiation.
A complex interplay of sex-biased processes such as: mutation, drift, selection,
migration, demography and mating strategies are likely to affect the relative rate of
divergence among autosomes and sex chromosomes (Kirkpatrick & Hall, 2004; Ellegren,
2009). As X chromosomes spend two thirds of their time in females, a sex-bias in mutation
rates could lead to substantial differences in genetic diversity between these two regions of
the genome (Mank et al., 2010). Increased mutation rates in males have often been observed
across many animals and have been suggested to occur due to increased meiosis during
spermatogenesis compared to oogenesis (Li et al., 2008). If the mutation rates are higher in
males than in females, it will increase diversity on the autosomes relative to the X
chromosomes, because of the fact that Xs spend only a third of their time in males,
potentially leading to greater population differentiation at autosomal vs. X-linked loci.
However, it is unlikely that mutation bias alone can explain our results, as the species
differed in the extent of population differentiation at X-linked loci, and sex-biased mutation
rates are unlikely to differ substantially between the species.
187
Additional factors that are likely to contribute to a reduction in X chromosome
diversity and differentiation are polygynous breeding systems, female-biased dispersal, and
unequal sex ratios (Ellegren, 2009; Mank et al., 2010). Polygynous mating systems, where
males mate multiply with females and there is a high variance among males in reproductive
success, will reduce the effective population size of males. Increasingly skewed male
reproductive success will decrease the relative genetic diversity of X chromosomes (Ellegren,
2014). If the species differ in the extent of male-biased reproductive success this could
contribute to the greater reduction in differentiation for X-linked loci among T. oceanicus
populations compared to T. commodus populations. Indeed, previous studies have shown that
males of both Teleogryllus species mate multiply and there is a strong skew in reproductive
success, with the bias being higher in T. oceanicus (Simmons & Beveridge, 2010b).
However, high skews in male reproductive success are also likely to generate strong sexual
selection on males which could increase diversity on X chromosomes and by extension
stronger population differentiation.
Sex-biased dispersal has been commonly observed among birds and mammals and
can alter the relative Ne between sex chromosomes and autosomes (Goudet et al., 2002).
However, surprisingly little is known about sex-biased dispersal among insects. If males
disperse more than females, autosomal gene flow may exceed that of X-linked loci, as males
carry only a single X chromosome but have two autosomes (Keinan et al., 2009). In contrast,
under female biased dispersal, gene flow should have equal effects across autosomal and Xlinked regions (Ellegren, 2009). Increased dispersal among T. oceanicus populations may
explain the overall weak population structuring (for both autosomal and X-linked markers),
and if females disperse at a much higher rate than males, this will tend to equalize the effect
of gene flow and reduce diversity on both autosomes and X chromosomes.
188
Changes in population size will also affect the relative diversity of autosomes and sex
chromosomes. Pool & Nielsen, (2007) suggested that population contractions can reduce Xlinked diversity, as the smaller Ne makes X chromosomes more susceptible to bottleneck
effects. The converse is that population expansions will tend to equalize autosomal and X
chromosome diversity. Populations of both species are likely to have undergone cycles of
inter-glacial contractions and expansions (Chapple et al., 2011), which may have
differentially affected the species and contributed to the observed discordance. A
combination of the above factors may contribute to the unexpected pattern of reduced
differentiation at X-linked loci among T. oceanicus populations. As the pattern is consistent
across all T. oceanicus populations it suggests the driving factor is likely to be common
across the species range rather than a localized environmentally induced effect.
What factors determine the frequency of hybridization?
When interfertile species overlap in sympatry, the strength of pre- and post-mating
barriers will largely determine the frequency of hybridization. Most reproductive barriers are
believed to accumulate indirectly through pleiotropic effects during adaptation to new
environments (Rogers & Bernatchez, 2007) or drift (Shuker et al., 2005). The strength of both
pre- and post-mating barriers have been shown to increase steadily with the level of genetic
divergence between species (Coyne & Orr, 1989, 1997; Price & Bouvier, 2002). The
frequency of hybridization may also exhibit a similar pattern (Edmands, 2002; Mallet, 2005).
The apparent absence of recent introgressive hybridization among T. commodus and T.
oceanicus may be unsurprising therefore, given the high level of genetic divergence (FST
values between species 0.39 – 0.52; Fig. S2, Table S3). However, hybrids are easily produced
in the lab (Hogan & Fontana, 1973; Hoy et al., 1977; Moran et al., in review) and the species
189
are in very close contact, with males of both species singing side by side across a large area
of sympatry, with no evidence of environmental differentiation.
Bimodal distributions of pure parental species have been suggested to be maintained
due to strong assortative mating (or fertilization) and/or habitat segregation, rather than
intrinsic genetic incompatibilities (Jiggins & Mallet, 2000). Indeed, Jiggins & Mallet (2000)
suggest that the overall level of genetic divergence is a poor predictor of species integrity,
with many unimodal hybrid zones occurring in taxa that are highly divergent in chromosome
structure and nucleotide sequence (e.g. Caleida and Vandiemenella grasshoppers which differ
in chromosomal inversions: Shaw et al. (1993); Kawakami et al. (2009)). Whereas in many
bimodal hybrid zones, taxa may be morphologically indistinguishable and/or only weakly
genetically differentiated (Johannesson et al., 1993). In some bimodal hybrid zones no
postyzygotic incompatibilities have been detected, such as in the ground crickets
Allonemobius fasciatus and A. socius, where the main barrier to gene exchange appears to be
conspecific sperm precedence which facilitates species coexistence over broad areas of
overlap (Howard et al., 1998; 2002)..
The apparent absence of recent hybridization between T. commodus and T. oceanicus,
across a broad area of sympatry (ca. 400km) suggests strong assortative mating which may
have arisen prior to secondary contact. A combination of both pre and post-mating barriers
have been detected in our study species, including discrimination against heterospecifics
during long and close range courtship behaviour (Hill et al., 1972; Hennig & Weber, 1997;
Bailey & Macleod, 2013; Chapter 3), reduced rates of heterospecific mating, and hybrid
female sterility in both cross directions (Fontana & Hogan, 1969; Hogan & Fontana, 1973;
Chapters 3 & 4). These studies have been based on allopatric populations, and selection, due
to the negative fitness effects of hybridization (e.g. hybrid female sterility), may have
enhanced assortative mating in sympatry through reinforcement (Butlin, 1987, Coyne & Orr,
190
2004). Little is known about ecological differences between these species. Divergent
ecological and sexual selection could be strong enough to overcome the potential for
hybridization and introgression.
Ancient or more recent timing for secondary contact?
The timing of secondary contact among our study species is unknown (alternatively
they may have diverged in parapatry) and could reflect a recent range expansion of either
species, promoted by human disturbance (Cairns et al., 2010) or a more ancient Holocene
expansion from glacial refugia (Moritz et al., 2009). Range expansions are predicted to lead
to a loss of genetic diversity along the axis of expansion due to recurrent bottleneck effects
(Excoffier et al., 2009). T. commodus populations exhibited a clear pattern of isolation by
distance with a steady increase in genetic divergence with distance (Figure 5-6). However,
there was only a slight decrease in heterozygosity among sympatric T. commodus
populations, which may be an artefact of the small sample sizes (Table 5-2). A rapid
population expansion after colonization of the new area may have masked the signal of a
spatial expansion (Excoffier et al., 2009). The pattern of isolation by distance may reflect
differentiation along an environmental gradient or limited migration. Indeed, northern and
southern populations of T. commodus differ in diapausing behaviour which is associated with
the environmental gradient (Hogan & Fontana, 1973). In addition, northern and southern
populations have been noted to differ in chromosomal structure with inversion
polymorphisms being common (Fontana & Hogan, 1969), which may influence gene flow
among populations and the efficiency of adaptive divergence (Noor et al., 2001; Feder &
Nosil, 2009). Distinguishing whether the pattern of IBD reflects adaptation along an
environmental gradient or limited dispersal requires further analysis such as tests for
191
environmental association (de Villemereuil & Gaggiotti, 2015; Rellstab et al., 2015) and
clinal analysis to examine the strength and pattern of selection (Machado et al., 2016).
The position of a contact zone (or hybrid zone) can inform about the processes
maintaining it, such as intrinsic selection due to genetic incompatibilities or extrinsic factors
such as environmental determinants (Barton & Hewitt, 1985). Interestingly, the two contact
zones encompassed in this study broadly overlap with areas where numerous taxa, including
grasshoppers, birds, frogs and reptiles, hybridize (Ford, 1987; Shaw et al., 1993; Chapple et
al., 2011; Moritz et al., 2009; Singhal & Moritz, 2013). Shared regions where sibling species
overlap and hybridize are known as suture zones and are often believed to indicate regions of
secondary contact after expansion from glacial refugia (Anderson, 1948; Remington, 1968;
Hewitt, 2000; Hewitt, 2001; but see Swenson & Howard, 2004). In the wet tropics of
northeastern Australia, where I encountered the third putative species T. marini, an unusually
high level of cryptic species diversity has been characterized, mainly in amphibians and
reptiles, and appears to be due to episodes of contraction and expansion during Quaternary
climatic cycles (Hewitt, 2000; Moritz et al., 2009; Singhal & Moritz, 2013). Consistent with
this area being a hotspot for species diversity, Shaw et al. (1993) described a species of
Caledia grasshoppers localized specifically to the Daintree region (DV and KH population).
T. marini has only been previously described once, and aside from its distinct calling song
(distinguished by a low carrier frequency of ca. 3kHz; Chapter 1), the males have reportedly
divergent genitalia (Otte & Alexander, 1983). The genomic results definitively confirm its
distinct species identity, sharing a similar level of divergence between the other species. The
absence of species admixture suggests it is reproductively isolated from T. oceanicus, which
it co-occurs with over a small area, with males found singing together in the same area.
Overall the concordant position of these contact zones with many other taxa suggests
192
secondary contact amongst the study species may predate large scale human disturbances,
thereby indicating an older interaction driven by post-glacial range expansions.
Conclusions
Sex chromosomes are predicted to exhibit increased population differentiation and
species divergence compared to autosomes, due to increased selection (e.g. exposure of
beneficial recessive mutations and sexual antagonistic selection) and drift (smaller Ne). In
line with the prediction, X-linked loci exhibited greater between species differentiation
compared to autosomal loci. Interestingly, species differentiation at X-linked loci appears to
be higher among sympatric populations suggesting a potential role for X chromosomes in
reproductive isolation. There was no evidence for recent hybridization or introgression
between the species suggesting the species are currently reproductive isolated. However, in
contrast to the higher level of genetic divergence between the species at X-linked loci, within
species analysis revealed a striking pattern of reduced population differentiation at X-linked
loci compared to the autosomes. A complex combination of selective sweeps (or background
selection) and sex-biased processes, which alter the relative Ne of autosomes and X
chromosomes, are likely to contribute to this unexpected discordance, such as female-biased
dispersal and a polygynous mating system. Further research into the causes and consequences
for reduced population differentiation at X loci, is likely to reveal important processes
involved in the evolution of X chromosomes and reproductive isolation.
193
Supplementary
Table S1. Sampled sites in eastern Australia (AU), population codes, species identity,
geographic coordinates and sample sizes for genomic data.
Population
Altona Meadows, AU
Bairnsdale, AU
Cooma Creek, AU
Moss Vale, AU
Bluey's Beach, AU
Coff's Harbour, AU
Brisbane, AU
Maleny, AU
Hervey Bay, AU
Tannum Sands, AU
Rockhampton, AU
Yeppoon, AU
Mount Pleasant, AU
Townsville, AU
Daintree, AU
Daintree (KH), AU
Pop
AM
BN
CC
MV
BL
CH
UQ
SV
HB
TS
RH
YP
PL
JC
DV
KH
Species
T. commodus
T. commodus
T. commodus
T. commodus
T. commodus
T. commodus
T. commodus
T. commodus
T. o / T. c mix
T. o / T. c mix
T. o / T. c mix
T. o / T. c mix
T. oceanicus
T. oceanicus
T. oceanicus
T. marini
194
Latitude
Longitude
-37.8827
-37.8264
-36.2353
-34.4867
-32.3497
-30.295
-27.4951
-26.7528
-25.3008
-23.9627
-23.3835
-23.1444
-21.1129
-19.3288
-16.249
-16.249
144.7834
147.6378
149.118
150.3736
152.5357
153.1167
153.0123
152.8472
152.8631
151.332
150.5067
150.7635
149.1558
146.7599
145.3223
145.3223
Female
DNA
Male
DNA
16
14
17
17
17
17
16
14
17
17
16
17
10
5
17
17
16
14
17
17
17
15
16
17
17
17
17
17
16
9
8
6
250
230
All = 480
Comparing the three different assemblies
If the species examined are highly divergent, there may be a large discrepancy in the number
of SNPs returned for the three different assembly methods. In particular, species-specific or
rare variants may be lost which may affect downstream analysis. This is an important
consideration for merging intraspecific and interspecific studies without a reference genome.
Table S2. Comparison of the number of SNPs obtained from the three different assemblies at
different thresholds of missing data allowed. First three rows indicate SNP numbers when
filtering is applied across data sets containing all populations (individuals from the KH
population were removed due to presence of the third species T.marini). The bottom six rows
indicate number of SNPs returned after subsetting into species specific groups. Filtered SNPs
indicates the number of SNPs obtained after removing sites which were missing across more
than 80% (0.8), 50% (0.5), 20% (0.2) of individuals.
Assembly
Combined
T. oceanicus
T. commodus
Combined
T. oceanicus
T. commodus
Total data
Species subset
T. oceanicus
T. commodus
T. oceanicus
T. commodus
T. oceanicus
T. commodus
SNPs
1,158,094
1,199,268
1,585,852
Samples
464
464
464
195
269
195
269
195
269
195
Filtered
SNPs
(0.8)
Filtered
SNPs
( 0.5)
Filtered
SNPs
(0.2)
178,062
186,851
223,403
103,958
104,388
108,795
36,500
36,653
33,186
38,708
41,757
41,928
35,972
35,131
38,121
Comparing the De novo assemblies
The Venn diagrams indicate that for species comparisons the three approaches result in a
similar number of shared SNPs, while for intraspecific analysis the choice of assembly has a
greater effect on the number of species specific SNPs obtained. A higher number of SNPs
unique to each species group were obtained when using the corresponding species specific
assembly. The combined assembly may be slightly biased towards T. commodus due to the
larger number of these individuals sampled. T.marini shares a low number of variants with
both species. The large number of species-specific variants may represent either fixed species
differences or sites that are missing (or due to allelic drop out) or were removed during
filtering as they were at a low frequency.
Figure. S1. Venn diagram illustrating the overlap of variants amongst the three putative
species for the three different assembly methods. Text box indicates the total number of SNPs
per group (after filtering). Individuals were assigned to their species group based on their
group assignment in the Bayesian clustering analysis. VCF files were filtered to remove
variants with > 0.5 missing data and comparisons were implemented using bcftools.
196
Figure. S2. Population pairwise FST for all 16 populations, based on combined assembly (n
=478, SNPs = 11,199). Individuals from sympatry assigned to either T. commodus or T.
oceanicus groups (T.com suffix or T.oc suffix). Individuals from KH population assigned to
T.marini are indicated with the suffix “T.mar”.
197
198
geographic order with AM the most southern population and KH the most northern. T.marini individuals are denoted as “KH_T.marini”.
females were included in estimates of FST. Colours correspond to increasing FST values (blue – low, red = high). Populations are arranged in
Table S3. Pairwise population FST comparisons for autosomal (below the diagonal) and X-linked markers (above the diagonal). Both males and
Figure. S3. FastStructure plot based on X-linked markers for K =2 - 3.
Table S4. Pairwise population FST comparisons for autosomal (below the diagonal) and Xlinked markers (above the diagonal). In contrast to Table 2 provided in the text, estimates of
FST among X-linked markers here encompass both males and females. Nonsignificant
comparisons highlighted in red.
199
Table S5. Pairwise population FST comparisons for autosomal (below the diagonal) and Xlinked markers (above the diagonal). In contrast to Table 2 and S4, only females included in
estimates of FST at both autosomal and X-linked markers. Significance was tested with 10,000
permutations and nonsignificant comparisons highlighted in red.
200
Figure S4. Relationship between geographic distance (Euclidean) and genetic distance
(Nei's) for autosomal and X-linked markers of both species. Significance was tested using
Mantel tests with 10,000 permutations.
201
Figure S5. Neighbour joining trees based on Nei’s genetic distances with bootstraps (1000
permutations) for both species autosomal and X-linked markers. Colours correspond to map
of sampling sites in Fig. 1.
Procrustes Analysis: PCs vs. geographic distances
There has been considerable debate about the reliability of Mantel tests and the
interpretability of results (e.g. distinguishing IBD from a hierarchical island model)
(Meirmans, 2012). A recently developed approach testing for an association between genetic
differentiation and geography combines principal component analysis and Procrustes analysis
(Wang et al., 2010; Wang et al., 2012). If populations conform to a strict IBD model, the
positioning of population clusters when projected onto a 2D principal component space
should match with their geographic distributions (provided the number of SNPs is evenly
distributed) (McVean, 2009; Fields et al., 2015; Knowles et al., 2016). Procrustes analysis
finds an optimal transformation to maximize the similarity between the genetic variation and
the geographic locations while preserving the relative distances between points within both
matrices. The statistical association is tested by permutations and the pattern of genetic and
geographic variation visualized by plotting the PC scores onto the geographic map. The
extent and direction that individuals or populations deviate from their geographic sampling
locations allows for inferences to be made on processes other than IBD, that structure
202
population genetic variation such as migration, admixture or selection. This analysis was
conducted using PC scores (PC1 & PC2) obtained using the Adegenet R package (Jombart et
al., 2010) combined with the geographic coordinates (latitude and longitude) of the sampling
locations obtained from Google maps. Procrustes analysis was implemented using the
procrustes and protest functions in the R packages MCMCpack (Martin et al., 2011) and
Vegan (Oksanen et al., 2008). The statistical association between genes and geography was
tested using 100,000 permutations, where geographical locations are randomly permutated
among sampling sites.
Figure. S6. Superimposed on the geographical maps are the major axes of genetic variation
(PC1 and PC2), which have been rotated to maximize similarity with geographic distance
(Procrustes analysis).
203
Table S6. Mantel test results for the strength and significance of the association between PC
scores with geographic distance. Significance tested using 10,000 permutations.
Species
Marker type
Samples
SNPs
Geographic association
T. commodus
Autosomal
265
40,728
r = 0.734, P < 0.0001
200
44,941
r = 0.4 , P < 0.0001
T. oceanicus
Figure. S7. Mean read depth for candidate X-linked loci. Read depth should be on average
2x higher in females than males.
204
Chapter 6
General discussion
What can the contact zone between T. oceanicus and T. commodus
tell us about speciation?
In this thesis I used behavioural and genomic approaches to examine prezygotic and
postzygotic barriers which contribute to reproductive isolation between two closely related
field cricket species, Teleogryllus oceanicus and T. commodus. From the results we can
conclude that the species are currently reproductively isolated (Chapter 5), even though they
are partially interfertile (Chapters 3 & 4) and co-occur over a broad area of sympatry, with no
obvious spatial or temporal segregation. Calling song has long been assumed to be the
principal barrier maintaining the species boundary. However, I identified numerous
additional pre-mating and post-mating barriers (Chapters 3 & 4). The presence of multiple
reproductive barriers may partly explain the maintenance of strong reproductive isolation in
sympatry. The absence of evidence for contemporary interspecies gene flow suggests that
selection against hybrids is unlikely to be occurring. However, if interspecies matings are
costly, such as it reduces male or female fecundity or viability, selection might enhance
assortative mating. There was weak support for reproductive character displacement for song
in T. oceanicus males from sympatry (Chapter 2). CHCs from T. oceanicus males and
females differed between sympatry and allopatry, but further work is required to determine if
this reflects species interactions that have increased prezygotic isolation (i.e. reinforcement),
or adaptation to an environmental gradient. The weak evidence for any reinforcement and the
apparent absence of recent hybridization or introgresssion suggest that pre-mating barriers,
205
such as song and CHCs, may have diverged sufficiently prior to secondary contact, to
maintain species integrity.
Contrary to my prediction, X chromosomes did not appear to disproportionately
contribute to post-zygotic barriers (Chapter 4). In particular, hybrid female sterility is not due
to an X-X incompatibility. Instead, autosomal incompatibilities appear to supersede X-linked
incompatibilities (although X-autosomal incompatibilities cannot be completely ruled out).
X-linked loci exhibited an unexpected pattern of reduced population differentiation within
species but increased species divergence, in particular in sympatry (Chapter 5). I discuss the
implications of these findings for the study system and for evolutionary biology more
generally.
What prevents hybridization and fusion of interfertile species?
One of the most surprising things about this contact zone is the absence of
hybridization or introgression (Chapter 5). This is unexpected as hybrids are easily produced
in the lab (albeit hybrid females are sterile) and the species are in very close contact across an
extensive area of sympatry, with no known ecological distinctions. The presence of multiple
behavioural barriers and intrinsic hybrid sterility may be sufficient to prevent gene flow
(Chapter 3 & 4). However, most hybrid zones persist in the presence of multiple reproductive
barriers, suggesting barriers are often leaky and even a few hybrids can provide a bridge for
alleles to pass between species (Maroja et al., 2009; Veen et al., 2011; Tyler et al., 2013;
Tyler et al., 2013; Maroja et al., 2014). The maintenance of reproductive isolation would
require strong assortative mating (or fertilization) and/or habitat segregation (Jiggins &
Mallet, 2005).
206
An interesting example of strong cryptic assortative fertilization comes from the
ground crickets Allonemobius fasciatus and A. socius, which form a mosaic hybrid zone
across the north-eastern United States. Extensive research failed to find any premating
barriers (calling song, mate preferences, habitat, phenology) or postzygotic barriers (Howard
et al., 2002). The only reproductive barrier identified that contributes to isolating the species
is conspecific sperm precedence - CSP (the preferential use of sperm from conspecific males
in the fertilization of eggs) (Howard et al., 1998). Among polyandrous species, CSP is
believed to be widespread and to evolve rapidly, possibly due to intraspecific sexually
antagonistic selection (Howard, 1999; Gavrilets, 2000; Howard et al., 2002). Indeed, CSP has
been suggested to be one of the earliest reproductive barriers to arise amongst incipient
species (Howard, 1999). If females mate multiply in the presence of strong CSP, the relative
costs of heterospecific matings can change, in some cases becoming higher for males rather
than females, which has important implications for the likelihood of reinforcement and the
targets of selection (Marshall et al., 2002; Shaw & Mendelson, 2013). Therefore in many
contact or hybrid zones, CSP may be important in species isolation and could explain why
reinforcement or reproductive character displacement does not often occur (Butlin, 1989;
Howard, 1993; Marshall et al., 2002).
Sperm utilisation provides an important avenue of future research in this study
system, as females of both Teleogryllus species are known to mate multiply (Simmons &
Beveridge, 2010). I predict that CSP plays an important role in maintaining the species
boundary, which may also explain the apparent absence of strong reproductive character
displacement (Hill et al., 1972; Chapter 2) or ecological character displacement (although no
studies have explicitly examined this). A broader question worth exploring is whether
multiple mating by females, in the presence of CSP, could be selected for in areas of
sympatry to reduce costly interspecific mismatings. I am unaware of any study which has
207
compared the propensity of females to mate multiply between sympatric and allopatric
populations.
Is ecological differentiation a requirement for species coexistence?
How important is ecological differentiation for the persistence and coexistence of
closely related species? The competitive exclusion principle precludes the stable coexistence
of closely related species in the absence of ecological differentiation (Hardin, 1960). When
sibling species occupy the same niche, competition for resources is predicted to be high,
therefore selection may drive ecological character divergence (Denno et al., 1995; Pfennig et
al., 2006; Pfennig & Pfennig, 2009; Ortiz-Barrientos et al., 2009) or extinction of the rarer
species (Liou & Price, 1994; Mallet, 2008). A comparative study of closely related species,
encompassing a broad range of taxa, has shown that the extent of ecological divergence
(differences in diet, habitat use and body size) is positively associated with the strength of
reproductive isolation (Funk et al., 2006). This has been suggested to reflect ecological
divergence facilitating reproductive isolation, although it could be vice versa (Funk et al.,
2006). Interestingly, ecological divergence was more strongly associated with prezygotic
barriers than postzygotic barriers across most of the taxa, suggesting traits involved in
prezygotic barriers may be more ecologically dependent.
The importance of ecological differentiation in facilitating species coexistence will
depend on the availability of resources and hence the strength of competition in sympatry.
The extent of species differences in resource utilization prior to secondary contact may be
important in determining the outcome of species interactions (Ortiz-Barrientos et al., 2009).
In the speciose group of Hawaiian endemic Laupala crickets, which have undergone an
extremely rapid diversification (Mendelson & Shaw, 2005), there are no obvious ecological
208
distinctions among the best studied species pairs (Otte, 1989). Instead, sexually selected
traits, namely calling song and female preferences, have been suggested to be sufficient in
maintaining the species boundaries (Shaw & Parsons, 2002; Shaw & Danley, 2003; Shaw et
al., 2007; Wiley et al., 2012). In contrast, in the field crickets Gryllus pennsylvanicus and G.
firmus environmental heterogeneity plays an important role in structuring a mosaic hybrid
zone, as these species exhibit differential fitness on sand and loam soil types (Harrison &
Rand, 1989; Ross & Harrison, 2002).
Little is known about whether the Teleogryllus species in this study are ecologically
divergent. During fieldwork, I encountered males of both species singing in very close
contact in sympatric populations, with no obvious environmental partitioning. Both species
are believed to be ecological generalists (Otte & Alexander, 1983), which may reduce
competition for resources. Further work is needed to determine if ecological factors
contribute to maintaining species isolation in sympatry. In particular, the species are known
to differ in their diapause behaviour: in the southern regions T. commodus is a univolitine
species, with the potential to egg-diapause as an adaptation to the colder winters, while T.
oceanicus has no egg diapause and breeds continuously (Fontana & Hogan, 1968; Hogan,
1971; Otte & Alexander, 1983). In sympatry both species may breed continuously, but
environmental effects could trigger diapause in T. commodus, leading to allochrony
(Alexander & Bigelow, 1960; Harrison, 1985). In these cases, this could reduce both
competition for resources and contribute to reproductive isolation.
209
Do sex chromosomes play a pronounced role in species isolation?
Theory predicts that sex chromosomes will be enriched for genes involved in both
pre- and postzygotic barriers, due to the asymmetry in inheritance and expression
(Charlesworth et al., 1987; Qvarnström & Bailey, 2009). Indeed, the prevalence of Haldane’s
rule is one of the main empirical findings that suggests sex chromosomes commonly play a
key role in postzygotic isolation and possibly speciation. The results from our study are
surprising in two ways:
1) they counter the commonly held view that X chromosomes play a pronounced role in
hybrid incompatibilities,
2) they suggest that deviations from Haldane’s rule might be more prevalent than previously
appreciated.
I discussed these two points in Chapter 4, so here I will focus on an important but
poorly understood aspect of sex chromosomes, dosage compensation. In addition, I will
examine some of the broader implications of the findings and some future avenues of
research.
Dosage compensation is an important consideration for understanding how X-linked
incompatibilities may manifest and lead to exceptions for Haldane’s rule (Fraïsse et al.,
2016). As X-linked genes occur in a lower dose compared to the autosomes in the
heterogametic sex, mechanisms are expected to develop to equalize X and autosome
expression levels (Mank et al., 2011). In XO taxa, as there is no sex-specific chromosome,
dosage effects during early development are likely to play an important role in sexual
development. Hybridization could disrupt dosage compensation mechanisms in hybrids (Orr,
1989; Jablonka & Lamb, 1991), potentially affecting both primary and secondary sexual trait
development. In this study, intrinsic hybrid female sterility may therefore reflect broader
210
disruption of sexual development in hybrid females. In this case, hybrid females may also be
predicted to be behaviourally sterile, thereby extending this rare exception to Haldane’s rule
to extrinsic postzygotic isolation (Davies et al., 1997). Contrary to this prediction, the results
in Chapter 3, suggested that hybrid female courtship behaviour is not disrupted by
hybridization.
Sex chromosome dosage compensation can be incredibly complex and diverse
(reviewed in Mank et al., 2011). The simplest method to balance expression levels of X and
autosomal loci is hyper-transcription of the X in the heterogametic sex (e.g. Drosophila,
Presgraves, 2008), however this could also result in over expression of X-linked loci when in
the homogametic sex (Mank et al., 2011). Alternatively, in some taxa (e.g. mammals) one of
the X chromosomes is inactivated during early development in the homogametic sex. In
eutherian mammals, one copy of the X is randomly inactivated in each cell, resulting in a
mosaic of maternal and paternal X-chromosome expression (Watson & Demuth, 2012). In
marsupials, it is always the paternal X copy that is inactivated. X-inactivation has the
interesting property of extending functional hemizygosity to females (which is normally
confined to males) thereby exposing females to both recessive and dominant
incompatibilities. Under the dominance theory, X-inactivation is not predicted to result in
Haldane’s rule for inviability (Fraïsse et al., 2016), as males and females are both
functionally hemizygous and incompatibility loci generally influence both sexes equally
(Coyne, 1985; Orr, 1993; Wu & Davis, 1993; Presgraves & Orr, 1998). However, loci which
confer sterility generally have sex-specific effects (Wu & Davis, 1993). Therefore in cases of
X-inactivation, hybrid females could suffer disproportionate sterility effects (Fraïsse et al.,
2016).
In our study system, if X chromosomes were selectively inactivated, this may explain
the surprising similarity in fecundity in the second-generation backcrosses, even though the
211
two cross types differ substantially in their X chromosome complement (Chapter 3, Fig. 1). I
predicted that backcross females that inherited two different species' X chromosomes would
suffer greater dysfunction compared to females with two of the same X chromosomes, due to
X-linked incompatibilities (X-X, X autosomal or X- cytoplasmic interactions), on a
controlled autosomal background. However, if the paternal X was inactivated in the BC1
females, they would share the same functional genotype, thereby one would not expect a
striking difference in female fertility between the groups of interest (Fig. 1).
Figure 6-1 Selective paternal X-inactivation?
Schematic of the cross design (adapted from Chapter 3). BC1 females backcrossed to males
from their maternal species. The paternal X chromosomes are highlighted in blue and a red
asterisk represents inactivation of these X chromosomes. X-inactivation renders the females
functionally hemizygous and, moreover, the females under comparison now share the same
functional genotype. The arrows indicate group comparisons. X in parentheses indicates an
inter-species recombinant X.
212
It is difficult to predict whether dosage compensation contributes to hybrid female
sterility in our study system, as dosage compensation is not well understood outside of
mammals. The few studies which have examined dosage compensation among Orthopterans
suggest it is highly variable (Rao & Padmaja, 1992; Rodrigo et al., 2010). In the true cricket
species Acheta domesticus there is no X-inactivation, instead the X is upregulated in males
(Rao & Ali., 1981), whereas X-inactivation has been found in the more distantly related mole
crickets (Gryllotalpa fossor) (Rao & Padmaja, 1992). Interestingly, Hoy and colleagues
speculated that there may be selective paternal X-inactivation in hybrid females of these
Teleogryllus species, to support their hypothesis of “genetic coupling” (Hoy et al., 1977;
Doherty & Hoy, 1985). However, the evidence for genetic coupling and even more so Xinactivation is rather weak (Butlin & Ritchie, 1989; Boake, 1991). How sexual development
is determined in taxa which lack dimorphic sex chromosomes (e.g. XO species) is largely
unknown, but dosage effects are likely to play an important role. Therefore, even if this
exception to Haldane’s rule proves to be a rare phenomenon, research on the genetic basis of
postzygotic incompatibilities in XO taxa is likely to provide important insights into the
mechanisms underlying sex determination and development.
The genetic basis to hybrid female sterility and whether this exception to Haldane’s
rule represents a broader pattern among XO systems warrants further research. Comparative
studies will be particularly informative in this respect. The most recent comparative review
by Schilthuizen and colleagues (2011), encompassing results from less well-studied taxa with
unconventional sex determination systems (such as Amphibia, Reptilia, Teleostei), revealed a
higher proportion of exceptions to Haldane’s rule for sterility compared to earlier reviews
(~20% vs. 2%). Previous reviews had been primarily based on species with conventional sex
determination systems (Drosophila, Lepidoptera, birds and mammals) (Turelli & Orr, 1995;
Laurie, 1997; Coyne & Orr., 2004). Almost all cases which contradict Haldane’s rule for
213
sterility occur in only one direction of the cross (exceptions: Limnoporus water striders (XO
system) and Xenopus frogs (female heterogamety), which exhibit sterility in both cross
directions, discussed in Chapter 4). Asymmetrical effects generally implicate cyto-nuclear
incompatibilities (Turelli & Moyle, 2007). Reciprocal exceptions to Haldane’s rule for
sterility, as is the case for our Teleogryllus species, remain exceedingly rare. Interestingly, in
the recent review, cases conforming to Haldane’s rule for inviability are more common than
for sterility (308 vs. 44 species pairs), while the exact opposite pattern was found in earlier
reviews (73 vs. 169). This discrepancy was primarily driven by the addition of results from
Aves (female heterogamety) and Amphibia (multiple different sex determination systems).
This highlights the importance of considering taxa with less conventional sex determination
systems, which may contradict previous generalizations and provide novel insights into
Haldane’s rule.
Identifying hybrid genetic incompatibilities
Reproductive isolation is a joint property of two divergent taxa. Intrinsic postzygotic
incompatibilities emerge due to complex genic and chromosomal interactions, which are
unique to hybrid backgrounds. Numerous approaches have been used to try and identify the
genetic mechanisms and the associated genes underlying hybrid incompatibilities (reviewed
in Maheshwari & Barbash, 2011). These methods include: genetic mapping in backcross or
F2 generations (requires a phenotype) (Tao et al., 2003; Moehring et al., 2006), segregation
distortion across backcross generations (no phenotype required) (Xu et al., 1997; Gadau et
al., 1999), controlled introgressions (requires a candidate region) (Tao et al., 2003; Masly &
Presgraves, 2007; Moyle & Nakazato, 2008), identifying mis-expressed genes in hybrids
(Michalak & Noor, 2003; Ranz et al., 2004; Ortíz-Barrientos et al., 2007; Malone et al., 2007)
214
and admixture mapping in hybrid zones (requires differential introgression) (reviewed in
Buerkle & Lexer, 2008; Payseur, 2010).
As the F1 hybrid females in our study system are almost completely sterile and
hybridization and introgression is rare or absent in the contact zone, this constrains the
approaches available for identifying the causal hybrid incompatibilities. Successive
backcrosses could be used to genetically map loci contributing to reduced female fertility
through QTL mapping, or examining segregation distortion of alleles across generations.
However, the genetic incompatibilities contributing to a reduction of female fertility in
backcross individuals may not be the same as those underlying F1 hybrid female sterility.
Gene expression studies may be particularly useful for identifying genes involved in the
disruption of F1 hybrid female fertility. Genes which are most highly mis-expressed in
hybrids compared to the parental species may be responsible for the hybrid incompatibilities,
although it is often difficult to determine the causal relationship (Michalak & Noor, 2003).
Examining expression differences between males and females and sex chromosomes and
autosomes, could also inform more broadly about sex determination and sexual development
in XO taxa.
Future research using hybrid crosses would also benefit from testing whether
Haldane’s rule extends to intraspecific incompatibilities. Conspecific crosses of allopatric and
sympatric populations could be used. This may be particularly interesting in T. commodus
populations as they have been suggested to be highly polymorphic for chromosomal
inversions (Fontana & Hogan, 1969; Hogan & Fontana, 1973), which may contribute to
population divergence (Noor et al., 2001; Feder & Nosil, 2009). Linking intraspecific
processes of divergence to the emergence of interspecific intrinsic incompatibilities remains
an important challenge in speciation research. In addition, reproductive character
displacement could lead to reproductive isolation within species. Premating divergence in
215
sympatry could constrain migration and gene flow between allopatric populations, thereby
mitigating the problem of gene “swamping” disrupting the genetic associations between traits
under selection and reproductive isolation (i.e. the ‘‘cascade reinforcement’’ hypothesis;
Ortiz-Barrientos et al., 2009; Nosil, 2012 p149). Conspecific mating trials between allopatric
and sympatric populations could be informative in this respect.
Testing the unusual patterns of X chromosome differentiation
The results from the population genomic analysis revealed an unexpected pattern of
reduced genetic differentiation for X-linked loci compared to autosomal loci within species.
As discussed in Chapter 5, a combination of selective sweeps and/or sex-biased processes are
likely to contribute to this pattern, such as female-biased dispersal or increased mutation rates
in males. The presence of sex-biased dispersal could be tested by comparing markers with
different modes of inheritance, in particular uni-parentally inherited markers (e.g. mtDNA vs.
autosomal loci) (Prugnolle & de Meeus, 2002; Bailey et al., 2007). Little is known about sexbiased dispersal in insects and its ecological and genetic causes and consequences.
Most surprisingly, there was greater divergence for X-linked loci between species in
sympatry (Chapter 5). This is interesting as it suggests X-linked loci may have played an
important role in species isolation at some point in the past. Future work is needed to test the
significance of this result, as sample sizes in sympatry were small. The availability of a
linkage map would greatly enhance the interpretability of our results and facilitate the
identification of additional X-linked loci. The genetic basis to population divergence could be
further investigated using outlier analyses (e.g. FST), which, combined with a linkage map,
will allow for a more comprehensive test of the relative role of X vs. autosomal loci in
divergence. Estimates of linkage disequilibrium (LD), i.e. the non-random association of
216
alleles, could also contribute important information about population structure, selection and
chromosomal inversions (Rieseberg, 2001).
The genetic architecture of traits involved in prezygotic isolation is likely to play a
critical role in determining the outcome of species interactions upon secondary contact
(discussed in Chapter 1). Strong LD between traits under divergent selection and those
involved in reproductive isolation is important for maintaining species boundaries
(Felsenstein, 1981; Ortiz-Barrientos et al., 2009). In this study, the pattern of inheritance for
song components (which may contribute to premating isolation) suggested some sex-linkage.
However, further work is needed to examine female preferences and whether both signal and
preference traits are genetically correlated, due to pleiotropy or LD. Quantitative trait loci
(QTL) mapping provides an important method for determining the genetic architecture
(number of genes, effect sizes and chromosomal position) of traits contributing to species
differences (Falconer & Mackay, 1996). The direction of allelic effects can be used to infer
whether selection or drift played a primary role in trait divergence (Shaw & Parsons, 2002;
Gleason & Ritchie, 2004). However, as hybrid females are sterile this restricts QTL
approaches to using backcross generations, which have lower power than F2 intercrosses.
Alternatively, intraspecific crosses could be used. There is considerable geographic variation
among populations in song and CHCs (Chapter 2), therefore the most phenotypically
divergent populations could be used for QTL mapping. However, the loci identified may not
be the same as those that contribute to species differences (Gleason & Ritchie, 2004). Once
candidate QTL for traits involved in prezygotic barriers (also postzygotic barriers) are
identified, they could be integrated with population genomic data to examine the geographic
distribution of variants and their role in population divergence (Rogers & Bernatchez, 2005;
Stinchcombe & Hoekstra, 2008; Via et al., 2012).
217
Conclusions
The complexity of factors involved in the origin and maintenance of species
boundaries makes speciation a challenging but profoundly interesting area of research. In this
thesis, I aimed to bridge a gap in our knowledge about the interactions between two species,
which are both model systems for sexual selection studies. The species appear to be largely
reproductively isolated in the wild, suggesting they are near the end of the speciation
continuum (Mallet et al., 2007; Nosil, 2012). In this respect, more work needs to be done to
understand the factors that facilitate species coexistence, such as ecological differences. Our
results highlight some testable predictions (e.g. conspecific sperm precedent, female-biased
dispersal) and interesting avenues of future research (e.g. gene expression studies in hybrids,
unusual patterns of intra- vs. interspecific divergence at X-linked loci), as discussed above.
How often do partially interfertile species, which overlap both spatially and temporally, with
no obvious environmental segregation, coexist and remain genetically isolated? Most
speciation research tends to focus on earlier stages of speciation and on species pairs that are
known to hybridize (Coyne & Orr, 2004; Nosil, 2012; Seehausen et al., 2014). This is
important, in order to determine which reproductive barriers arise first and contribute most to
species divergence (Coyne & Orr, 2004). However, given that allopatric divergence is widely
accepted as the primary mode of speciation, a deeper understanding of the evolutionary
outcomes of secondary contact among taxa that are interfertile but do not hybridize is
important for understanding species diversity. This is especially important as climate change
is likely to lead to dramatic shifts in species distributions, bringing many sibling species into
secondary contact (Walther et al., 2002; Chen, 2012). Understanding the role that pre-existing
genetic and ecological divergence plays in determining whether secondary contact leads to
species coexistence or extinction will require a deeper integration of ecological and genetic
approaches (Ortiz-Barrientos et al., 2009; Butlin et al., 2012).
218
(Via, 2001)
(Feder, Egan, & Nosil, 2012; Gagnaire, Pavey, Normandeau, & Bernatchez, 2013; R.
G. Harrison & Larson, 2014; Mallet, Beltrán, Neukirchen, & Linares, 2007; Turelli, Barton,
& Coyne, 2001)
(Basolo, 1990; Doorn, Sander, Edelaar, & Weissing, 2009; Fisher, 1930; Kokko,
Jennions, & Brooks, 2013; Kraaijeveld, Kraaijeveld-Smit, & Maan, 2011; Maan &
Seehausen, 2011; Mead & Arnold, 2004; M Nei, Maruyama, & Wu, 1983; Payne &
Krakauer, 1997; Michael J Ryan & Cummings, 2013; Tyler, Frances, Xavier A. Harrison,
Bretman, Veen, & Rolando Rodríguez‐Muñoz, Tregenza, 2013; West-Eberhard, 1983)
(E. Anderson & Stebbins, 1954; Andersson & Simmons, 2006; Andrews, Good,
Miller, Luikart, & Hohenlohe, 2016; Andrews & Luikart, 2014; N. a. Baird et al., 2008;
Brown & Wilson, 1956; R. K. Butlin, 1998; Davey & Blaxter, 2010; Gompert & Buerkle,
2016; P. R. Grant, Grant, Keller, Markert, & Petren, 2003; Richard G. Harrison & Larson,
2016; Godfrey M. Hewitt, 1988; Kelly & Noor, 1996; Mark Kirkpatrick & Ravigné, 2002; A.
R. Lemmon & Kirkpatrick, 2006; Prowell, 1998; Pryke, 2010; Pryke & Griffith, 2009; Rice,
1984; Servedio & Saetre, 2003; Stebbins, 1959)
(Alexander, 1961, 1962; Bierne, Welch, Loire, Bonhomme, & David, 2011; Bousquet
et al., 2012; Broughton & Harrison, 2003; Carling & Brumfield, 2008; Fitzpatrick, 2013;
Fukamachi et al., 2009; Gompert & Buerkle, 2011; Richard G. Harrison & Rand, 1989;
Ronald R Hoy, Pollack, & Moiseff, 1982; Huber, Moore, & Loher., 1989; Kronforst et al.,
2013; Limousin et al., 2012; L. S. Maroja, Larson, Bogdanowicz, & Harrison, 2015; Luana S
Maroja, Andrés, & Harrison, 2009; Mullen, Mendelson, Schal, & Shaw, 2007; Pollack &
Hoy, 1979; Shaw & Lesnick, 2009; Teeter et al., 2010; Veen, Faulks, Tyler, Lloyd, &
Tregenza, 2012; Wiley, Ellison, & Shaw, 2012; M Zuk, Rotenberry, & Simmons, 2001)
(Marlene Zuk, Rotenberry, & Tinghitella, 2006)
(Arnegard & Kondrashov, 2004; M W Blows & Allan, 1998; Bussière, Hunt,
Jennions, & Brooks, 2006; Casaday & Hoy, 1977; Denno, McClure, & Ott, 1995; Fraïsse,
Gunnarsson, Roze, Bierne, & Welch, 2016; Gavrilets, 2000; Gwynne, 1991; Hardin, 1960;
Huber et al., 1989; Lucotte, Laurent, Heyer, Ségurel, & Toupance, 2016; Mallet, 2008; Mank,
Hosken, & Wedell, 2011; Luana S Maroja et al., 2014; Pascoal et al., 2014; D. W. Pfennig,
Rice, & Martin, 2006; Karin S. Pfennig & Pfennig, 2009; Shaw & Danley, 2003; Sousa &
Hey, 2013; Stein & Uy, 2006; E. Svensson, Nordén, Waller, & Runemark, 2016; Tinghitella
& Zuk, 2009; Veen, Faulks, Rodríguez-Muñoz, & Tregenza, 2011)
(Buerkle & Lexer, 2008; Chen, 2012; Gadau, Page, & Werren, 1999; Moehring,
Llopart, Elwyn, Coyne, & Mackay, 2006; Moyle & Nakazato, 2008; Ortíz-Barrientos,
Counterman, & Noor, 2007; Prugnolle & de Meeus, 2002; Rao & Ali., 1981; Rao &
Padmaja, 1992; Rodrigo, Marti, Bidau, Castillo, & Bidau, 2010; Stinchcombe & Hoekstra,
2008; Yun Tao, Chen, Hartl, & Laurie, 2003; Walther et al., 2002; Xu, Zhu, Xiao, Huang, &
McCouch, 1997) (Hopkins, Guerrero, Rausher, & Kirkpatrick, 2014)(Cruickshank & Hahn,
2014; Roux et al., 2016) (Mallet, 2007; Wood et al., 2009)
(Zirkle, 1934) (Hopkins et al., 2014; Marshall et al., 2002)
219
References
Abbott, R., Albach, D., Ansell, S., Arntzen, J. W., Baird, S. J. E., Bierne, N., … Zinner, D.
(2013). Hybridization and speciation. Journal of Evolutionary Biology, 26(2), 229–46.
http://doi.org/10.1111/j.1420-9101.2012.02599.x
Abe, T. A., Spence, J. R., & Sperling, F. A. H. (2005). Mitochondrial introgression is
restricted relative to nuclear markers in a water strider ( Hemiptera : Gerridae ) hybrid
zone. Canadian Journal of Zoology, 83, 432–444. http://doi.org/10.1139/Z05-030
Albert, A. Y. K., & Schluter, D. (2004). Reproductive character displacement of male
stickleback mate preference: reinforcement or direct selection? Evolution; International
Journal of Organic Evolution, 58(5), 1099–1107. http://doi.org/10.1111/j.00143820.2004.tb00443.x
Alexander, R. D. (1961). Aggressiveness , Territoriality, and Sexual Behavior in Field
Crickets (Orthoptera : Gryllidae). Behaviour, 17(2), 130–223.
Alexander, R. D. (1962). Evolutionary change in cricket acoustical communication.
Evolution, 16(4), 443–467.
Alexander, R. D., & Bigelow, R. S. . (1960). Allochronic Speciation in Field Crickets , and a
New Species , Acheta veletis. Evolution, 14(3), 334–346.
Alexander, R. D., & Otte, D. (1967). The Evolution ofGenitalia and Mating Behaviour in
Crickets (Gryllidae) and other Orthoptera. Museum of Zoology, University of
Michigan.
Anderson, E. (1948). Hybridization of the Habitat. Evolution, 2(1), 1–9.
http://doi.org/10.1038/157619d0
Anderson, E., & Stebbins, G. L. (1954). Hybridization as an Evolutionary Stimulus.
Evolution, 8(4), 378–388.
Andersson, M., & Simmons, L. W. (2006). Sexual selection and mate choice. Trends in
Ecology & Evolution, 21(6), 296–302. http://doi.org/10.1016/j.tree.2006.03.015
Andrews, K. R., Good, J. M., Miller, M. R., Luikart, G., & Hohenlohe, P. A. (2016).
Harnessing the power of RADseq for ecological and evolutionary genomics. Nat Rev
Genet, advance on(2), 81–92. http://doi.org/10.1038/nrg.2015.28
Andrews, K. R., & Luikart, G. (2014). Recent novel approaches for population genomics data
analysis. Molecular Ecology, 23, 1661–1667. http://doi.org/10.1111/mec.12686
Arnegard, M. E., & Kondrashov, A. S. (2004). Sympatric speciation by sexual selection alone
is unlikely. Evolution; International Journal of Organic Evolution, 58(2), 222–37.
Arnold, M. L. (1997). Natural Hybridization and Evolution. Oxford, UK: Oxford Univ.
Press.
Arnold, M. L., Ballerini, E. S., & Brothers, a N. (2012). Hybrid fitness, adaptation and
evolutionary diversification: lessons learned from Louisiana Irises. Heredity, 108(3),
159–66. http://doi.org/10.1038/hdy.2011.65
Arnold, M. L., Bulger, M. R., Burke, J. M., Hempel, A. L., & Williams, J. H. (1999). Natural
hybridization: How low can you go and still be important? Ecology, 80(2), 371–381.
220
http://doi.org/10.1890/0012-9658(1999)080[0371:NHHLCY]2.0.CO;2
Arnold, S. J., Verrell, P. A., & Tilley, S. G. (1996). The Evolution of Asymmetry in Sexual
Isolation : A Model and a Test Case. Evolution, 50(3), 1024–1033.
Arthur, N. J., & Dyer, K. A. (2015). Asymmetrical sexual isolation but no postmating
isolation between the closely related species Drosophila suboccidentalis and Drosophila
occidentalis. BMC Evolutionary Biology, 15(1), 38. http://doi.org/10.1186/s12862-0150328-y
Bailey, N. W. (2011). Mate choice plasticity in the field cricket Teleogryllus oceanicus:
effects of social experience in multiple modalities. Behavioral Ecology and
Sociobiology, 65(12), 2269–2278.
Bailey, N. W., Gwynne, D. T., & Ritchie, M. G. (2007). Dispersal differences predict
population genetic structure in Mormon crickets. Molecular Ecology, 16(10), 2079–
2089. http://doi.org/10.1111/j.1365-294X.2007.03265.x
Bailey, N. W., & Macleod, E. (2013). Socially flexible female choice and premating isolation
in field crickets (Teleogryllus spp.). Journal of Evolutionary Biology, 27, 170–180.
http://doi.org/10.1111/jeb.12285
Bailey, N. W., & Moore, A. J. (2012). Runaway sexual selection without genetic correlations:
social environments and flexible mate choice initiate and enhance the Fisher process.
Evolution, 66(9), 2674–2684.
Bailey, N. W., & Zuk, M. (2008). Acoustic experience shapes female mate choice in field
crickets. Proceedings. Biological Sciences / The Royal Society, 275(1651), 2645–50.
http://doi.org/10.1098/rspb.2008.0859
Baird, N. a., Etter, P. D., Atwood, T. S., Currey, M. C., Shiver, A. L., Lewis, Z. a., …
Johnson, E. a. (2008). Rapid SNP discovery and genetic mapping using sequenced RAD
markers. PLoS ONE, 3(10), 1–7. http://doi.org/10.1371/journal.pone.0003376
Baird, S. E. (2002). Haldane’s rule by sexual transformation in caenorhabditis. Genetics,
161(3), 1349–1353.
Baird, S. E., & Yen, W. C. (2000). Reproductive isolation in Caenorhabditis: Terminal
phenotypes of hybrid embryos. Evolution and Development, 2(1), 9–15.
http://doi.org/10.1046/j.1525-142X.2000.00031.x
Balakrishnan, R., & Pollack, G. (1997). The role of antennal sensory cues in female
responses to courting males in the cricket Teleogryllus oceanicus. The Journal of
Experimental Biology, 200(Pt 3), 511–22.
Baldassarre, D. T., White, T. A., Karubian, J., & Webster, M. S. (2014). Genomic and
morphological analysis of a semipermeable avian hybrid zone suggests asymmetrical
introgression of a sexual signal. Evolution; International Journal of Organic Evolution,
68(9), 2644–57. http://doi.org/10.1111/evo.12457
Barton, N. H., & Charlesworth, B. (1984). Genetic Revolutions , Founder Effects , and
Speciation. Annual Review of Ecology and Systematics, 15(1984), 133–164.
Barton, N. H., & Hewitt, G. M. (1985). Analysis of Hybrid Zones. Annual Review of Ecology
and Systematics, 16(1), 113–148. http://doi.org/10.1146/annurev.es.16.110185.000553
221
Baselga, A. (2010). Partitioning the turnover and nestedness components of beta diversity.
Global Ecology and Biogeography, 19(1), 134–143. http://doi.org/10.1111/j.14668238.2009.00490.x
Basolo, a L. (1990). Female preference predates the evolution of the sword in swordtail fish.
Science (New York, N.Y.). http://doi.org/10.1126/science.250.4982.808
Bentley, D. R., & Hoy, R. R. (1972). Genetic Control of the Neuronal Network Generating
Cricket (Teleogryllus Gryllus) Song Patterns. Animal Behaviour, (1968), 478–492.
Bentsen, C. L., Hunt, J., Jennions, M. D., & Brooks, R. (2006). Complex multivariate sexual
selection on male acoustic signaling in a wild population of Teleogryllus commodus.
The American Naturalist, 167(4), E102-16. http://doi.org/10.1086/501376
Bierne, N., Welch, J., Loire, E., Bonhomme, F., & David, P. (2011). The coupling
hypothesis: why genome scans may fail to map local adaptation genes. Molecular
Ecology, 20(10), 2044–72. http://doi.org/10.1111/j.1365-294X.2011.05080.x
Blows, M. W. (2002). Interaction between natural and sexual selection during the evolution
of mate recognition. The Royal Society, 269(1496), 1113–8.
http://doi.org/10.1098/rspb.2002.2002
Blows, M. W., & Allan, R. a. (1998). Levels of mate recognition within and between two
Drosophila species and their hybrids. The American Naturalist, 152(6), 826–37.
http://doi.org/10.1086/286211
Boake, C. R. (1991). Coevolution of senders and receivers of sexual signals: Genetic
coupling and genetic correlations. Trends in Ecology & Evolution, 6(7), 225–7.
http://doi.org/10.1016/0169-5347(91)90027-U
Bonduriansky, R. (2001). The evolution of male mate choice in insects: a synthesis of ideas
and evidence. Biological Reviews, 76(3), 305–339.
http://doi.org/10.1017/s1464793101005693
Bordenstein, S. R., Drapeau, M. D., & Werren, J. H. (2000). Intraspecific Variation in Sexual
Isolation in the Jewel Wasp Nasonia, 54(2), 567–573.
Borge, T., Lindroos, K., Nadvornik, P., SYVÄNEN, A. C., & Satre, G. P. (2005). Amount of
introgression in flycatcher hybrid zones reflects regional differences in pre and postzygotic barriers to gene exchange. Journal of Evolutionary Biology, 18(6), 1416–1424.
http://doi.org/10.1111/j.1420-9101.2005.00964.x
Boughman, J. W. (2001). Divergent sexual selection enhances reproductive isolation in
sticklebacks. Nature, 411(6840), 944–8. http://doi.org/10.1038/35082064
Bousquet, F., Nojima, T., Houot, B., Chauvel, I., Chaudy, S., Dupas, S., … Ferveur, J.-F.
(2012). Expression of a desaturase gene, desat1, in neural and nonneural tissues
separately affects perception and emission of sex pheromones in Drosophila.
Proceedings of the National Academy of Sciences of the United States of America,
109(1), 249–54. http://doi.org/10.1073/pnas.1109166108
Bridle, J. R., & Butlin, R. K. (2002). Mating signal variation and bimodality in a mosaic
hybrid zone between Chorthippus grasshopper species. Evolution, 56(6), 1184–1198.
http://doi.org/10.1111/j.0014-3820.2002.tb01431.x
Bridle, J. R., Saldamando, C. I., Koning, W., & Butlin, R. K. (2006). Assortative preferences
222
and discrimination by females against hybrid male song in the grasshoppers Chorthippus
brunneus and Chorthippus jacobsi (Orthoptera: Acrididae). Journal of Evolutionary
Biology, 19(4), 1248–56. http://doi.org/10.1111/j.1420-9101.2006.01080.x
Brooks, R., Hunt, J., Blows, M. W., Smith, M. J., Bussière, L. F., & Jennions, M. D. (2005).
Experimental evidence for multivariate stabilizing sexual selection. Evolution;
International Journal of Organic Evolution, 59(4), 871–80.
Brothers, A. N., & Delph, L. F. (2010). Haldane ’ s rule is extended to plants with sex
chromosomes. Evolution, 3643–3648. http://doi.org/10.1111/j.1558-5646.2010.01095.x
Broughton, R. E., & Harrison, R. G. (2003). Nuclear gene genealogies reveal historical,
demographic and selective factors associated with speciation in field crickets. Genetics,
163(4), 1389–401.
Brown, W. L., & Wilson, E. (1956). Character Displacement. Systematic Zoology, 5(2), 49–
64.
Broyles, S. B. (2002). Hybrid bridges to gene flow: a case study in milkweeds (Asclepias).
Evolution, 56(10), 1943–1953. http://doi.org/10.1111/j.0014-3820.2002.tb00120.x
Buerkle, C. A., & Lexer, C. (2008). Admixture as the basis for genetic mapping. Trends in
Ecology & Evolution, 23(12), 686–94. http://doi.org/10.1016/j.tree.2008.07.008
Bundus, J. D., Alaei, R., & Cutter, A. D. (2015). Gametic selection, developmental
trajectories, and extrinsic heterogeneity in Haldane’s rule. Evolution, 69(8), 2005–2017.
http://doi.org/10.1111/evo.12708
Burke, J. M., & Arnold, M. L. (2001). Genetics and the Fitness of Hybrids, 31–52.
Bussière, L. F., Hunt, J., Jennions, M. D., & Brooks, R. (2006). Sexual conflict and cryptic
female choice in the black field cricket, Teleogryllus commodus. Evolution, 60(4), 792–
800.
Butlin, R. K. & Ritchie, M. G. (1989). Genetic coupling in mate recognition systems: what is
the evidence? Biological Journal of the Linnean Society, 37, 237–246.
Butlin, R., Debelle, A., Kerth, C., Snook, R. R., Beukeboom, L. W., Castillo Cajas, R. F., …
Schilthuizen, M. (2012). What do we need to know about speciation? Trends in Ecology
and Evolution, 27(1), 27–39. http://doi.org/10.1016/j.tree.2011.09.002
Butlin, R. K. (1987). Speciation and Reinforcement. The American Naturalist, 130(3), 461–
464.
Butlin, R. K. (1989). Reinforcement of premating isolation. In Speciaiton and it’s
consequences (pp. 158–179).
Butlin, R. K. (1998). What do hybrid zones in general and the Chorthippus parallelus zone in
particular, tell us about speciation?. In D. J. Howard & S. H. Berlocher (Eds.), Endless
Forms: Species and Speciation. (pp. 367–389). Oxford University Press.
Cain, M. L., Andreasen, V., & Howard, D. J. (1999). Reinforcing Selection is Effective under
a Relatively Broad Set of Conditions in a Mosaic Hybrid Zone. Evolution, 53(5), 1343–
1353.
Cairns, K. M., Wolff, J. N., Brooks, J. N., & Ballard, J. W. O. (2010). Evidence of recent
population expansion in the field cricket Teleogryllus commodus. Australian Journal of
223
Zoology, 58(1), 33–38.
Candolin, U. (2008). The use of multiple cues in mate choice. Biol. Rev., 78, 575–595.
http://doi.org/10.1017/S1464793103006158
Carling, M. D., & Brumfield, R. T. (2008). Haldane’s rule in an avian system: using cline
theory and divergence population genetics to test for differential introgression of
mitochondrial, autosomal, and sex‐linked loci across the passerina bunting hybrid zone.
Evolution, 62(10), 2600–2615. http://doi.org/10.1111/j.1558-5646.2008.00477.x
Casaday, G. B., & Hoy, R. R. (1977). Auditory Interneurons in the Cricket TeleogryUus
oceanicus : Physiological and Anatomical Properties. Journal of Comparative
Physiology A: Neuroethology, Sensory, Neural, and Behavioral Physiology, 121(1), 1–
13.
Catchen, J., Hohenlohe, P. a, Bassham, S., Amores, A., & Cresko, W. a. (2013). Stacks: an
analysis tool set for population genomics. Molecular Ecology, 22(11), 3124–40.
http://doi.org/10.1111/mec.12354
Chapple, D. G., Hoskin, C. J., Chapple, S. N. J., & Thompson, M. B. (2011).
Phylogeographic divergence in the widespread delicate skink (Lampropholis delicata)
corresponds to dry habitat barriers in eastern Australia. BMC Evolutionary Biology,
11(1), 191. http://doi.org/10.1186/1471-2148-11-191
Charlesworth, B. (1998). Measures of divergence between populations and the effect of
forces that reduce variability. Molecular Biology and Evolution, 15(5), 538–543.
http://doi.org/10.1093/oxfordjournals.molbev.a025953
Charlesworth, B., Coyne, J. A., & Barton, N. H. (1987). The Relative Rates of Evolution of
Sex Chromosomes and Autosomes. American Naturalist, 130(1), 113–146.
Charlesworth, B., Nordborg, M., & Charlesworth, D. (1997). The effects of local selection,
balanced polymorphism and background selection on equilibrium patterns of genetic
diversity in subdivided populations. Genet. Res., Camb, 70, 155–174.
http://doi.org/10.1017/S0016672397002954
Chatfield, M. W. H., Kozak, K. H., Fitzpatrick, B. M., & Tucker, P. K. (2010). Patterns of
differential introgression in a salamander hybrid zone: inferences from genetic data and
ecological niche modelling. Molecular Ecology, 19(19), 4265–82.
http://doi.org/10.1111/j.1365-294X.2010.04796.x
Chen, I. (2012). Rapid Range Shifts of Species. Science, 1024(2011), 17–20.
http://doi.org/10.1126/science.1206432
Clark, M. E., O’Hara, F. P., Chawla, a, & Werren, J. H. (2010). Behavioral and
spermatogenic hybrid male breakdown in Nasonia. Heredity, 104(3), 289–301.
http://doi.org/10.1038/hdy.2009.152
Coyne, J. A. (1985). The genetic basis of Haldane’s rule. Nature, 314(25).
Coyne, J. A., Charlesworth, B., & Orr, H. A. (1991). Haldane’s Rule Revisited. Evolution,
45(7), 1710–1714.
Coyne, J. A., Crittenden, A. P., & Mah, K. (1994). Genetics of a pheromonal difference
contributing to reproductive isolation in Drosophila. Science (New York, N.Y.),
265(5177), 1461–4.
224
Coyne, J. A., & Elwyn, S. (2006). Does the desaturase-2 locus in Drosophila melanogaster
cause adaptation and sexual isolation? Evolution, 60(2), 279–291.
http://doi.org/10.1111/j.0014-3820.2006.tb01106.x
Coyne, J. A., & Orr., H. A. (2004). Speciation. Sinauer Associates, Sunderland, MA.
Coyne, J. A., & Orr, H. A. (1989a). Patterns of Speciation in Drosophila. Evolution, 43(2),
362–381.
Coyne, J. A., & Orr, H. A. (1989b). Two rules of speciation. In Speciation and its
Consequences (pp. 180–207).
Coyne, J. A., & Orr, H. A. (1997). “Patterns of Speciation in Drosophila ” Revisited.
Evolution, 51(1), 295–303.
Coyne, J. A., Wicker-Thomas, C., & Jallon, J. M. (1999). A gene responsible for a cuticular
hydrocarbon polymorphism in Drosophila melanogaster. Genetical Research, 73, 189–
203. http://doi.org/10.1017/S0016672398003723
Cruickshank, T. E., & Hahn, M. W. (2014). Reanalysis suggests that genomic islands of
speciation are due to reduced diversity, not reduced gene flow. Molecular Ecology,
23(13), 3133–3157. http://doi.org/10.1111/mec.12796
Currat, M., Ruedi, M., Petit, R. J., & Excoffier, L. (2008). The hidden side of invasions:
Massive introgression by local genes. Evolution, 62(8), 1908–1920.
http://doi.org/10.1111/j.1558-5646.2008.00413.x
Cutter, A. D. (2012). The polymorphic prelude to Bateson-Dobzhansky-Muller
incompatibilities. Trends in Ecology and Evolution, 27(4), 209–218.
http://doi.org/10.1016/j.tree.2011.11.004
Danecek, P., Auton, A., Abecasis, G., Albers, C. a., Banks, E., DePristo, M. a., … Durbin, R.
(2011). The variant call format and VCFtools. Bioinformatics, 27(15), 2156–2158.
http://doi.org/10.1093/bioinformatics/btr330
Darling, J. A., Tsai, Y. E., April, M. H., Roman, J., & Drive, T. W. A. (2014). Are genes
faster than crabs ? Mitochondrial introgression exceeds larval dispersal during
population expansion of the invasive crab Carcinus maenas Subject Areas : Author for
correspondence : Proceedings of The Royal Society B.
Darwin, C. (1871). Sexual selection and the descent of man. Murray, London.
Davey, J., & Blaxter, M. (2010). RADSeq: next-generation population genetics. Briefings in
Functional Genomics, 9(5–6), 416–23.
Davies, N., Aiello, a., Mallet, J., Pomiankowski, a., & Silberglied, R. E. (1997). Speciation
in two neotropical butterflies: Extending Haldane’s rule. Proceedings of the Royal
Society B: Biological Sciences, 264(1383), 845–851.
http://doi.org/10.1098/rspb.1997.0118
Davies, N., & Pomiankowski, a. (1995). Haldane’s rule: Old theories are the best. Trends in
Ecology and Evolution, 10(9), 350–351. http://doi.org/10.1016/S0169-5347(00)89131-4
de Villemereuil, P., & Gaggiotti, O. E. (2015). A new FST-based method to uncover local
adaptation using environmental variables. Methods in Ecology and Evolution, 1–24.
http://doi.org/10.1111/2041-210X.12418
225
Debelle, A., Ritchie, M. G., & Snook, R. R. (2016). Sexual selection and assortative mating:
an experimental test. Journal of Evolutionary Biology, 29, n/a-n/a.
http://doi.org/10.1111/jeb.12855
Denno, R. F., McClure, M. S., & Ott, J. R. (1995). Interspecific interactions in phytophagous
insects: competition reexamined and resurrected. Annual Review of Entomology, 40(1),
297–331.
Dey, A., Jin, Q., Chen, Y. C., & Cutter, A. D. (2014). Gonad morphogenesis defects drive
hybrid male sterility in asymmetric hybrid breakdown of Caenorhabditis nematodes.
Evolution and Development, 16(6), 362–372. http://doi.org/10.1111/ede.12097
Dobzhansky, T. (1935). "A critique of the species concept in biology. Philosophy of Science,
344–355.
Dobzhansky, T. (1937). Genetics and the Origin of Species. Columbia University Press.
Dobzhansky, T. (1940). Speciation as a Stage in Evolutionary Divergence. The American
Naturalist, 74(753), 312–321.
Doherty, J., & Hoy, R. (1985). The auditory behavior of crickets: some views of genetic
coupling, song recognition, and predator detection. Quarterly Review of Biology, 60(4),
457–472.
Doorn, V., Sander, G., Edelaar, P., & Weissing, F. J. (2009). On the origin of species by
natural and sexual selection. Science, 326(5960), 1704–1707.
Dougherty, L. R., & Shuker, D. M. (2015). The effect of experimental design on the
measurement of mate choice: A meta-analysis. Behavioral Ecology, 26(2), 311–319.
http://doi.org/10.1093/beheco/aru125
Dowling, T. E., & Secor, C. L. (1997). The Role of Hybridization and Introgression in the
Diversification of Animals. Annual Review of Ecology and Systematics, (33), 593–619.
Dray, S., & Dufour, A. B. (2007). The ade4 Package: Implementing the Duality Diagram for
Ecologists. Journal of Statistical Software, 22(4), 1–20. http://doi.org/10.1.1.177.8850
Drayton, J. M., Hunt, J., Brooks, R., & Jennions, M. D. (2007). Sounds different: inbreeding
depression in sexually selected traits in the cricket Teleogryllus commodus. Journal of
Evolutionary Biology, 20(3), 1138–47. http://doi.org/10.1111/j.1420-9101.2006.01286.x
Drayton, J. M., Milner, R. N. C., Hunt, J., & Jennions, M. D. (2010). Inbreeding and
advertisement calling in the cricket Teleogryllus commodus: laboratory and field
experiments. Evolution; International Journal of Organic Evolution, 64(10), 3069–83.
http://doi.org/10.1111/j.1558-5646.2010.01053.x
Eaton, D. A. R., Hipp, A. L., Gon Alez-Rodríguez, A., & Cavender-Bares, J. (2015).
Historical introgression among the American live oaks and the comparative nature of
tests for introgression. Evolution, 2587–2601. http://doi.org/10.1111/evo.12758
Edmands, S. (2002). Does parental divergence predict reproductive compatibility? Trends in
Ecology and Evolution, 17(11), 520–527. http://doi.org/10.1016/S0169-5347(02)025855
Edward, D. A., & Chapman, T. (2011). The evolution and significance of male mate choice.
Trends in Ecology and Evolution, 26(12), 647–654.
226
http://doi.org/10.1016/j.tree.2011.07.012
Ellegren, H. (2009). The different levels of genetic diversity in sex chromosomes and
autosomes. Trends in Genetics, 25(6), 278–284. http://doi.org/10.1016/j.tig.2009.04.005
Ellegren, H. (2014). Genome sequencing and population genomics in non-model organisms.
Trends in Ecology and Evolution, 29(1), 51–63.
http://doi.org/10.1016/j.tree.2013.09.008
Ellison, C. K., Wiley, C., & Shaw, K. L. (2011). The genetics of speciation: genes of small
effect underlie sexual isolation in the Hawaiian cricket Laupala. Journal of Evolutionary
Biology, 24(5), 1110–9.
Endler, J. A. (1992). Signals , Signal Conditions , and the Direction of Evolution. The
American Naturalist, 139(May).
Excoffier, L., Foll, M., & Petit, R. J. (2009). Genetic Consequences of Range Expansions.
Annual Review of Ecology, Evolution, and Systematics, 40(1), 481–501.
http://doi.org/10.1146/annurev.ecolsys.39.110707.173414
Excoffier, L., & Lischer, H. E. L. (2010). Arlequin suite ver 3.5: A new series of programs to
perform population genetics analyses under Linux and Windows. Molecular Ecology
Resources, 10(3), 564–567. http://doi.org/10.1111/j.1755-0998.2010.02847.x
Falconer, D. S., & Mackay, T. F. C. (1996). Introduction to Quantitative Genetics. Pearson
Education Limited.
Feder, J. L., Egan, S. P., & Nosil, P. (2012). The genomics of speciation-with-gene-flow.
Trends in Genetics : TIG, 28(7), 342–50. http://doi.org/10.1016/j.tig.2012.03.009
Feder, J. L., & Nosil, P. (2009). Chromosomal inversions and species differences: when are
genes affecting adaptive divergence and reproductive isolation expected to reside within
inversions? Evolution; International Journal of Organic Evolution, 63(12), 3061–75.
http://doi.org/10.1111/j.1558-5646.2009.00786.x
Felsenstein, J. (1981). Skepticism Towards Santa Rosalia , or Why are There so Few Kinds
of Animals ? Evolution, 35(1), 124–138.
Ferveur, J. F. (2005). Cuticular hydrocarbons: Their evolution and roles in Drosophila
pheromonal communication. Behavior Genetics, 35(3), 279–295.
http://doi.org/10.1007/s10519-005-3220-5
Fields, P. D., Reisser, C., Dukic, M., Haag, C. R., & Ebert, D. (2015). Genes mirror
geography in Daphnia magna. Molecular Ecology, 24(17), 4521–4536.
http://doi.org/10.1111/mec.13324
Fisher, R. A. (1930). The genetical theory of natural selection: a complete variorum edition.
Oxford University Press.
Fitzpatrick, B. M. (2013). Alternative forms for genomic clines. Ecology and Evolution, 3(7),
1951–66. http://doi.org/10.1002/ece3.609
Fontana, P. G., & Hogan, T. W. (1969). Cytogenetic and Hybridization studies of geographic
populations of Teleogryllus commodus (Walker) and T. oceanicus (Le
Guillou)(Orthoptera: Gryllidae). Australian Journal of Zoology, 17(1), 13–35.
Ford, J. (1987). Hybrid Zones in Australian Birds. Emu, 87(December), 158–178.
227
Ford, M. J., & Aquadro, C. F. (1996). Selection on X-linked genes during speciation in the
Drosophila athabasca complex. Genetics, 144(2), 689–703.
Foster, S. A., & Endler, J. A. (1999). Geographic Variation in Behaviour.
Fraïsse, C., Gunnarsson, P. A., Roze, D., Bierne, N., & Welch, J. J. (2016). The genetics of
speciation : insights from Fisher’s Geometric Model. Evolution, (1937), 1450–1464.
http://doi.org/10.1111/evo.1296
Frank, S. A. (1991). Haldane’s Rule: A Defense of the Meiotic Drive Theory. Society for the
Study of Evolution, 45(7), 1714–1717.
Frentiu, F. D., & Chenoweth, S. F. (2010). Clines in cuticular hydrocarbons in two drosophila
species with independent population histories. Evolution, 64(6), 1784–1794.
http://doi.org/10.1111/j.1558-5646.2009.00936.x
Fukamachi, S., Kinoshita, M., Aizawa, K., Oda, S., Meyer, A., & Mitani, H. (2009). Dual
control by a single gene of secondary sexual characters and mating preferences in
medaka. BMC Biology, 7, 64. http://doi.org/10.1186/1741-7007-7-64
Funk, D. J., Nosil, P., & Etges, W. J. (2006). Ecological divergence exhibits consistently
positive associations with reproductive isolation across disparate taxa. Proceedings of
the National Academy of Sciences of the United States of America, 103(9), 3209–3213.
http://doi.org/10.1073/pnas.0508653103
Gadau, J., Page, R. E., & Werren, J. H. (1999). Mapping of hybrid incompatibility loci in
Nasonia. Genetics, 153(4), 1731–1741.
Gagnaire, P. A., Pavey, S. a., Normandeau, E., & Bernatchez, L. (2013). The genetic
architecture of reproductive isolation during speciation-with-gene-flow in lake whitefish
species pairs assessed by rad sequencing. Evolution, 67(9), 2483–2497.
http://doi.org/10.1111/evo.12075
Gavrilets, S. (2000). Rapid evolution of reproductive barriers driven by sexual conflict.
Nature, 403(6772), 886–889. http://doi.org/10.1038/35002564
Gerhardt, H. C. (2013). Geographic variation in acoustic communication: Reproductive
character displacement and speciation. In Evolutionary Ecology Research (Vol. 15, pp.
605–632).
Gerhardt, H. C., & Huber, F. (2002). Acoustic Communication in Insects and Anurans.
Gleason, J. M., & Ritchie, M. G. (1998). Evolution of Courtship Song and Reproductive
Isolation in the Drosophila willistoni Species Complex : Do Sexual Signals Diverge the
Most Quickly ? Society for the Study of Evolution, 52(5), 1493–1500.
Gompert, Z., & Alex Buerkle, C. (2010). Introgress: a Software Package for Mapping
Components of Isolation in Hybrids. Molecular Ecology Resources, 10(2), 378–84.
http://doi.org/10.1111/j.1755-0998.2009.02733.x
Gompert, Z., & Buerkle, C. A. (2009). A powerful regression-based method for admixture
mapping of isolation across the genome of hybrids. Molecular Ecology, 18, 1207–1224.
http://doi.org/10.1111/j.1365-294X.2009.04098.x
Gompert, Z., & Buerkle, C. A. (2011). A hierarchical bayesian model for next-generation
population genomics. Genetics, 187(3), 903–917.
228
http://doi.org/10.1534/genetics.110.124693
Gompert, Z., & Buerkle, C. A. (2016). What, if anything, are hybrids: Enduring truths and
challenges associated with population structure and gene flow. Evolutionary
Applications. http://doi.org/10.1111/eva.12380
Gompert, Z., Lucas, L. K., Buerkle, C. A., Forister, M. L., Fordyce, J. A., & Nice, C. C.
(2014). Admixture and the organization of genetic diversity in a butterfly species
complex revealed through common and rare genetic variants. Molecular Ecology,
23(18), 4555–4573. http://doi.org/10.1111/mec.12811
Gompert, Z., Parchman, T. L., & Buerkle, C. A. (2012). Genomics of isolation in hybrids.
Philosophical Transactions of the Royal Society of London. Series B, Biological
Sciences, 367(1587), 439–50. http://doi.org/10.1098/rstb.2011.0196
Gottipati, S., Arbiza, L., Siepel, A., Clark, A. G., & Keinan, A. (2011). Analyses of X-linked
and autosomal genetic variation in population-scale whole genome sequencing. Nature
Genetics, 43(8), 741–3. http://doi.org/10.1038/ng.877
Goudet, J. (2005). HIERFSTAT, a package for R to compute and test hierarchical F-statistics.
Molecular Ecology Notes, 2, 184–186. http://doi.org/10.1111/j.1471-8278
Goudet, J., Perrin, N., & Waser, P. (2002). Tests for sex-biased dispersal using bi-parentally
inherited genetic markers. Molecular Ecology, 11(6), 1103–1114.
http://doi.org/10.1046/j.1365-294X.2002.01496.x
Gow, J. L., Peichel, C. L., & Taylor, E. B. (2007). Ecological selection against hybrids in
natural populations of sympatric threespine sticklebacks. Journal of Evolutionary
Biology, 20(6), 2173–2180. http://doi.org/10.1111/j.1420-9101.2007.01427.x
Grant, B. R., & Grant, P. R. (1993). Evolution of Darwin’s Finches Caused by a Rare
Climatic Event. Proceedings of The Royal Society B, 251, 111–117.
Grant, P. R., Grant, B. R., Keller, L. F., Markert, J. A., & Petren, K. (2003). Inbreeding and
Interbreeding in Darwin ’ S Finches. Evolution, 57(12), 2911–2916.
Gray, D. a. (2005). Does courtship behavior contribute to species-level reproductive isolation
in field crickets? Behavioral Ecology, 16(1), 201–206.
http://doi.org/10.1093/beheco/arh144
Gray, D. a., & Cade, W. H. (2000). Sexual selection and speciation in field crickets.
Proceedings of the National Academy of Sciences of the United States of America,
97(26), 14449–54. http://doi.org/10.1073/pnas.97.26.14449
Gray, D. A., Gutierrez, N. J., Chen, T. L., Weissman, D. B., & Cole, J. A. (2016). Species
divergence in field crickets : genetics , song , ecomorphology , and pre- and postzygotic
isolation. Biological Journal of the Linnean Society, 117(192–205), 192–205.
Greenfield, M. D. (2002). Signalers and receivers mechanisms and evolution of arthropod
communication.
Greenfield, M. D., Alem, S., Limousin, D., & Bailey, N. W. (2014). The dilemma of
Fisherian sexual selection: Mate choice for indirect benefits despite rarity and overall
weakness of trait-preference genetic correlation. Evolution, 68(12), 3524–3536.
http://doi.org/10.1111/evo.12542
229
Greig, E. I., Baldassarre, D. T., & Webster, M. S. (2015). Differential rates of phenotypic
introgression are associated with male behavioral responses to multiple signals.
Evolution, n/a-n/a. http://doi.org/10.1111/evo.12756
Gwynne, D. T. (1981). Sexual difference theory: mormon crickets show role reversal in mate
choice. Science (New York, N.Y.). http://doi.org/10.1126/science.213.4509.779
Gwynne, D. T. (1985). Role-Reversal in Katydids : Habitat Influences Reproductive
Behaviour ( Orthoptera : Tettigoniidae , Metaballus sp .). Behavioral Ecology and
Sociobiology, 16(4), 355–361.
Gwynne, D. T. (1991). Sexual competition among females: What causes courtship-role
reversal? Trends in Ecology and Evolution, 6(4), 118–121. http://doi.org/10.1016/01695347(91)90089-G
Gwynne, D. T., & Simmons, L. W. (1990). Experimental reversal of courtship roles in an
insect. Nature, 346(6280), 172–174. http://doi.org/10.1016/0021-9797(80)90501-9
Haavie, J., Borge, T., Bures, S., Garamszegi, L. Z., Lampe, H. M., Moreno, J., … Sætre, G.
P. (2004). Flycatcher song in allopatry and sympatry - Convergence, divergence and
reinforcement. Journal of Evolutionary Biology, 17(2), 227–237.
http://doi.org/10.1111/j.1420-9101.2003.00682.x
Hack, M. A. (1998). The energetics of male mating strategies in field crickets (Orthoptera:
Gryllinae: Gryllidae). Journal of Insect Behavior, 11(6), 853–867.
http://doi.org/10.1023/a:1020864111073
Haldane, J. B. S. (1922). Sex ratio and unisexual sterility in hybrid animals. Journal of
Genetics, 12(2), 101–109.
Hall, D. W. (2004). Meiotic Drive and Sex Chromosome Cycling. Evolution, 58(5), 925–931.
Hall, M. D., Lailvaux, S. P., Blows, M. W., & Brooks, R. C. (2010). Sexual conflict and the
maintenance of multivariate genetic variation. Evolution; International Journal of
Organic Evolution, 64(6), 1697–703. http://doi.org/10.1111/j.1558-5646.2009.00932.x
Hardin, G. (1960). The competitive exclusion principle. Science.
http://doi.org/10.1126/science.131.3409.1292
Harrison, R. G. (1986). Pattern and process in a narrow hybrid zone, 56(August 1985).
Harrison, R. G. (1993). Hybrid zones and the evolutionary process. (R. G. Harrison, Ed.).
Oxford University Press.
Harrison, R. G., & Larson, E. L. (2014). Hybridization, Introgression, and the Nature of
Species Boundaries. Journal of Heredity, 105(S1), 795–809.
http://doi.org/10.1093/jhered/esu033
Harrison, R. G., & Larson, E. L. (2016). Heterogeneous genome divergence, differential
introgression, and the origin and structure of hybrid zones. Molecular Ecology, 2454–
2466. http://doi.org/10.1111/mec.13582
Harrison, R. G., & Rand, D. M. (1989). Mosaic Hybrid Zones and the Nature of Species
Boundaries. In Speciation and its consequences.
Head, M. L., Price, E. A., & Boughman, J. W. (2009). Body size differences do not arise
from divergent mate preferences in a species pair of threespine stickleback. Biology
230
Letters, 5(May), 517–520. http://doi.org/10.1098/rsbl.2009.0216
Hedrick, P. W. (2013). Adaptive introgression in animals: Examples and comparison to new
mutation and standing variation as sources of adaptive variation. Molecular Ecology, 22,
4606–4618. http://doi.org/10.1111/mec.12415
Hennig, R. M., & Weber, T. (1997). Filtering of temporal parameters of the calling song by
cricket females of two closely related species: a behavioral analysis. Journal of
Comparative Physiology A: Sensory, Neural, and Behavioral Physiology, 180(6), 621–
630.
Henry, C. S., Martínez Wells, M. L., & Holsinger, K. E. (2002). The inheritance of mating
songs in two cryptic, sibling lacewing species (Neuroptera: Chrysopidae: Chrysoperla).
Genetica, 116(2–3), 269–89.
Hewitt, G. M. (1988). Hybrid zones-natural laboratories for evolutionary studies. Trends in
Ecology and Evolution, 7, 158–167.
Hewitt, G. M. (2000). The genetic legacy of the Quaternary ice ages. Nature, 405(6789),
907–913. http://doi.org/10.1038/35016000
Hewitt, G. M. (2001). Speciation, hybrid zones and phylogeography — or seeing genes in
space and time. Molecular Ecology, 10(3), 537–549. http://doi.org/10.1046/j.1365294X.2001.01202.x
Higgie, M., & Blows, M. W. (2007). Are Traits That Experience Reinforcement Also Under
Sexual Selection? The American Naturalist, 170(3), 409–420.
http://doi.org/10.1086/519401
Higgie, M., Chenoweth, S. F., & Blows, M. W. (2000). Natural Selection and the
Reinforcement of Mate Recognition. Science, 290(5491), 519–521.
Higgins, L. a., & Waugaman, R. D. (2004). Sexual selection and variation: a multivariate
approach to species-specific calls and preferences. Animal Behaviour, 68(5), 1139–
1153. http://doi.org/10.1016/j.anbehav.2003.10.035
Hill, K. G., Loftus-Hills, J. J., & Gartside, D. F. (1972). Pre-mating isolation between the
Australian field crickets Teleogryllus commodus and T. oceanicus (Orthoptera :
Gryllidae). Aust. J. Zool., (20), 153–163.
Hogan, T. W. (1971). An evaluation of a genetic method for population suppression of
Teleogryllus commodus ( Wlk .) ( Orth ., Gryllidae ) in Victoria. Bull. Entomol. Res, 60,
383–390.
Hogan, T. W., & Fontana, P. G. (1973). Restoration of meiotic stability following artificial
hybridisation and selection in Teleogryllus ( Orth ., Gryllidae ). Bulletin of
Entomological Research, 62(4), pp557-563.
Hollocher, H., & Wu, C. I. (1996). The genetics of reproductive isolation in the Drosophila
simulans clade: X vs. autosomal effects and male vs. female effects. Genetics, 143(3),
1243–1255.
Holsinger, K. E., & Weir, B. S. (2009). Genetics in geographically structured populations:
defining, estimating and interpreting F(ST). Nature Reviews. Genetics, 10(9), 639–650.
http://doi.org/10.1038/nrg2611
231
Hondasumi, E. (2005). Difference in calling song of three field crickets of the genus : the role
in premating isolation. Animal Behaviour, 69(4), 881–889.
http://doi.org/10.1016/j.anbehav.2004.05.015
Hopkins, R., Guerrero, R. F., Rausher, M. D., & Kirkpatrick, M. (2014). Strong reinforcing
selection in a Texas wildflower. Current Biology, 24(17), 1995–1999.
http://doi.org/10.1016/j.cub.2014.07.027
Hothorn, Torsten, Bretz, F., & Westfall, P. (2008). Simultaneous inference in general
parametric models. Biometrical Journal, 3, 346–363.
Howard, D. J. (1993). Reinforcement: origin, dynamics, and fate of an evolutionary
hypothesis. In Hybrid zones and the evolutionary process (pp. 46–69).
Howard, D. J., Gregory, P. G., Chu, J., Cain, M. L., & Url, S. (1998). Conspecific Sperm
Precedence is an Effective Barrier to Hybridization Between Closely Related Species.
Evolution, 52(2), 511–516.
Howard, D. J., Marshall, J. L., Hampton, D. D., Britch, S. C., Draney, M. L., Chu, J., &
Cantrell, R. G. (2002). The genetics of reproductive isolation: a retrospective and
prospective look with comments on ground crickets. The American Naturalist, 159
Suppl(May), S8–S21. http://doi.org/10.1086/338369
Howard, R. W., & Blomquist, G. J. (2005). Ecological , Behavioral , and Biochemical
Aspects of Insect Hydrocarbons. Annual Review of Entomology, 50, 371–393.
http://doi.org/10.1146/annurev.ento.50.071803.130359
Hoy, R. R. . (1974). Genetic Control of Acoustic Behavior in Crickets. American Zoologist,
14(3), 1067–1080.
Hoy, R. R., Hahn, J., & Paul, R. C. (1977). Hybrid Cricket Auditory Behavior : Evidence for
Genetic Coupling in Animal Communication. Science.
Hoy, R. R., & Paul, R. C. (1973). Genetic control of song specificity in crickets. Science
(New York, N.Y.), 180(4081), 82–3. http://doi.org/10.1126/science.180.4081.82
Hoy, R. R., Pollack, G. S., & Moiseff, A. (1982). Species-Recognition in the Field Cricket,
Teleogryllus oceanicus: Behavioral and Neural Mechanisms. Amer. Zool, 607, 597–607.
Huber, F., Moore, T. E., & Loher., W. (1989). Cricket behavior and neurobiology. Ithaca:
Comstock Pub. Associates.
Hunt, J., Blows, M. W., Zajitschek, F., Jennions, M. D., & Brooks, R. (2007). Reconciling
strong stabilizing selection with the maintenance of genetic variation in a natural
population of black field crickets (Teleogryllus commodus). Genetics, 177(2), 875–880.
http://doi.org/10.1534/genetics.107.077057
Jablonka, E., & Lamb, M. J. (1991). Sex chromosomes and speciation. Proceedings.
Biological Sciences / The Royal Society, 243(1308), 203–208.
http://doi.org/10.1098/rspb.1991.0032
Jallon, J.-M. (1984). A Few Chemical Words Exchanged by Drosophila During Courtship
and Mating. Behavior Genetics, 14(5).
Janoušek, V., Wang, L., Luzynski, K., Dufková, P., Vyskočilová, M. M., Nachman, M. W.,
… Tucker, P. K. (2012). Genome-wide architecture of reproductive isolation in a
232
naturally occurring hybrid zone between Mus musculus musculus and M. m.
domesticus. Molecular Ecology, 21(12), 3032–47. http://doi.org/10.1111/j.1365294X.2012.05583.x
Jean-Marc Jallon, & David, J. R. . (1987). Variation in Cuticular Hydrocarbons Among the
Eight Species of the Drosophila melanogaster Subgroup. Evolution; International
Journal of Organic Evolution, 41(2), 294–302.
Jiggins, C., & Mallet, J. (2000). Bimodal hybrid zones and speciation. Trends in Ecology &
Evolution, 15(6), 250–255.
Johannesson, K., Johannesson, B., & Rolan-Alvarez, E. (1993). Morphological
Differentiation and Genetic Cohesiveness Over a Microenvironmental Gradient in the
Marine Snail Littorina saxatilis. Evolution, 47(6), 1770–1787.
Johnson, N. A. (2010). Hybrid incompatibility genes : remnants of a genomic battlefield ?
Trends in Genetics, 26(7), 317–325. http://doi.org/10.1016/j.tig.2010.04.005
Jombart, T. (2008). Adegenet: A R package for the multivariate analysis of genetic markers.
Bioinformatics, 24(11), 1403–1405. http://doi.org/10.1093/bioinformatics/btn129
Jombart, T., & Ahmed, I. (2011). adegenet 1.3-1: New tools for the analysis of genome-wide
SNP data. Bioinformatics, 27(21), 3070–3071.
http://doi.org/10.1093/bioinformatics/btr521
Jombart, T., Devillard, S., & Balloux, F. (2010). Discriminant analysis of principal
components: a new method for the analysis of genetically structured populations. BMC
Genetics, 11, 94. http://doi.org/10.1186/1471-2156-11-94
Kamvar, Z. N., Brooks, J. C., & Grunwald, N. J. (2015). Novel R tools for analysis of
genome-wide population genetic data with emphasis on clonality. Frontiers in Genetics,
6(JUN), 1–10. http://doi.org/10.3389/fgene.2015.00208
Kamvar, Z. N., Tabima, J. F., & Grünwald, N. J. (2014). Poppr: an R package for genetic
analysis of populations with clonal, partially clonal, and/or sexual reproduction. PeerJ,
2, e281. http://doi.org/10.7717/peerj.281
Kaneshiro, K. Y. (1980). Sexual Isolation , Speciation and the Direction of Evolution.
Evolution, 34(3), 437–444.
Kauer, M., Zangerl, B., Dieringer, D., & Schlötterer., C. (2002). Chromosomal patterns of
microsatellite variability contrast sharply in African and non-African populations of
Drosophila melanogaster. Genetics, 160(1), 247–256. http://doi.org/north-american
populations DNA polymorphism data x-linked loci natural-populations recombination
rates inversion polymorphism genetic hitchhiking molecular variation sexual isolation
mutation-rates
Kawakami, T., Butlin, R. K., Adams, M., Paull, D. J., & Cooper, S. J. B. (2009). Genetic
analysis of a chromosomal hybrid zone in the Australian morabine grasshoppers
(Vandiemenella, viatica species group). Evolution, 63(1), 139–52.
http://doi.org/10.1111/j.1558-5646.2008.00526.x
Keinan, A., Mullikin, J. C., Patterson, N., & Reich, D. (2009). Accelerated genetic drift on
chromosome X during the human dispersal out of Africa. Nature Genetics, 41(1), 66–70.
http://doi.org/10.1038/ng.303
233
Keller, I., Wagner, C. E., Greuter, L., Mwaiko, S., Selz, O. M., Sivasundar, a., … Seehausen,
O. (2013). Population genomic signatures of divergent adaptation, gene flow and hybrid
speciation in the rapid radiation of Lake Victoria cichlid fishes. Molecular Ecology,
22(11), 2848–2863. http://doi.org/10.1111/mec.12083
Kelly, J. K., & Noor, M. A. F. (1996). Speciation by reinforcement: A model derived from
studies of Drosophila. Genetics, 143(3), 1485–1497.
Kirkpatrick, M. (1982). Sexual selection and the evolution of female choice. Evolution, 36(1),
1–12.
Kirkpatrick, M., & Hall, D. W. (2004). Sexual selection and sex linkage. Evolution, 58(4),
683–691.
Kirkpatrick, M., & Ravigné, V. (2002). Speciation by natural and sexual selection: models
and experiments. The American Naturalist, 159(S3), S22–S35.
Knowles, L. L., Massatti, R., He, Q., Olson, L. E., & Lanier, H. C. (2016). Quantifying the
similarity between genes and geography across Alaska’s alpine small mammals. Journal
of Biogeography. http://doi.org/10.1111/jbi.12728
Koevoets, T., & Beukeboom, L. W. (2009). Genetics of postzygotic isolation and Haldane’s
rule in haplodiploids. Heredity, 102(1), 16–23. http://doi.org/10.1038/hdy.2008.44
Kokko, H., Jennions, M. D., & Brooks, R. (2013). Unifying and Testing Unifying Models of
Sexual Selection. Annual Review of Ecology, Evolution, and Systematics and
Systematics, 37(2006), 43–66.
http://doi.org/10.2307/annurev.ecolsys.37.091305.30000004
Kozlowska, J. L., Ahmad, A. R., Jahesh, E., & Cutter, A. D. (2012). Genetic variation for
postzygotic reproductive isolation between caenorhabditis briggsae and caenorhabditis
sp. 9. Evolution, 66(4), 1180–1195. http://doi.org/10.1111/j.1558-5646.2011.01514.x
Kraaijeveld, K., Kraaijeveld-Smit, F. J. L., & Maan, M. E. (2011). Sexual selection and
speciation: The comparative evidence revisited. Biological Reviews, 86(2), 367–377.
http://doi.org/10.1111/j.1469-185X.2010.00150.x
Kronforst, M. R., Hansen, M. E. B., Crawford, N. G., Gallant, J. R., Zhang, W., Kulathinal,
R. J., … Mullen, S. P. (2013). Hybridization Reveals the Evolving Genomic
Architecture of Speciation. Cell Reports, 5(3), 666–677.
http://doi.org/10.1016/j.celrep.2013.09.042
Lande, R. (1981). Models of speciation by sexual selection on polygenic traits. Proceedings
of the National Academy of Sciences of the United States of America, 78(6), 3721–5.
Lande, R. (1982). Rapid Origin of Sexual Isolation and Character Divergence in a Cline.
Society for the Study of Evolution, 36(2), 213–223.
Laurent, S., Pfeifer, S. P., Settles, M. L., Hunter, S. S., Hardwick, K. M., Ormond, L., …
Rosenblum, E. B. (2016). The population genomics of rapid adaptation: Disentangling
signatures of selection and demography in white sands lizards. Molecular Ecology,
25(1), 306–323. http://doi.org/10.1111/mec.13385
Laurie, C. C. (1997). The weaker Sex is Heterogametic: 75 Years of Haldane’s Rule.
Genetics, 154(4), 1419–1426.
234
Lavretsky, P., Dacosta, J. M., Hern??ndez-Ba??os, B. E., Engilis, A., Sorenson, M. D., &
Peters, J. L. (2015). Speciation genomics and a role for the Z chromosome in the early
stages of divergence between Mexican ducks and mallards. Molecular Ecology, 24(21),
5364–5378. http://doi.org/10.1111/mec.13402
Lê, S., Josse, J., & Husson, F. (2008). FactoMineR : An R package for multivariate analysis.
J Stat Softw, 25(1), 1–18. http://doi.org/10.1016/j.envint.2008.06.007
Lemmon, A. R., & Kirkpatrick, M. (2006). Reinforcement and the genetics of hybrid
incompatibilities. Genetics, 173(2), 1145–1155.
http://doi.org/10.1534/genetics.105.048199
Lemmon, E. M., & Lemmon, A. R. (2010). Reinforcement in chorus frogs: lifetime fitness
estimates including intrinsic natural selection and sexual selection against hybrids.
Evolution; International Journal of Organic Evolution, 64(6), 1748–61.
http://doi.org/10.1111/j.1558-5646.2010.00955.x
Leonard, A. S., & Hedrick, A. V. (2009). Single versus multiple cues in mate discrimination
by males and females. Animal Behaviour, 77(1), 151–159.
http://doi.org/10.1016/j.anbehav.2008.09.029
Lexer, C., Mangili, S., Bossolini, E., Forest, F., Stölting, K. N., Pearman, P. B., … Salamin,
N. (2013). “Next generation” biogeography: Towards understanding the drivers of
species diversification and persistence. Journal of Biogeography, 40(6), 1013–1022.
http://doi.org/10.1111/jbi.12076
Li, H. (2011). A statistical framework for SNP calling, mutation discovery, association
mapping and population genetical parameter estimation from sequencing data.
Bioinformatics, 2987–2993.
Li, H., & Durbin, R. (2009). Fast and accurate short read alignment with Burrows-Wheeler
transform. Bioinformatics, 25(14), 1754–1760.
http://doi.org/10.1093/bioinformatics/btp324
Li, H., Handsaker, B., Wysoker, A., Fennell, T., Ruan, J., Homer, N., … Goncalo Abecasis,
R. D. (2009). The sequence alignment/map format and SAMtools. Bioinformatics, 16,
2078–2079.
Li, W.-H., Soojin Yi, & Makova, K. (2008). Male-driven Evolution. Current Opinion in
Genetics & Development, 1–3. http://doi.org/10.1002/9780470015902.a0005303.pub2
Limousin, D., Streiff, R., Courtois, B., Dupuy, V., Alem, S., & Greenfield, M. D. (2012).
Genetic architecture of sexual selection: QTL mapping of male song and female receiver
traits in an acoustic moth. PloS One, 7(9), e44554.
http://doi.org/10.1371/journal.pone.0044554
Liou, L. W., & Price, T. D. (1994). Speciation by Reinforcement of Premating Isolation.
Evolution, 48(5), 1451–1459.
Lischer, H. E. L., & Excoffier, L. (2012). PGDSpider: An automated data conversion tool for
connecting population genetics and genomics programs. Bioinformatics, 28(2), 298–299.
http://doi.org/10.1093/bioinformatics/btr642
Loher, W., & Rence, B. (1978). The Mating Behaviour of Teleogryllus commodus (Walker)
and its Central and Peripheral Control. Zeitschrift Für Tierpsychologie, 46(3), 225–259.
235
Lowry, D. B., Modliszewski, J. L., Wright, K. M., Wu, C. a, & Willis, J. H. (2008). Review.
The strength and genetic basis of reproductive isolating barriers in flowering plants.
Philosophical Transactions of the Royal Society of London. Series B, Biological
Sciences, 363(1506), 3009–3021. http://doi.org/10.1098/rstb.2008.0064
Lucotte, E. A., Laurent, R., Heyer, E., Ségurel, L., & Toupance, B. (2016). Detection of
allelic frequency differences between the sexes in humans: a signature of sexually
antagonistic selection. Genome Biology and Evolution, evw090.
http://doi.org/10.1093/gbe/evw090
Luikart, G., England, P. R., Tallmon, D., Jordan, S., & Taberlet, P. (2003). The power and
promise of population genomics: from genotyping to genome typing. Nature Reviews.
Genetics, 4(12), 981–94. http://doi.org/10.1038/nrg1226
Lynch;, M., & Walsh, B. (1999). Genetics and analysis of quantitative traits (2nd ed.).
Sunderland, Mass. : Sinauer.
Maan, M. E., & Seehausen, O. (2011). Ecology, sexual selection and speciation. Ecology
Letters, 14(6), 591–602. http://doi.org/10.1111/j.1461-0248.2011.01606.x
Machado, H. E., Bergland, A. O., O’Brien, K. R., Behrman, E. L., Schmidt, P. S., & Petrov,
D. A. (2016). Comparative population genomics of latitudinal variation in Drosophila
simulans and Drosophila melanogaster. Molecular Ecology, 25(3), 723–740.
http://doi.org/10.1111/mec.13446
Maheshwari, S., & Barbash, D. a. (2011). The genetics of hybrid incompatibilities. Annual
Review of Genetics, 45, 331–55. http://doi.org/10.1146/annurev-genet-110410-132514
Mallet, J. (2005). Hybridization as an invasion of the genome. Trends in Ecology &
Evolution, 20(5), 229–37. http://doi.org/10.1016/j.tree.2005.02.010
Mallet, J. (2007). Hybrid speciation. Nature, 446(7133), 279–283.
http://doi.org/10.1038/nature05706
Mallet, J. (2008). A Century of Evolution : Ernst Mayr ( 1904 – 2005 ) Mayr ’ s view of
Darwin : was Darwin wrong about speciation ? Biological Journal of the Linnean
Society, 95, 3–16.
Mallet, J., Beltrán, M., Neukirchen, W., & Linares, M. (2007). Natural hybridization in
heliconiine butterflies: the species boundary as a continuum. BMC Evolutionary
Biology, 7, 28. http://doi.org/10.1186/1471-2148-7-28
Malone, J. H., & Michalak, P. (2008). Physiological Sex Predicts Hybrid Sterility Regardless
of Genotype. Science, 319(January), 2008.
Mank, J. E., Axelsson, E., & Ellegren, H. (2007). Fast-X on the Z: Rapid evolution of sexlinked genes in birds. Genome Research, 17(5), 618–624.
http://doi.org/10.1101/gr.6031907
Mank, J. E., Hosken, D. J., & Wedell, N. (2011). Some inconvenient truths about sex
chromosome dosage compensation and the potential role of sexual conflict. Evolution,
65(8), 2133–2144. http://doi.org/10.1111/j.1558-5646.2011.01316.x
Mank, J. E., Vicoso, B., Berlin, S., & Charlesworth, B. (2010). Effective population size and
the Faster-X effect: Empirical results and their interpretation. Evolution, 64(3), 663–674.
http://doi.org/10.1111/j.1558-5646.2009.00853.x
236
Mantovani, B., & Scali, V. (1992). Hybridogenesis and Androgenesis in the Stick-Insect
Bacillus rossius- Grandii benazzii. Evolution, 46(3), 783–796.
Maroja, L. S., Andrés, J. a, & Harrison, R. G. (2009). Genealogical discordance and patterns
of introgression and selection across a cricket hybrid zone. Evolution; International
Journal of Organic Evolution, 63(11), 2999–3015. http://doi.org/10.1111/j.15585646.2009.00767.x
Maroja, L. S., Larson, E. L., Bogdanowicz, S. M., & Harrison, R. G. (2015). Genes with
Restricted Introgression in a Field Cricket (Gryllus firmus/Gryllus pennsylvanicus)
Hybrid Zone Are Concentrated on the X Chromosome and a Single Autosome. G3,
5(11), 2219–2227. http://doi.org/10.1534/g3.115.021246
Maroja, L. S., McKenzie, Z. M., Hart, E., Jing, J., Larson, E. L., & Richardson, D. P. (2014).
Barriers to gene exchange in hybridizing field crickets: the role of male courtship effort
and cuticular hydrocarbons. BMC Evolutionary Biology, 14(1), 65.
http://doi.org/10.1186/1471-2148-14-65
Marshall, J. L., Arnold, M. L., & Howard, D. J. (2002). Reinforcement : the road not taken.
Trends in Ecology and Evolution, 17(12), 558–563.
Martín-Coello, J., Benavent-Corai, J., Roldan, E. R. S., & Gomendio, M. (2009). Sperm
competition promotes asymmetries in reproductive barriers between closely related
species. Evolution, 63(3), 613–623. http://doi.org/10.1111/j.1558-5646.2008.00585.x
Martin, A. D., Quinn, K. M., & Park, J. H. (2011). MCMCpack: Markov chain Monte Carlo
in R. Journal Of Statistical Software, 42(Ii), 1–22.
Martin, A., McCulloch, K. J., Patel, N. H., Briscoe, A. D., Gilbert, L. E., & Reed, R. D.
(2014). Multiple recent co-options of Optix associated with novel traits in adaptive
butterfly wing radiations. EvoDevo, 5(1), 7. http://doi.org/10.1186/2041-9139-5-7
Masly, J. P., & Presgraves, D. C. (2007). High-resolution genome-wide dissection of the two
rules of speciation in Drosophila. PLoS Biology, 5(9), 1890–1898.
http://doi.org/10.1371/journal.pbio.0050243
Mavárez, J., Salazar, C. a, Bermingham, E., Salcedo, C., Jiggins, C. D., & Linares, M.
(2006). Speciation by hybridization in Heliconius butterflies. Nature, 441(7095), 868–
71. http://doi.org/10.1038/nature04738
Mayr, E. (1942). Systematics and the origin of species, from the viewpoint of a zoologist.
Harvard University Press.
McDermott, S. R., & Noor, M. a F. (2010). The role of meiotic drive in hybrid male sterility.
Philosophical Transactions of the Royal Society of London. Series B, Biological
Sciences, 365(1544), 1265–1272. http://doi.org/10.1098/rstb.2009.0264
McVean, G. (2009). A genealogical interpretation of principal components analysis. PLoS
Genetics, 5(10). http://doi.org/10.1371/journal.pgen.1000686
Mead, L. S., & Arnold, S. J. (2004). Quantitative genetic models of sexual selection. Trends
in Ecology & Evolution, 19(5), 264–71. http://doi.org/10.1016/j.tree.2004.03.003
Meiklejohn, C. D., & Tao, Y. (2010). Genetic conflict and sex chromosome evolution.
Trends in Ecology and Evolution, 25(4), 215–223.
http://doi.org/10.1016/j.tree.2009.10.005
237
Meirmans, P. G. (2012). The trouble with isolation by distance. Molecular Ecology, 21(12),
2839–2846. http://doi.org/10.1111/j.1365-294X.2012.05578.x
Mendelson, T. C., & Shaw, K. L. (2005). Rapid speciation in an arthropod. Nature,
433(7024), E9; discussion E9.
Mendelson, T. C., & Shaw, K. L. (2012). The (mis)concept of species recognition. Trends in
Ecology & Evolution, 27(8), 421–7.
Mendelson, T. C., & Shaw, K. L. (2013). Further misconceptions about species recognition: a
reply to Padian and Horner. Trends in Ecology & Evolution, 28(5), 252–3.
http://doi.org/10.1016/j.tree.2013.02.012
Michalak, P., & Noor, M. A. F. (2003). Genome-wide patterns of expression in Drosophila
pure species and hybrid males. Molecular Biology and Evolution, 20(7), 1070–1076.
http://doi.org/10.1093/molbev/msg119
Moehring, A. J., Llopart, A., Elwyn, S., Coyne, J. A., & Mackay, T. F. C. (2006). The genetic
basis of postzygotic reproductive isolation between Drosophila santomea and D. yakuba
due to hybrid male sterility. Genetics, 173(1), 225–233.
http://doi.org/10.1534/genetics.105.052985
Moritz, C., Hoskin, C. J., MacKenzie, J. B., Phillips, B. L., Tonione, M., Silva, N., …
Graham, C. H. (2009). Identification and dynamics of a cryptic suture zone in tropical
rainforest. Proceedings. Biological Sciences / The Royal Society, 276(1660), 1235–44.
http://doi.org/10.1098/rspb.2008.1622
Moura, A. E., Kenny, J. G., Chaudhuri, R., Hughes, M. A., J. Welch, A., Reisinger, R. R., …
Hoelzel, A. R. (2014). Population genomics of the killer whale indicates ecotype
evolution in sympatry involving both selection and drift. Molecular Ecology, 23(21),
5179–5192. http://doi.org/10.1111/mec.12929
Mousseau, T. (1991). Maternal Effects In Insect Life Histories. Annual Review of
Entomology, 36(1), 511–534. http://doi.org/10.1146/annurev.ento.36.1.511
Moyle, L. C., & Nakazato, T. (2008). Comparative genetics of hybrid incompatibility:
Sterility in two solanum species crosses. Genetics, 179(3), 1437–1453.
http://doi.org/10.1534/genetics.107.083618
Mullen, S. P., Mendelson, T. C., Schal, C., & Shaw, K. L. (2007). Rapid evolution of
cuticular hydrocarbons in a species radiation of acoustically diverse Hawaiian crickets
(Gryllidae: trigonidiinae: Laupala). Evolution; International Journal of Organic
Evolution, 61(1), 223–31. http://doi.org/10.1111/j.1558-5646.2007.00019.x
Muller, H. J. (1942). Isolating mechanisms, evolution, and temperature. Biol. Symp, 6, 71–
125.
Nadeau, N. J., Ruiz, M., Salazar, P., Counterman, B., Medina, J. A., Ortiz-Zuazaga, H., …
Papa, R. (2014). Population genomics of parallel hybrid zones in the mimetic butterflies,
H. melpomene and H. erato. Genome Research, 24(8), 1316–1333.
http://doi.org/10.1101/gr.169292.113
Nadeau, N. J., Whibley, A., Jones, R. T., Davey, J. W., Dasmahapatra, K. K., Baxter, S. W.,
… Jiggins, C. D. (2012). Genomic islands of divergence in hybridizing Heliconius
butterflies identified by large-scale targeted sequencing. Philosophical Transactions of
the Royal Society of London. Series B, Biological Sciences, 367(1587), 343–53.
238
http://doi.org/10.1098/rstb.2011.0198
Nagel, L. & Schluter, D. (1998). Body Size, Natural Selection, and Speciation in
Sticklebacks. Evolution, 209–218.
Naisbit, R. E., Jiggins, C. D., & Mallet, J. (2001). Disruptive sexual selection against hybrids
contributes to speciation between Heliconius cydno and Heliconius melpomene.
Proceedings of the Royal Society of London B: Biological Sciences, 268(1478), 1849–
1854. http://doi.org/10.1098/rspb.2001.1753
Nei, M. (1987). Molecular evolutionary genetics. Columbia university press.
Nei, M., Maruyama, T., & Wu, C. I. (1983). Models of evolution of reproductive isolation.
Genetics, 103(3), 557–579.
Noor, M. A. F. (1999). Reinforcement and other consequences of sympatry. Heredity, 83(5),
503–508.
Noor, M. A. F., Grams, K. L., Bertucci, L. a, & Reiland, J. (2001). Chromosomal inversions
and the reproductive isolation of species. Proceedings of the National Academy of
Sciences of the United States of America, 98(21), 12084–8.
http://doi.org/10.1073/pnas.221274498
Nosil, P. (2012). Ecological speciation. Oxford University Press.
Nosil, P., Crespi, B. J., Gries, R., & Gries, G. (2007). Natural selection and divergence in
mate preference during speciation. Genetica, 129(3), 309–327.
http://doi.org/10.1007/s10709-006-0013-6
Nosil, P., Crespi, B. J., & Sandoval, C. P. (2003). Reproductive isolation driven by the
combined effects of ecological adaptation and reinforcement. Proceedings. Biological
Sciences / The Royal Society, 270(1527), 1911–1918.
http://doi.org/10.1098/rspb.2003.2457
Nosil, P., Egan, S. P., & Funk, D. J. (2007). Heterogeneous genomic differentiation between
walking‐stick ecotypes:“isolation by adaptation” and multiple roles for divergent
selection. Evolution, 316–336. http://doi.org/10.1111/j.1558-5646.2007.00299.x
Nosil, P., & Feder, J. L. (2012). Genomic divergence during speciation: causes and
consequences. Philosophical Transactions of the Royal Society of London. Series B,
Biological Sciences, 367(1587), 332–42. http://doi.org/10.1098/rstb.2011.0263
Nosil, P., Funk, D. J., & Ortiz-Barrientos, D. (2009). Divergent selection and heterogeneous
genomic divergence. Molecular Ecology, 18(3), 375–402. http://doi.org/10.1111/j.1365294X.2008.03946.x
Nosil, P., Vines, T. H., & Funk, D. J. (2005). Reproductive isolation caused by natural
selection against immigrants from divergent habitats. Evolution; International Journal
of Organic Evolution, 59, 705–719.
Oh, K. P., Fergus, D. J., Grace, J. L., & Shaw, K. L. (2012). Interspecific genetics of
speciation phenotypes: song and preference coevolution in Hawaiian crickets. Journal of
Evolutionary Biology, 25(8), 1500–12. http://doi.org/10.1111/j.1420-9101.2012.02531.x
Ohmachi, F., & Masaki, S. (1964). Interspecific Crossing and Development of Hybrids
between the Japanese Species of Teleogryllus (Orthoptera: Gryllidae). Society for the
239
Study of Evolution, 18(3), 405–416.
Oksanen, A. J., Kindt, R., Legendre, P., Hara, B. O., Simpson, G. L., Stevens, M. H. H., …
Wagner, H. (2008). The vegan package. … Ecology Package, (January), 190.
Orr, H. A. (1987). Genetics of Male and Female Sterility in Hybrids of, 563.
Orr, H. A. (1993a). A mathematical Model of Haldane’s Rule. Evolution, 47(5), 1606–1611.
Orr, H. A. (1993b). Haldane’s rule has multiple genetic causes. Nature, 361(6412), 532–533.
http://doi.org/10.1038/361532a0
Orr, H. A., & Coyne, J. A. (1989). The genetics of postzygotic isolation in the Drosophila
virilis group. Genetics, 121, 527–537.
Orr, H. A., & Irving, S. (2001). Complex epistasis and the genetic basis of hybrid sterility in
the Drosophila pseudoobscura Bogota-USA hybridization. Genetics, 158(3), 1089–1100.
http://doi.org/10.1534/genetics.104.033274
Ortíz-Barrientos, D., Counterman, B. a, & Noor, M. a F. (2007). Gene expression divergence
and the origin of hybrid dysfunctions. Genetica, 129(1), 71–81.
http://doi.org/10.1007/s10709-006-0034-1
Ortiz-Barrientos, D., Grealy, A., & Nosil, P. (2009). The genetics and ecology of
reinforcement: Implications for the evolution of prezygotic isolation in sympatry and
beyond. Annals of the New York Academy of Sciences, 1168, 156–182.
http://doi.org/10.1111/j.1749-6632.2009.04919.x
Otte, D. (1989). Speciation in Hawaiian crickets. In Speciation and its Consequences (pp.
482–526).
Otte, D. (1992). Evolution of Cricket Songs. Journal of Orthoptera Research, 1(1), 25–49.
Otte, D., & Alexander, R. D. (1983). The Australian crickets (Orthoptera: Gryllidae).
Padian, K., & Horner, J. R. (2013). Misconceptions of sexual selection and species
recognition: a response to Knell et al. and to Mendelson and Shaw. Trends in Ecology &
Evolution, 28(5), 249–50. http://doi.org/10.1016/j.tree.2013.01.011
Panhuis, T. M., Butlin, R., Zuk, M., & Tregenza, T. (2001). Sexual selection and speciation.
Trends in Ecology & Evolution, 16(7), 364–371.
Parchman, T. L., Gompert, Z., Braun, M. J., Brumfield, R. T., McDonald, D. B., Uy, J. a C.,
… Buerkle, C. a. (2013). The genomic consequences of adaptive divergence and
reproductive isolation between species of manakins. Molecular Ecology, 22(12), 3304–
3317. http://doi.org/10.1111/mec.12201
Pascoal, S., Cezard, T., Eik-Nes, A., Gharbi, K., Majewska, J., Payne, E., … Bailey, N. W.
(2014). Rapid Convergent Evolution in Wild Crickets. Current Biology, 1–6.
http://doi.org/10.1016/j.cub.2014.04.053
Pascoal, S., Mendrok, M., Mitchell, C., Wilson, A. J., Hunt, J., & Bailey, N. W. (2015).
Sexual selection and population divergence I. the influence of socially flexible cuticular
hydrocarbon expression in male field crickets (Teleogryllus oceanicus ). Evolution, n/an/a. http://doi.org/10.1111/evo.12839
Paterson, H. E. H. (1985). The Recognition Concept of Species. Species and speciation (Vol.
240
4). Species and Speciation. Transvaal Museum Monograph.
Payne, R. J. H., & Krakauer, D. C. (1997). Sexual Selection, Space, and Speciation. Evol,
51(1), 1–9.
Payseur, B. A. (2010). Using differential introgression in hybrid zones to identify genomic
regions involved in speciation. Molecular Ecology Resources, 10(5), 806–20.
http://doi.org/10.1111/j.1755-0998.2010.02883.x
Payseur, B. A., & Rieseberg, L. H. (2016). A Genomic Perspective on Hybridization and
Speciation. Molecular Ecology, n/a-n/a. http://doi.org/10.1111/mec.13557
Peng, Y., Leung, H. C. M., Yiu, S. M., & Chin, F. Y. L. (2012). IDBA-UD: A de novo
assembler for single-cell and metagenomic sequencing data with highly uneven depth.
Bioinformatics, 28(11), 1420–1428. http://doi.org/10.1093/bioinformatics/bts174
Peterson, M. A., Honchak, B. M., Locke, S. E., Beeman, T. E., Mendoza, J., Green, J., …
Monsen, K. J. (2005). Relative abundance and the species-specific reinforcement of
male mating preference in the Chrysochus (Coleoptera: Chrysomelidae) hybrid zone.
Evolution, 59(12), 2639–2655. http://doi.org/10.1554/05-120.1
Pfennig, D. W., Rice, A. M., & Martin, R. A. (2006). Ecological opportunity and phenotypic
plasticity interact to promote character displacement and species coexistence. Ecology,
87(3), 769–779. http://doi.org/10.1890/05-0787
Pfennig, K. S. (1998). The evolution of mate choice and the potential for conflict between
species and mate-quality recognition. Proceedings of the Royal Society B: Biological
Sciences, 265(1407), 1743–1748. http://doi.org/10.1098/rspb.1998.0497
Pfennig, K. S. (2007). Facultative mate choice drives adaptive hybridization. Science (New
York, N.Y.), 318(5852), 965–7. http://doi.org/10.1126/science.1146035
Pfennig, K. S., & Pfennig, D. W. (2009). Character displacement: ecological and
reproductive responses to a common evolutionary problem. Quarterly Review of
Biology, 84(3), 253–276. http://doi.org/10.1016/j.surg.2006.10.010.Use
Pitchers, W. R., Brooks, R., Jennions, M. D., Tregenza, T., Dworkin, I., & Hunt, J. (2013).
Limited plasticity in the phenotypic variance-covariance matrix for male advertisement
calls in the black field cricket, Teleogryllus commodus. Journal of Evolutionary
Biology, 26(5), 1060–78. http://doi.org/10.1111/jeb.12120
Poelstra, J. W., Vijay, N., Bossu, C. M., Lantz, H., Ryll, B., Müller, I., … Wolf, J. B. W.
(2014). The genomic landscape underlying phenotypic integrity in the face of gene flow
in crows. Science (New York, N.Y.), 344(6190), 1410–4.
http://doi.org/10.1126/science.1253226
Pollack, G. S. ;, & Hoy, R. R. . (1979). Temporal Pattern as a cue for Species-Specific
Calling Song Recognition in Circkets. Science, 204(4391), 429–432.
Pool, J. E., & Nielsen, R. (2007). Population size changes reshape genomic patterns of
diversity. Evolution, 61(12), 3001–3006. http://doi.org/10.1111/j.15585646.2007.00238.x
Presgraves, D. C. (2002). Patterns of postzygotic isolation in Lepidoptera. Evolution;
International Journal of Organic Evolution, 56(6), 1168–1183.
http://doi.org/10.2307/2411351
241
Presgraves, D. C. (2008). Sex chromosomes and speciation in Drosophila. Trends in
Genetics, 24(7), 336–343. http://doi.org/10.1016/j.tig.2008.04.007
Presgraves, D. C. (2010). Darwin and the origin of interspecific genetic incompatibilities. The
American Naturalist, 176 Suppl(December 2010), S45-60.
http://doi.org/10.1086/657058
Presgraves, D. C., & Orr, H. a. (1998). Haldane’s rule in taxa lacking a hemizygous X.
Science (New York, N.Y.), 282(5390), 952–954.
http://doi.org/10.1126/science.282.5390.952
Price, T. D., & Bouvier, M. M. (2002). The evolution of F1 postzygotic incompatibilities in
birds. Evolution; International Journal of Organic Evolution, 56(10), 2083–9.
Pritchard, J. K., Stephens, M., & Donnelly, P. (2000). Inference of population structure using
multilocus genotype data. Genetics, 155(2), 945–959. http://doi.org/10.1111/j.14718286.2007.01758.x
Prowell, D. (1998). Sex linkage and speciation in Lepidoptera. In Endless forms: species and
speciation (pp. 309–319). New York: Oxford University Press.
Prugnolle, F., & de Meeus, T. (2002). Inferring sex-biased dispersal from population genetic
tools: a review. Heredity, 88, 161–165. http://doi.org/10.1038/sj/hdy/6800060
Pryke, S. R. (2010). Sex chromosome linkage of mate preference and color signal maintains
assortative mating between interbreeding finch morphs. Evolution, 64(5), 1301–1310.
http://doi.org/10.1111/j.1558-5646.2009.00897.x
Pryke, S. R., & Griffith, S. C. (2009). Postzygotic genetic incompatibility between sympatric
color morphs. Evolution, 63(3), 793–798. http://doi.org/10.1111/j.15585646.2008.00584.x
Qvarnström, a, & Bailey, R. I. (2009). Speciation through evolution of sex-linked genes.
Heredity, 102(1), 4–15. http://doi.org/10.1038/hdy.2008.93
Raj, A., Stephens, M., & Pritchard, J. K. (2014). fastSTRUCTURE: Variational Inference of
Population Structure in Large SNP Data Sets. Genetics, 197(2), 573–589.
http://doi.org/10.1534/genetics.114.164350
Ramachandran, S., Rosenberg, N. A., Zhivotovsky, L. A., & Feldman, M. W. (2004). On the
robustness of the inference of human population structure. Hum. Genomics, 1(2), 87–97.
Ramsey, J., Bradshaw, H. D., & Schemske, D. W. (2003). Components of reproductive
isolation between the monkeyflowers Mimulus lewisii and M. cardinalis (Phrymaceae).
Evolution; International Journal of Organic Evolution, 57(7), 1520–34.
Ranz, J. M., Namgyal, K., Gibson, G., & Hartl, D. L. (2004). Anomalies in the expression
profile of interspecific hybrids of Drosophila melanogaster and D. simulans. Genome
Res., 14(3), 373–379. http://doi.org/10.1101/gr.2019804
Rao, S. R. V, & Ali., S. (1981). Insect sex chromosomes. VI. A presumptive hyperactivation
of the male X chromosome in Acheta domesticus (L.). Chromosoma, 3, 325–339.
Rao, S. R. V, & Padmaja, M. (1992). Mammalian-type dosage compensation mechanism in
an insect—Gryllotalpa fossor (Scudder)—Orthoptera. Journal of Biosciences, 17(3),
253–273.
242
Rebar, D., Zuk, M., & Bailey, N. W. (2011). Mating experience in field crickets modifies
pre- and postcopulatory female choice in parallel. Behavioral Ecology, 22(2), 303–309.
http://doi.org/10.1093/beheco/arq195
Reinhold, K. (1998). Sex linkage among genes controlling sexually selected traits.
Behavioral Ecology and Sociobiology, 44, 1–7.
Rellstab, C., Gugerli, F., Eckert, A. J., Hancock, A. M., & Holderegger, R. (2015). A
practical guide to environmental association analysis in landscape genomics. Molecular
Ecology, 24(17), 4348–4370. http://doi.org/10.1111/mec.13322
Remington, C. L. (1968). Suture-zones of hybrid interaction between recently joined biotas.
Evolutionary biology. Springer US,.
Rhymer, J. M. ., & Simberloff, D. (1996). Extinction by Hybridization and Introgression.
Annual Review of Ecology and Systematics, 27(1996), 83–109.
Rice, W. R. (1984). Sex Chromosomes and the Evolution of Sexual Dimorphism. Evolution,
38(4), 735–742.
Rieseberg, L. H. (2009). Evolution: Replacing Genes and Traits through Hybridization.
Current Biology, 19(3), 119–122. http://doi.org/10.1016/j.cub.2008.12.009
Rieseberg, L. H., Archer, M. A., & Wayne, R. K. (1999). Short Review Transgressive
segregation , adaptation and speciation. Heredity, 83(July).
Rieseberg, L. H., Widmer, A., Arntz, a M., & Burke, J. M. (2003). The genetic architecture
necessary for transgressive segregation is common in both natural and domesticated
populations. Philosophical Transactions of the Royal Society of London. Series B,
Biological Sciences, 358(1434), 1141–1147. http://doi.org/10.1098/rstb.2003.1283
Ritchie, M. G. (2000). The inheritance of female preference functions in a mate recognition
system. Proceedings of the Royal Society of London. Series B: Biological Sciences,
267(1441), 327–332.
Ritchie, M. G. (2007). Sexual Selection and Speciation. Annual Review of Ecology,
Evolution, and Systematics, 38(1), 79–102.
Ritchie, M. G., Butlin, R. K., & Hewitt, G. M. (1989). Assortative mating across a hybrid
zone in Chorthippus parallelus (Orthoptera: Acrididae). Journal of Evolutionary
Biology, 2(5), 339–352. http://doi.org/10.1046/j.1420-9101.1989.2050339.x
Ritchie, M. G., & Phillips, S. D. F. (1998). The genetics of Sexual isolation.
Rodrigo, E., Marti, D. A., Bidau, C. J., Castillo, E. R., & Bidau, J. (2010). Sex and neo-sex
chromosomes in Orthoptera : a review Sex and neo-sex chromosomes in Orthoptera : a
review. Journal of O, 19(2), 213–231.
Rogers, S. M., & Bernatchez, L. (2007). The genetic architecture of ecological speciation and
the association with signatures of selection in natural lake whitefish (Coregonus sp.
Salmonidae) species pairs. Molecular Biology and Evolution, 24(6), 1423–38.
Rosenthal, G. G. (2013). Individual mating decisions and hybridization. Journal of
Evolutionary Biology, 26(2), 252–5. http://doi.org/10.1111/jeb.12004
Ross, C. L., & Harrison, R. G. (2002). A fine-scale spatial analysis of the mosaic hybrid zone
between Gryllus firmus and Gryllus pennsylvanicus. Evolution; International Journal of
243
Organic Evolution, 56(11), 2296–312.
Rouault, J., Capy, P., & Jallon, J. M. (2001). Variations of male cuticular hydrocarbons with
geoclimatic variables: An adaptative mechanism in Drosophila melanogaster? Genetica,
110(2), 117–130. http://doi.org/10.1023/A:1017987220814
Roux, C., Fraïsse, C., Romiguier, J., Anciaux, Y., Galtier, N., & Bierne, N. (2016). Shedding
light on the grey zone of speciation along a continuum of genomic divergence. PLoS
Biology, 14(12). http://doi.org/10.1007/s11548-012-0737-y
Rowe, C. (1999). Receiver psychology and evolution of multicomponent signals. Animal
Behaviour, 58, 921–931.
Rundle, H. D., & Dyer, K. A. (2015). Reproductive character displacement of female mate
preferences for male cuticular hydrocarbons in Drosophila subquinaria. Evolution, n/an/a. http://doi.org/10.1111/evo.12761
Ryan, M. J., & Cummings, M. E. (2013). Perceptual Biases and Mate Choice. Annual Review
of Ecology, Evolution, and Systematics, 44, 437–459. http://doi.org/10.1146/annurevecolsys-110512-135901
Ryan, M. J., & Rand, A. S. (1995). Female responses to ancestral advertisement calls in
túngara frogs. Science, 269(5222), 390.
Ryan, M. J., & Wagner, Jr., W. E. (1987). Asymmetries in mating preferences between
species - female swordtails prefer heterospecific males. Science, 236, 595–597.
http://doi.org/10.1126/science.236.4801.595
Saenz-Agudelo, P., DiBattista, J. D., Piatek, M. J., Gaither, M. R., Harrison, H. B., Nanninga,
G. B., & Berumen, M. L. (2015). Seascape genetics along environmental gradients in the
Arabian Peninsula: Insights from ddRAD sequencing of anemonefishes. Molecular
Ecology, 24(24), 6241–6255. http://doi.org/10.1111/mec.13471
Sæther, S. A., Sætre, G.-P., Borge, T., Wiley, C., Svedin, N., Andersson, G., … Qvarnstrom,
A. (2007). Sex Chromosome-Linked Species Recognition and Evolution of
Reproductive Isolation in Flycatchers. Science, 318(5847), 95–97.
http://doi.org/10.1126/science.1141506
Sætre, G.-P., Borge, T., Lindroos, K., Haavie, J., Sheldon, B. C., Primmer, C., & Syvänen,
A.-C. (2003). Sex chromosome evolution and speciation in Ficedula flycatchers.
Proceedings. Biological Sciences / The Royal Society, 270(1510), 53–9.
http://doi.org/10.1098/rspb.2002.2204
Safran, R. J., Scordato, E. S. C., Symes, L. B., Rodríguez, R. L., & Mendelson, T. C. (2013).
Contributions of natural and sexual selection to the evolution of premating reproductive
isolation: A research agenda. Trends in Ecology and Evolution, 28(11), 643–650.
http://doi.org/10.1016/j.tree.2013.08.004
Sánchez-Guillén, Anna, R., Wellenreuther, M., & Rivera., A. C. (2012). Strong asymmetry in
the relative strengths of prezygotic and postzygotic barriers between two damselfly
sister species. Evolution, 66, 690–707. http://doi.org/10.5061/dryad.d34t730j
Savarit, F., & Ferveur, J.-F. (2002). Temperature affects the ontogeny of sexually dimorphic
cuticular hydrocarbons in Drosophila melanogaster. The Journal of Experimental
Biology, 205(Pt 20), 3241–9. http://doi.org/10.1017/S0016672301005481
244
Sawamura, K. (1996). Maternal Effect as a Cause of Exceptions for Haldane’s Rule.
Genetics, 143(1), 609. http://doi.org/10.1016/j.clinthera.2013.05.010
Sawamura, K., Taira, T., & K. Watanabe, T. (1993). Hybrid lethal systems in the Drosophila
melanogaster species complex. I. The maternal hybrid rescue (mhr) gene of Drosophila
simulans. Genetics, 305(1 993), 299–305.
Schilthuizen, M., Giesbers, M., & Beukeboom, L. W. (2011). Haldane ’ s rule in the 21st
century. Heredity, (107), 95–102. http://doi.org/10.1038/hdy.2010.170
Schluter, D. (2001). Ecology and the origin of species. Introductory statement. Trends in
Ecology and Evolution, 16(7).
Schluter, D. (2009). Evidence for ecological speciation and its alternative. Science,
323(2008), 737–741. http://doi.org/10.1126/science.1160006
Schmidt, E. M., & Pfennig, K. S. (2015). Hybrid female mate choice as a species isolating
mechanism: environment matters. Journal of Evolutionary Biology, 29, n/a-n/a.
http://doi.org/10.1111/jeb.12818
Schumer, M., Rosenthal, G. G., & Andolfatto, P. (2014). How common is homoploid hybrid
speciation? Evolution, 68(6), 1553–1560. http://doi.org/10.1111/evo.12399
Scordato, E. S. C., Symes, L. B., Mendelson, T. C., & Safran, R. J. (2014). The role of
ecology in speciation by sexual selection: A systematic empirical review. Journal of
Heredity, 105(S1), 782–794. http://doi.org/10.1093/jhered/esu037
Seehausen, O. (2004). Hybridization and adaptive radiation. Trends in Ecology and
Evolution, 19(4), 198–207. http://doi.org/10.1016/j.tree.2004.01.003
Seehausen, O. (2006). African cichlid fish: a model system in adaptive radiation research.
Proceedings of The Royal Society B, 273(May), 1987–1998.
http://doi.org/10.1098/rspb.2006.3539
Seehausen, O., Butlin, R. K., Keller, I., Wagner, C. E., Boughman, J. W., Hohenlohe, P. a, …
Widmer, A. (2014). Genomics and the origin of species. Nature Reviews. Genetics,
15(3), 176–92. http://doi.org/10.1038/nrg3644
Seehausen, O., Terai, Y., Magalhaes, I., Carleton, K., Mrosso, H., Miyagi, R., … Okada, N.
(2008). Speciation through sensory drive in cichlid fish. Nature, 455(2), 620–627.
http://doi.org/10.1038/nature07285
Seehausen, O., Van Alphen, J. J. M., & Witte, F. (1999). Can ancient colour polymorphisms
explain why some cichlid lineages speciate rapidly under disruptive sexual selection?
Belgian Journal of Zoology, 129(1), 43–60.
Seehausen, O., van Alphen, J. J., & Witte, F. (1997). Cichlid Fish Diversity Threatened by
Eutrophication That Curbs Sexual Selection. Science, 277(5333), 1808–1811.
http://doi.org/10.1126/science.277.5333.1808
Selz, O. M., Thommen, R., Maan, M. E., & Seehausen, O. (2014). Behavioural isolation may
facilitate homoploid hybrid speciation in cichlid fish. Journal of Evolutionary Biology,
27(2), 275–289. http://doi.org/10.1111/jeb.12287
Servedio, M. R. (2004a). The evolution of premating isolation: local adaptation and natural
and sexual selection against hybrids. Evolution; International Journal of Organic
245
Evolution, 58(5), 913–24.
Servedio, M. R. (2004b). The what and why of research on reinforcement. PLoS Biology,
2(12), e420. http://doi.org/10.1371/journal.pbio.0020420
Servedio, M. R., & Noor., M. A. (2003). The role of reinforcement in speciation: theory and
data. Annual Review of Ecology, Evolution, and Systematics, 34(1), 339–364.
http://doi.org/10.1146/annurev.ecolsys.34.011802.132412
Servedio, M. R., & Saetre, G.-P. (2003). Speciation as a positive feedback loop between
postzygotic and prezygotic barriers to gene flow. Proceedings. Biological Sciences / The
Royal Society, 270(1523), 1473–9. http://doi.org/10.1098/rspb.2003.2391
Sharma, M. D., Hunt, J., & Hosken, D. J. (2011). Antagonistic responses to natural and
sexual selection and the sex-specific evolution of cuticular hydrocarbons in drosophila
simulans. Evolution, 66(3), 665–677. http://doi.org/10.5061/dryad.227f5pn9
Shaw, K. L. (1996). Polygenic Inheritance of a Behavioral Phenotype : Interspecific Genetics
of Song in the Hawaiian Cricket Genus Laupala. Evolution, 50(1), 256–266.
Shaw, K. L. (2000). Interspecific genetics of mate recognition: inheritance of female acoustic
preference in Hawaiian crickets. Evolution, 54(4), 1303–12.
Shaw, K. L. (2002). Conflict between nuclear and mitochondrial DNA phylogenies of a
recent species radiation: what mtDNA reveals and conceals about modes of speciation in
Hawaiian crickets. Proceedings of the National Academy of Sciences of the United
States of America, 99(25), 16122–7. http://doi.org/10.1073/pnas.242585899
Shaw, K. L., & Danley, P. D. (2003). Behavioral genomics and the study of speciation at a
porous species boundary. Zoology (Jena, Germany), 106(4), 261–73.
http://doi.org/10.1078/0944-2006-00129
Shaw, K. L., Ellison, C. K., Oh, K. P., & Wiley, C. (2011). Pleiotropy, “sexy” traits, and
speciation. Behavioral Ecology, 22(6), 1154–1155.
Shaw, K. L., & Lesnick, S. C. (2009). Genomic linkage of male song and female acoustic
preference QTL underlying a rapid species radiation. Proceedings of the National
Academy of Sciences of the United States of America, 106(24), 9737–42.
http://doi.org/10.1073/pnas.0900229106
Shaw, K. L., & Parsons, Y. M. (2002). Divergence of mate recognition behavior and its
consequences for genetic architectures of speciation. The American Naturalist, 159
Suppl, S61-75.
Shaw DD, Marchant AD, Contreras NE, Arnold ML, Groeters F, K. (1993). Genomic and
environmental determinants of a narrow hybrid zone: cause or coincidence. In Hybrid
zones and the evolutionary process (p. 165–95.).
Shine, R., Reed, R. N., Shetty, S., Lemaster, M., & Mason, R. T. (2002). Reproductive
isolating mechanisms between two sympatric sibling species of sea snakes. Evolution,
56(8), 1655–1662. http://doi.org/Doi 10.1554/00143820(2002)056[1655:Rimbts]2.0.Co;2
Shuker, D. M., Underwood, K., King, T. M., & Butlin, R. K. (2005). Patterns of male sterility
in a grasshopper hybrid zone imply accumulation of hybrid incompatibilities without
selection. Proceedings. Biological Sciences / The Royal Society, 272(1580), 2491–7.
246
http://doi.org/10.1098/rspb.2005.3242
Simmons, L. W. (2004). Genotypic variation in calling song and female preferences of the
field cricket Teleogryllus oceanicus. Animal Behaviour, 68(2), 313–322.
http://doi.org/10.1016/j.anbehav.2003.12.004
Simmons, L. W., & Beveridge, M. (2010a). The strength of postcopulatory sexual selection
within natural populations of field crickets. Behavioral Ecology, 21(6), 1179–1185.
http://doi.org/10.1093/beheco/arq132
Simmons, L. W., & Beveridge, M. (2010b). The strength of postcopulatory sexual selection
within natural populations of field crickets. Behavioral Ecology, 21(6), 1179–1185.
http://doi.org/10.1093/beheco/arq132
Simmons, L. W., & Ritchie, M. G. (1996). Symmetry in the Songs of Crickets. Proceedings
of the Royal Society B: Biological Sciences, 263(1375), 1305–1311.
http://doi.org/10.1098/rspb.1996.0191
Simmons, L. W., Thomas, M. L., Simmons, F. W., & Zuk, M. (2013). Female preferences for
acoustic and olfactory signals during courtship: male crickets send multiple messages.
Behavioral Ecology, 24(5), 1099–1107. http://doi.org/10.1093/beheco/art036
Simmons, L. W., Tinghitella, R. M., & Zuk, M. (2010). Quantitative genetic variation in
courtship song and its covariation with immune function and sperm quality in the field
cricket Teleogryllus oceanicus. Behavioral Ecology, 21(6), 1330–1336.
http://doi.org/10.1093/beheco/arq154
Simmons, L. W., Wernham, J., & Garcı, F. (2003). Variation in paternity in the field cricket
Teleogryllus oceanicus : no detectable influence of sperm numbers or sperm length,
14(4), 539–545.
Simmons, L. W., Zuk, M., & Rothenberry, J. T. (2001). Geographic variation in female
preference functions and male songs of the field cricket Teleogryllus oceanicus.
Evolution, 55(7), 1386–1394.
Singhal, S., & Moritz, C. (2013). Reproductive isolation between phylogeographic lineages
scales with divergence. Proceedings. Biological Sciences / The Royal Society,
280(1772), 20132246. http://doi.org/10.1098/rspb.2013.2246
Smadja, C., & Butlin, R. K. (2009). On the scent of speciation : the chemosensory system and
its role in premating isolation, 77–97. http://doi.org/10.1038/hdy.2008.55
Sousa, V., & Hey, J. (2013). Understanding the origin of species with genome-scale data:
modelling gene flow. Nature Reviews. Genetics, 14(6), 404–14.
http://doi.org/10.1038/nrg3446
Spence, J. R. (1990). Introgressive hybridization in Heteroptera: the example of Limnoporus
Stål (Gerridae) species in western Canada. Canadian Journal of Zoology, 68(8), 1770–
1782.
Stebbins, G. L. (1959). The Role of Hybridization in Evolution. Proceedings of the American
Philosophical Society, 103(2), 231–251.
Stein, A. C. ., & Uy, J. . A. C. . (2006). Unidirectional Introgression of a Sexually Selected
Trait across an Avian Hybrid Zone : A Role for Female Choice ? Evolution;
International Journal of Organic Evolution, 60(7), 1476–1485.
247
Stelkens, R., & Seehausen, O. (2009). Genetic distance between species predicts novel trait
expression in their hybrids. Evolution, 63(4), 884–897. http://doi.org/10.1111/j.15585646.2008.00599.x
Stinchcombe, J. R., & Hoekstra, H. E. (2008). Combining population genomics and
quantitative genetics: finding the genes underlying ecologically important traits.
Heredity, 100(2), 158–70. http://doi.org/10.1038/sj.hdy.6800937
Stratton, G. E. ., & Uetz, G. W. . (1986). The Inheritance of Courtship Behavior and Its Role
as a Reproductive Isolating Mechanism in Two Species of Schizocosa Wolf Spiders (
Araneae ; Lycosidae. Evolution, 40(1), 129–141.
Suzuki, T. A., & Nachman, M. W. (2015). Speciation and reduced hybrid female fertility in
house mice. Evolution, 69(9), 2468–2481. http://doi.org/10.1111/evo.12747
Svensson, E. I., Karlsson, K., Friberg, M., & Eroukhmanoff, F. (2007). Gender differences in
species recognition and the evolution of asymmetric sexual isolation. Current Biology :
CB, 17(22), 1943–7. http://doi.org/10.1016/j.cub.2007.09.038
Svensson, E., Nordén, A., Waller, J., & Runemark, A. (2016). Linking intra- and interspecific
assortative mating : consequences for asymmetric sexual isolation. Evolution, 1–41.
http://doi.org/10.1111/evo.12939.This
Swenson, N. G., & Howard, D. J. (2004). Do Suture Zones Exist? Evolution, 58(11), 2391–
2397. http://doi.org/10.1038/157619d0
Tao, Y., Chen, S., Hartl, D. L., & Laurie, C. C. (2003). Genetic dissection of hybrid
incompatibilities between Drosophila simulans and D. mauritiana. I. Differential
accumulation of hybrid male sterility effects on the X and autosomes. Genetics, 164(4),
1383–1397.
Tao, Y., Hartl, D. L., & Laurie, C. C. (2001). Sex-ratio segregation distortion associated with
reproductive isolation in Drosophila. Proceedings of the National Academy of Sciences
of the United States of America, 98(23), 13183–8.
http://doi.org/10.1073/pnas.231478798
Taylor, E. B., Boughman, J. W., Groenenboom, M., Sniatynski, M., Schluter, D., & Gow, J.
L. (2006). Speciation in reverse: Morphological and genetic evidence of the collapse of
a three-spined stickleback (Gasterosteus aculeatus) species pair. Molecular Ecology,
15(2), 343–355. http://doi.org/10.1111/j.1365-294X.2005.02794.x
Teeter, K. C., Thibodeau, L. M., Gompert, Z., Buerkle, C. A., Nachman, M. W., & Tucker, P.
K. (2010). The variable genomic architecture of isolation between hybridizing species of
house mice. Evolution; International Journal of Organic Evolution, 64(2), 472–85.
http://doi.org/10.1111/j.1558-5646.2009.00846.x
Telschow, A., Yamamura, N., & Werren, J. H. (2005). Bidirectional cytoplasmic
incompatibility and the stable coexistence of two Wolbachia strains in parapatric host
populations. Journal of Theoretical Biology, 235(2), 265–274.
http://doi.org/10.1016/j.jtbi.2005.01.008
The Heliconius Genome Consortium, Dasmahapatra, K. K., Walters, J. R., Briscoe, A. D.,
Davey, J. W., Whibley, A., … Jiggins, C. D. (2012). Butterfly genome reveals
promiscuous exchange of mimicry adaptations among species. Nature, 487(7405), 94–
98. http://doi.org/10.1038/nature11041
248
Thomas, M. L., & Simmons, L. W. (2008a). Cuticular hydrocarbons are heritable in the
cricket Teleogryllus oceanicus. Journal of Evolutionary Biology, 21(3), 801–6.
Thomas, M. L., & Simmons, L. W. (2008b). Sexual dimorphism in cuticular hydrocarbons of
the Australian field cricket Teleogryllus oceanicus (Orthoptera: Gryllidae). Journal of
Insect Physiology, 54(6), 1081–9.
Thomas, M. L., & Simmons, L. W. (2009a). Male-derived cuticular hydrocarbons signal
sperm competition intensity and affect ejaculate expenditure in crickets. Proceedings.
Biological Sciences / The Royal Society, 276(1655), 383–8.
Thomas, M. L., & Simmons, L. W. (2009b). Sexual selection on cuticular hydrocarbons in
the Australian field cricket, Teleogryllus oceanicus. BMC Evolutionary Biology, 9, 162.
Thomas, M. L., & Simmons, L. W. (2010). Cuticular hydrocarbons influence female
attractiveness to males in the Australian field cricket, Teleogryllus oceanicus. Journal of
Evolutionary Biology, 23(4), 707–14.
Thomas, M. L., & Simmons, L. W. (2011). Crickets detect the genetic similarity of mating
partners via cuticular hydrocarbons. Journal of Evolutionary Biology, 24(8), 1793–1800.
http://doi.org/10.1111/j.1420-9101.2011.02319.x
Thomas, M. L., & Simmons, L. W. (2011). Short-term phenotypic plasticity in long-chain
cuticular hydrocarbons. Proceedings. Biological Sciences / The Royal Society,
278(1721), 3123–3128. http://doi.org/10.1098/rspb.2011.0159
Tinghitella, R. M., & Zuk, M. (2009). Asymmetric mating preferences accommodated the
rapid evolutionary loss of a sexual signal. Evolution, 63(8), 2087–2098.
http://doi.org/10.1111/j.1558-5646.2009.00698.x
Tobias, J. a, Planqué, R., Cram, D. L., & Seddon, N. (2014). Species interactions and the
structure of complex communication networks. Proceedings of the National Academy of
Sciences of the United States of America, 111(3), 1020–5.
http://doi.org/10.1073/pnas.1314337111
Tregenza, T., Pritchard, V. L., & Butlin, R. K. (2000). The origins of premating reproductive
isolation : testing hypothesies in the grasshopper Chorthippus. Evolution, 54(5), 1687–
1698.
Tregenza, T., & Wedell, N. (1997). Definitive evidence for cuticular pheromones in a cricket.
Animal Behaviour, 54(4), 979–84.
Turelli, M., Barton, N. H., & Coyne, J. A. (2001). Theory and speciation. Trends in Ecology
& Evolution, 16(7), 330–343.
Turelli, M., & Begun, D. J. (1997). Haldane’s rule and X-chromosome size in Drosophila.
Genetics, 147(4), 1799–1815.
Turelli, M., & Moyle, L. C. (2007). Asymmetric postmating isolation: Darwin’s corollary to
Haldane’s rule. Genetics, 176(2), 1059–88. http://doi.org/10.1534/genetics.106.065979
Turelli, M., & Orr, H. A. (1995). The Dominance Theory of Haldane’s Rule. Genetics,
140(1), 389–402.
Turelli, M., & Orr, H. A. (2000). Dominance, epistasis and the genetics of postzygotic
isolation. Genetics, 154(4), 1663–1679.
249
Tyler, Frances, Xavier A. Harrison, Bretman, A., Veen, T., & Rolando Rodríguez‐Muñoz,
Tregenza, T. (2013). Multiple post-mating barriers to hybridization in field crickets.
Molecular Ecology, 22(6), 1640–9. http://doi.org/10.1111/mec.12187
Veen, T., Faulks, J., Rodríguez-Muñoz, R., & Tregenza, T. (2011). Premating reproductive
barriers between hybridising cricket species differing in their degree of polyandry. PloS
One, 6(5), e19531. http://doi.org/10.1371/journal.pone.0019531
Veen, T., Faulks, J., Tyler, F., Lloyd, J., & Tregenza, T. (2012). Diverse reproductive barriers
in hybridising crickets suggests extensive variation in the evolution and maintenance of
isolation. Evolutionary Ecology, 27(5), 993–1015. http://doi.org/10.1007/s10682-0129610-2
Venables, W. N. & Ripley, B. D. (2002). Modern Applied Statistics with S. Fourth Edition.
Springer, New York.
Via, S. (2001). Sympatric speciation in animals: the ugly duckling grows up. Trends in
Ecology & Evolution, 16(7), 381–390.
Vicoso, B., & Charlesworth, B. (2006). Evolution on the X chromosome: unusual patterns
and processes. Nature Reviews Genetics, 7(8), 645–653. http://doi.org/10.1038/nrg1914
Vines, T. H. (2014). Stuck in the middle with you: close concordance between geographical
clines in a cricket hybrid zone. Molecular Ecology, 23(7), 1647–1649.
http://doi.org/10.1111/mec.12692
Virdee, S. R., & Hewitt, G. M. (1992). Postzygotic isolation and Haldane â€TM s rule in a
grasshopper. Heredity, 69(March), 527–538.
Virdee, S. R., & Hewitt, G. M. (1994). Clines for Hybrid Dysfunction in a Grasshopper
Hybrid Zone. Evolution, 48(2), 392–407.
Waage, J. K. (1975). Reproductive isolation and potential for character displacement in
damselflies, Calopteryx maculata and Calopteryx aequabilis (Odonata-Calopterygidae).
Systematic Zoology, 24, 24–36.
Wade, M. J., Johnson, N. A., & Toquenaga, Y. (1999). Temperature Effects and Genotypeby-Environment Interactions in Hybrids : Haldane â€TM s Rule in Flour Beetles.
Evolution, 53(3), 855–865.
Walther, G. R., Post, E., Convey, P., Menzel, A., Parmesan, C., Beebee, T. J. C., … Bairlein,
F. (2002). Ecological responses to recent climate change. Nature, 416(6879), 389–395.
http://doi.org/10.1038/416389a
Wang, C., Szpiech, Z. A., Degnan, J. H., Jakobsson, M., Pemberton, T. J., Hardy, J. A., …
Rosenberg., N. A. (2010). Comparing Spatial Maps of Human Population-Genetic
Variation Using Procrustes Analysis Comparing Spatial Maps of Human PopulationGenetic Variation Using Procrustes. Statistical Applications in Genetics and Molecular
Biology, 9(1).
Wang, C., Zollner, S., & Rosenberg, N. A. (2012). A Quantitative Comparison of the
Similarity between Genes and Geography in Worldwide Human Populations. PLoS
Genetics, 8(8). http://doi.org/10.1371/journal.pgen.1002886
Watanabe, T. K., & Kawanishi, M. (1979). Mating preference and the direction of evolution
of Drosophila. Science, 205(August), 906–907.
250
Watson, E. T., & Demuth, J. P. (2012). Haldane’s rule in marsupials: What happens when
both sexes are functionally hemizygous? Journal of Heredity, 103(3), 453–458.
http://doi.org/10.1093/jhered/esr154
Weir, B.S. and Cockerham, C. C. (1984). Estimating F-Statistics for the Analysis of
Population Structure. Evolution, 38(6), 1358–1370.
Wells, M., & Henry, C. (1998). Songs, reproductive isolation, and speciation in cryptic
species of insects. In Endless forms: species and speciation (pp. 217–233).
West-Eberhard. (1983). Sexual Selection, Social Competition, and Speciation. Quarterly
Review of Biology, 58(2), 155–183.
Westfall, T. H. and F. B. and P. (2008). Simultaneous Inference in General Parametric
Models. Biometrical Journal, 50(3), 346--363.
Wiley, C., Ellison, C. K., & Shaw, K. L. (2012). Widespread genetic linkage of mating
signals and preferences in the Hawaiian cricket Laupala. Proceedings. Biological
Sciences / The Royal Society, 279(1731), 1203–9. http://doi.org/10.1098/rspb.2011.1740
Wilkins, M. R., Seddon, N., & Safran, R. J. (2012). Evolutionary divergence in acoustic
signals: causes and consequences. Trends in Ecology & Evolution, 1–11.
Wood, T. E., Takebayashi, N., Barker, M. S., Mayrose, I., Greenspoon, P. B., & Rieseberg,
L. H. (2009). The frequency of polyploid speciation in vascular plants. Proceedings of
the National Academy of Sciences, 106(33), 13875–13879.
http://doi.org/10.1073/pnas.0811575106
Woodruff, G. C., Eke, O., Baird, S. E., Félix, M. A., & Haag, E. S. (2010). Insights into
species divergence and the evolution of hermaphroditism from fertile interspecies
hybrids of Caenorhabditis nematodes. Genetics, 186(3), 997–1012.
http://doi.org/10.1534/genetics.110.120550
Wright, S. (1943). Isolation by Distance. Genetics, 28(2), 114–138. http://doi.org/Article
Wright, S. (1946). Isolation by distance under diverse systems of mating. Genetics,
31(January), 39–59.
Wu, C., & Davis, A. W. (1993). Evolution of Postmating Reproductive Isolation : The
Composite Nature of Haldane â€TM s Rule and Its Genetic Bases. The American
Naturalist, 142(2), 187–212.
Xu, Y., Zhu, L., Xiao, J., Huang, N., & McCouch, S. R. (1997). Chromosomal regions
associated with segregation distortion of molecular markers in F2, backcross, doubled
haploid, and recombinant inbred populations in rice (Oryza sativa L.). Molecular and
General Genetics, 253(5), 535–545. http://doi.org/10.1007/s004380050355
Yukilevich, R. (2012). Asymmetrical patterns of speciation uniquely support reinforcement
in Drosophila. Evolution, 66(5), 1430–46. http://doi.org/10.1111/j.15585646.2011.01534.x
Zeileis, a, Kleiber, C., & Jackman, S. (2008). Regression Models for Count Data in R.
Journal of Statistical Software, 27(8), 1–25.
Zheng, X., Levine, D., Shen, J., Gogarten, S. M., Laurie, C., & Weir, B. S. (2012). A highperformance computing toolset for relatedness and principal component analysis of SNP
251
data. Bioinformatics, 28(24), 3326–3328. http://doi.org/10.1093/bioinformatics/bts606
Zirkle, C. (1934). More records of plant hybridization before Koelreuter. Journal of Heredity,
25, 3–18.
Zuk, M., Rebar, D., & Scott, S. P. (2008). Courtship song is more variable than calling song
in the field cricket Teleogryllus oceanicus. Animal Behaviour, 76(3), 1065–1071.
http://doi.org/10.1016/j.anbehav.2008.02.018
Zuk, M., Rotenberry, J. T., & Simmons, L. W. (2001). Geographical variation in calling song
of the field cricket Teleogryllus oceanicus: the importance of spatial scale. Journal of
Evolutionary Biology, 14(5), 731–741. http://doi.org/10.1046/j.1420-9101.2001.00329.x
Zuk, M., Rotenberry, J. T., & Tinghitella, R. M. (2006). Silent night: adaptive disappearance
of a sexual signal in a parasitized population of field crickets. Biology Letters, 2(4),
521–524. http://doi.org/10.1098/rsbl.2006.0539
Zuur, A. F., Saveliev, A. A., & Ieno, E. N. (2012). Zero inflated models and generalized
linear mixed models with R.
252