Content uploaded by Frantisek Marec
Author content
All content in this area was uploaded by Frantisek Marec on Oct 10, 2014
Content may be subject to copyright.
Available via license: CC BY 3.0
Content may be subject to copyright.
Extensive variation in chromosome number and genome
size in sexual and parthenogenetic species of the
jumping-bristletail genus Machilis (Archaeognatha)
Melitta Gassner
1,a
, Thomas Dejaco
1,a
, Peter Sch€
onswetter
2
, Franti
sek Marec
3
, Wolfgang Arthofer
1
,
Birgit C. Schlick-Steiner
1,b
& Florian M. Steiner
1,b
1
Institute of Ecology, University of Innsbruck, Technikerstraße 25, Innsbruck 6020, Austria
2
Institute of Botany, University of Innsbruck, Sternwartestraße 15, Innsbruck 6020, Austria
3
Institute of Entomology, Biology Centre ASCR, Branisovska 31, Cesk
e Budejovice 37005, Czech Republic
Keywords
Asexuality, chromosomal speciation, genome
downsizing, parthenogenesis, polyploidy.
Correspondence
Thomas Dejaco, Institute of Ecology,
University of Innsbruck, Technikerstraße 25,
Innsbruck 6020, Austria. Tel: (+43) 0512 507
51756; Fax: (+43) 0512 507-51799;
E-mail: Thomas.Dejaco@uibk.ac.at
Funding Information
TD was funded by the association for the
support of South-Tyrolean students at
University of Innsbruck and by the South-
Tyrolean science fund (Project-ID: 40.3/
22306/27.01.2014). MG received funding
from the University of Innsbruck. FM
acknowledges the support of the Grant
Agency of the Czech Republic (Project-ID:
14-22765S).
Received: 17 July 2014; Revised: 5
September 2014; Accepted: 10 September
2014
doi: 10.1002/ece3.1264
a
These authors contributed equally to this
work as first authors.
b
These authors contributed equally to this
work as senior authors.
Abstract
Parthenogenesis in animals is often associated with polyploidy and restriction
to extreme habitats or recently deglaciated areas. It has been hypothesized that
benefits conferred by asexual reproduction and polyploidy are essential for col-
onizing these habitats. However, while evolutionary routes to parthenogenesis
are manifold, study systems including polyploids are scarce in arthropods. The
jumping-bristletail genus Machilis (Insecta: Archaeognatha) includes both sexual
and parthenogenetic species, and recently, the occurrence of polyploidy has
been postulated. Here, we applied flow cytometry, karyotyping, and mitochon-
drial DNA sequencing to three sexual and five putatively parthenogenetic East-
ern-Alpine Machilis species to investigate whether (1) parthenogenesis
originated once or multiply and (2) whether parthenogenesis is strictly associ-
ated with polyploidy. The mitochondrial phylogeny revealed that parthenogene-
sis evolved at least five times independently among Eastern-Alpine
representatives of this genus. One parthenogenetic species was exclusively trip-
loid, while a second consisted of both diploid and triploid populations. The
three other parthenogenetic species and all sexual species were diploid. Our
results thus indicate that polyploidy can co-occur with parthenogenesis, but
that it was not mandatory for the emergence of parthenogenesis in Machilis.
Overall, we found a weak negative correlation of monoploid genome size (Cx)
and chromosome base number (x), and this connection is stronger among par-
thenogenetic species alone. Likewise, monoploid genome size decreased with
elevation, and we therefore hypothesize that genome downsizing could have
been crucial for the persistence of alpine Machilis species. Finally, we discuss
the evolutionary consequences of intraspecific chromosomal rearrangements
and the presence of B chromosomes. In doing so, we highlight the potential of
Alpine Machilis species for research on chromosomal and genome-size altera-
tions during speciation.
Introduction
In animals, sexual reproduction prevails and asexuality is
uncommon in most taxonomic groups (Suomaleinen
et al. 1987). It is assumed that asexual organisms have
originated from sexual ancestors (Bell 1982), and this
transition can occur multiple times independently within
a taxonomic group (Stenberg et al. 2003; Schwander and
Crespi 2009; Bode et al. 2010; Elzinga et al. 2013). Apart
from some entirely asexual clades (e.g., bdelloid rotifers
or darwinulid ostracods, the so-called ancient asexuals;
Judson and Normark 1996), strictly asexual animal
ª2014 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd.
This is an open access article under the terms of the Creative Commons Attribution License, which permits use,
distribution and reproduction in any medium, provided the original work is properly cited.
1
species are rare (Bengtsson 2009; Vrijenhoek and Parker
2009). Instead, asexuality often occurs as alternative
reproductive strategy within a sexual taxon. Vandel
(1928) first used the term geographic parthenogenesis to
describe the phenomenon of distinct geographic distribu-
tion of sexual and asexual populations within a species.
Since then, many examples have been reported of parthe-
nogenetic lineages being more common at higher alti-
tudes, higher latitudes, and in anthropogenically
disturbed or extreme habitats (e.g., recently deglaciated
areas), compared with their sexual relatives. Following
this, the advantage of asexuals over sexual congeners for
colonizing new habitats has been attributed to their dou-
bled reproductive potential (as predicted by the twofold
cost of sex (Maynard Smith 1978)) and to highly special-
ized (“frozen niche variation model”) or generalized
(“general purpose model”) genotypes (Vrijenhoek and
Parker 2009). It has been unclear, although, to what
extent asexuality itself (Cuellar 1977; Glesener and Tilman
1978; Law and Crespi 2002; Maniatsi et al. 2011) or fre-
quent correlates like polyploidy (Zhang and Lefcort 1991;
Stenberg et al. 2003; Comai 2005; Adolfsson et al. 2010)
or hybridization (Kearney 2003, 2005; Ghiselli et al. 2007)
promote the spread of organisms into these habitats. The
proposed advantages of polyploidy and hybridization
include protection against deleterious mutations (Comai
2005) and increased heterozygosity (Kearney 2005),
respectively. However, in the latest review of empirical
data (Lundmark and Saura 2006), polyploidy turned out
to be the factor best explaining distributional patterns in
species containing asexual populations.
Parthenogenesis is mostly associated with polyploidy in
plants, and, to a lesser extent, this correlation holds for
animals as well (Suomalainen 1940; Otto and Whitton
2000; Choleva and Janko 2013). The low occurrence of
polyploid lineages in animals has been explained by
dioecy (i.e., male and female gametes come from different
individuals) and chromosomal sex determination (for a
detailed discussion, see Otto and Whitton 2000 and refer-
ences therein), but in fact, polyploidy may have been sim-
ply overlooked in many animal species (Mable 2004).
Polyploidy in animals is thought to emerge more likely in
parthenogenetic diploid populations, while in plants,
polyploidy usually predates asexuality (Otto and Whitton
2000). Already Bell (1982) suggested that in animals,
polyploidy may be fundamental for the persistence of par-
thenogenetic lineages but not essential for their emer-
gence. Despite these theoretical predictions, empirical
studies on the interactions between asexuality and poly-
ploidy and their ultimate effect on the evolution and
spatial distribution of species remain scarce.
To enhance our understanding of the interrelation
between asexuality and polyploidy, it is thus fundamental
to collect empirical data from previously neglected animal
groups with high incidence of parthenogenesis. Jumping
bristletails (Insecta: Archaeognatha) of the genus Machilis
meet these criteria. Within the Eastern Alps, 25 nominal
species are known and at least nine of them putatively
reproduce via parthenogenesis, as only females have been
reported (Wygodzinsky 1941; Janetschek 1954). Moreover,
Dejaco et al. (2012) hypothesized the occurrence of poly-
ploidy in a study including three Machilis species.
In this study, we apply relative genome-size measure-
ments, karyotyping, and mitochondrial DNA sequencing
to sexual and putative parthenogenetic Machilis species to
address the following questions: (1) Did parthenogenesis
arise once or multiple times independently across species?
(2) Is parthenogenesis always coupled with polyploidy?
Materials and Methods
Specimen collection
We focused on eight species of which three reproduce
sexually (M. helleri,M. hrabei, and M. lehnhoferi) and five
putatively via parthenogenesis (M. fuscistylis,M. pallida,
M. engiadina,M. ticinensis, and M. tirolensis). In the par-
thenogenetic species M. ticinensis, males have been
reported from the southern Swiss Alps (Wygodzinsky
1941), potentially indicating a pattern of geographic par-
thenogenesis. In the absence of species-specific multilocus
markers to prove asexual reproduction, we base our defi-
nition of parthenogenetic species on the following obser-
vations: (1) Four species, M. engiadina,M. fuscistylis,
M. pallida, and M. tirolensis, have been collected numer-
ously over the past 60 years but males were never found.
(2) Between 2009 and 2013, we collected over 2000 speci-
mens from these and other parthenogenetic Machilis spe-
cies throughout the Eastern Alps. In doing so, we
considerably widened the known distribution ranges of
these four species (e.g., Rinnhofer et al. 2012; T. Dejaco,
unpubl. data) but again did not find males. (3) In the
sexual species we sampled, males occur at roughly even
ratios and males do not differ from females in their habi-
tat preferences. It is thus unlikely that we overlooked
males in presumably parthenogenetic species. (4) The fifth
parthenogenetic species included in this study, M. ticinen-
sis, was originally described as sexual species from the
southern Swiss Alps (Wygodzinsky 1941). Between 2009
and 2013, we repeatedly sampled female M. ticinensis on
the northern side of the Alps (Austria and Switzerland),
but never encountered male specimens. In contrast, male
and female M. ticinensis were found at an even ratio in
one sampling location in the southern Swiss Alps (MIR).
In total, 209 specimens were sampled at 41 geographi-
cally representative sites throughout the known
2ª2014 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd.
Genomic Variation in Machilis Jumping Bristletails M. Gassner et al.
distribution range of each species (Fig. 1; online supple-
mentary material Table S1). Moreover, 12 specimens of
four additional Machilis species were sampled: M. glacialis,
M. inermis,M. mesolcinensis, and one unknown species,
which morphologically keyed out as M. glacialis, but was
named M. sp. A due to its distant position from M. glacial-
is in the mitochondrial phylogeny. These additional species
were included in our phylogeny but excluded from other
analyses due to poor geographic coverage. One specimen of
Lepismachilis y-signata was used as out-group, resulting in
a total of 222 specimens. Species were determined using the
identification key in Palissa (1964) and original species
descriptions (Riezler 1941; Wygodzinsky 1941; Kratochv
ıl
1945; Janetschek 1949). Sexing of individuals was based on
secondary sexual organs –females are easily recognized by
the presence of an ovipositor, detectable via screening a
specimen’s external morphology. In some cases, specimens
were kept alive in a climate cabinet at 19°C until further
use.
DNA extraction, PCR conditions, and
phylogenetic reconstruction
Genomic DNA was extracted from muscle tissue using
the Mammalian Genomic DNA Miniprep Kit (Sigma-
Aldrich, St. Louis, MO). Two newly designed primers,
MachF5 (50-TAGTTATACCYATYATAATYGGHGG-30)
and MachR7 (50-CCTATRATAGCAAATACTGCYCC-30),
were used to amplify approx. 750 bp of the mitochondrial
cytochrome c oxidase 1 gene (CO1). PCR conditions were
as follows: 95°C for 2 min, 35 cycles (94°C for 30 sec,
50°C for 45 sec, 72°C for 90 sec), and 72°C for 10 min.
Amplicons were checked via gel electrophoresis and puri-
fied using a mastermix containing the enzymes Exo1
(1 U/lL) and FastAP (0.05 U/lL) (Thermo Fisher Scien-
tific Inc., Waltham, MA) and applying an incubation step
at 37°C for 15 min, followed by denaturation of enzymes
at 80°C for 15 min. Sanger sequencing was conducted by
a commercial sequencing facility (Eurofins MWG Operon,
Munich, Germany) using the forward primer MachF5. All
sequences were deposited in GenBank (accession numbers
KJ501697 –KJ501918). Sequences were checked for cor-
rect amino acid translation and aligned using the Clu-
stalW algorithm implemented in MEGA 5 (Tamura et al.
2011), yielding a final alignment of 703 bp. The HKY+G
model of nucleotide substitution was determined as best
fitting our data based on the Akaike information criterion
using jModeltest 2 (Guindon and Gascuel 2003; Darriba
et al. 2012). Net between- and within-species genetic
p-distances were calculated in MEGA 5. The Mann–Whit-
ney U-test was used to test whether within-species p-dis-
tances differed significantly between sexual and
parthenogenetic species.
A Bayesian inference tree was constructed with MrBa-
yes 3.2 (Ronquist et al. 2012). Three partitions were spec-
ified corresponding to codon positions. Two parallel runs,
each consisting of three heated and one cold chain, were
run for 10
6
generations, sampling trees every 1000 genera-
tions. Convergence was checked using the average stan-
dard deviation of splits frequencies, which fell below 0.01
after approx. 8.5 910
5
generations. The first 500 trees
were discarded as burn-in.
Figure 1. Geographic locations of all
populations sampled in this study. White and
black symbols correspond to sexual and
parthenogenetic species, respectively.
Whenever more than one species was found
at a location, species-specific symbols are
connected by lines with that location.
ª2014 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd. 3
M. Gassner et al.Genomic Variation in Machilis Jumping Bristletails
Karyotyping
With minor adaptations, we followed the protocol of
Sahara et al. (1999) for chromosome preparations. Speci-
mens were anesthetized with carbon dioxide prior to dis-
section, and three legs were snap-frozen in liquid
nitrogen and stored at 70°C for use in flow cytometry.
Specimens were dissected in physiological solution con-
taining 154 mmol/L NaCl, 5 mmol/L KCl, 1.8 mmol/L
CaCl
2
, and 2.9 mmol/L NaHCO
3
. Gonads were removed
and placed into hypotonic solution (0.075 mol/L KCl) for
swelling (15–20 min), followed by fixation in Carnoy’s
solution (ethanol: chloroform: acetic acid =6:3:1) for 20
to 30 min. Immediately after the removal of gonads, spec-
imens were preserved in 99% ethanol and stored at
20°C for further use. Glass slides were incubated in acid
ethanol (1% HCl: 96% ethanol =1:100) for at least half
an hour before use. Gonads were dispersed on dry-
cleaned slides in a drop of acetic acid (60%) with tung-
sten needles and spread using a heating plate (40°C).
Chromosome preparations were first checked for meta-
phase spreads under a Nikon Eclipse E600 phase contrast
microscope and then stained with 0.02 lg/mL DAPI in an-
tifade solution based on DABCO (1,4-diazabicyclo[2.2.2]
octane; Sigma-Aldrich). Pictures of chromosome spreads
were taken with a Leica DFC 495 digital camera (Leica
Microsystems, Wetzlar, Germany) mounted on a Leica
DM 5000B fluorescence microscope. Single chromosomes
were arranged according to their size using Adobe Creative
Suite 4 software (Adobe Systems, San Jose, CA). Chromo-
some number was counted in at least three metaphase
spreads per individual. Chromosome base number (x; i.e.,
the haploid number, which is occurring twice in a diploid
and three times in a triploid) was calculated by dividing
diploid chromosome number by two and triploid chromo-
some numbers by 3. Fundamental number of arms (FN;
i.e., the total number of chromosome arms in a 2n meta-
phase complement) was determined for all species from
representative karyotypes of as many sampling locations as
possible.
In M. ticinensis, chromosome lengths (in lm) were
measured in each one karyotype of nine asexual and five
sexual individuals using the software MicroMeasure ver-
sion 3.3 (Reeves 2000). We calculated mean chromosome
length (MChL, equal to the sum of chromosome lengths
divided by the number of chromosomes within one
karyotype) and tested for significant difference in MChL
among sexual and asexual individuals. Statistical tests
were calculated in SigmaPlot v.12.5 (Systat Software Inc.,
San Jose, CA), applying a significance level of a=0.05.
When normality and equal variances were given, we
applied parametric tests. Whenever one of these criteria
failed, we used nonparametric tests instead (Reeves 2000).
Flow cytometry measurements
We applied flow cytometry to obtain relative genome-size
estimates from leg muscle tissue using Bellis perennis
(Asteraceae; 2C =3.38 pg, Sch€
onswetter et al. 2007) as
internal standard (i.e., measured together with each sam-
ple). Samples were treated following Suda et al. (2007). In
detail, single legs and approx. 0.5 cm
2
of a B. perennis leaf
were chopped with 500 lL of ice cold Otto 1 Buffer
(0.1 mol/L citric acid, 0.5% Tween20 (Merck KGaA,
Darmstadt, Germany)) in a small Petri dish. The suspen-
sion was then filtered through a 42-lm nylon mesh and
incubated for a few minutes. Finally, 1 mL of staining
solution (DAPI (4 lL/mL) and 2-mercaptoethanol (2 lL/
mL) in Otto 2 buffer (0.4 mol/L Na
2
HPO
4
∙12 H
2
O)) was
added. Fluorescence intensity of 3000 cell nuclei per sam-
ple was measured using a Partec CyFlow
space flow
cytometer (Partec GmbH, M€
unster, Germany). Gating and
peak analysis were performed automatically using the Par-
tec Flo Max software. In a few samples that were defrosted
too fast, fluorescence peaks showed increased noise on the
left side. In these cases, left gates were adjusted manually
to prevent overestimation of peak width.
Genome size measurements based on DAPI fluores-
cence have been criticized because of DAPI’s AT-specific
DNA-binding properties. Potential differences in genomic
AT/GC content among species may introduce some bias
to fluorescence intensity and thus genome size estimates.
However, Johnston et al. (1999) showed that genome size
estimates based on DAPI largely agree with estimates
based on other fluorochromes. As we used a plant (B. pe-
rennis) as internal standard, we report relative genome
size (i.e., fluorescence intensity of sample divided by fluo-
rescence intensity of internal standard), instead of calcu-
lating nuclear DNA content in picograms. Relative
genome size values were divided by ploidy level to obtain
monoploid genome size (Cx), which is comparable
among different ploidy levels. Fig. 4 was drawn in Sigma-
plot v.12.5. Fig. 5 was generated using the R package
ggplot2 (http://cran.r-project.org/web/packages/ggplot2/
index.html).
Results
For each species, an overview of sampling locations is
given in Table 1, including number of female and male
individuals sampled, mean chromosome number and rel-
ative genome size, ploidy level, elevation, and FN.
Mitochondrial phylogeny
Our phylogeny included eight target species, as well as four
additional species mentioned in the specimen collection
4ª2014 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd.
Genomic Variation in Machilis Jumping Bristletails M. Gassner et al.
Table 1. Summary of average chromosome number and genome size (mean SD) values for all Machilis populations sampled in this study,
including additional information on reproductive mode, number of individuals per population, and sex.
Species name
Repr.
mode
Sampling
location
Number of
specimens
Number of
chromosomes
FN
Mean relative genome
size (SD) Ploidy
levelF M F (2n) M (n) F M
Machilis helleri S ADM (AT) 2 2 52 26 124 1.49 0.01 1.46 0.02 2n
S EIS (AT) 4 1 52 26 124 1.56 0.3 1.45 2n
S GIE (AT) 5 4 n/a 25, 51
1
n/a 1.60 0.01 1.60 0.01 2n
S HWD (AT) 8 3 50 25 124 1.60 0.02 1.57 0.01 2n
S RAX (AT) 3 1 n/a 25 n/a 1.55 0.04 1.49 2n
S UST (CZ) 4 0 52 n/a n/a 1.45 0.02 n/a 2n
Machilis hrabei S BRN (CZ) 4 6 52 26 140 1.90 0.02 1.89 0.02 2n
S KNB (AT) 3 5 52 26 140 1.83 0.01 1.77 0.02 2n
S VIE (AT) 4 2 52 26 140 1.86 0.02 1.85 0.03 2n
Machilis lehnhoferi S GLK (AT) 4 1 52 26, 52
1
128 1.73 0.3 1.77 2n
S HAI (AT) 3 0 52 n/a 128 5.82 0.03 n/a 2n
S HUN (AT) 1 1 52 26, 52
1
n/a 1.7 1.77 2n
S OBE (DE) 3 2 52 26 n/a 1.76 0.02 1.76 0.08 2n
S OBL (AT) 4 0 52 n/a 128 1.71 0.05 n/a 2n
S SAA (AT) 0 3 n/a 26 n/a n/a 1.77 0.01 2n
S SAL (AT) 0 4 n/a 26 n/a n/a 1.76 0.02 2n
S STZ (AT) 3 0 52 n/a n/a 1.72 0.02 n/a 2n
Machilis distincta P EIS (AT) 10 50 140 1.73 0.03 2n
P KRK (AT) 1 50 n/a 1.73 2n
P LDC (IT) 1 50 n/a 1.74 2n
P MUR (AT) 9 50 n/a 1.73 0.01 2n
P NIK (AT) 8 50 n/a 1.72 0.02 2n
P SAA (AT) 2 50 n/a 1.74 0.00 2n
P SLD (AT) 4 50 140 1.74 0.02 2n
P SMT (IT) 4 50 140 1.72 0.02 2n
P STA (IT) 2 50 n/a 1.73 0.03 2n
P TOB (IT) 1 50 n/a 1.73 2n
Machilis fuscistylis P FOT (AT) 3 54 148 1.46 0.02 2n
P HIN (AT) 6 56 156 1.50 0.01 2n
P MAR (IT) 1 54 148 1.47 2n
P OBG (AT) 4 54 148 1.47 0.03 2n
P SRK (AT) 1 54 n/a 1.46 2n
Machilis pallida P MAD (IT) 8 78 n/a 2.30 0.03 3x
P SEI (IT) 4 78 186 2.29 0.02 3x
P TRI (AT) 5 78 n/a 2.31 0.03 3x
Machilis tirolensis P EBK (AT) 4 75 198 2.54 0.03 3x
P KRK (AT) 2 75 n/a 2.54 0.04 3x
P LDC (IT) 6 50 132 1.61 0.02 2n
P LDL (IT) 3 50 132 1.62 0.02 2n
P RAU (AT) 1 75 n/a 2.54 3x
P SAR (IT) 3 75 n/a 2.54 0.02 3x
P STA (IT) 3 50 n/a 1.60 0.01 2n
Machilis ticinensis P BRA (AT) 2 46 128 1.67 0.06 2n
P KRK (AT) 1 46 n/a 1.65 2n
P NEN (AT) 7 46 128 1.69 0.01 2n
P RAN (AT) 1 46 n/a 1.67 2n
P RAU (AT) 2 46 128 1.68 0.02 2n
S MIR (CH) 8 2 46, 46 +Bs 23 128 2.01 0.06 1.95 0.01 2n
Total 41 172 37
FN, fundamental number of arms; F, females; M, males; S, sexual; P, parthenogenetic; n/a, not available (or not applicable); AT, Austria; CZ,
Czech Republic; DE, Germany; IT, Italy; CH, Switzerland; Bs, B chromosomes.
1
2n chromosome numbers are additionally given in males when mitotic chromosomes were found.
ª2014 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd. 5
M. Gassner et al.Genomic Variation in Machilis Jumping Bristletails
section. All species formed monophyletic clusters sup-
ported by posterior probabilities above 0.95 (Fig. 2). In
contrast, some deeper nodes were only weakly supported.
The three sexual species formed a monophyletic group,
while the parthenogenetic species did not. Mean net p-dis-
tances between species ranged between 3.00% and 16.22%,
while p-distances within-species were equal or lower than
1.00% (Table 2). Mean within-group distances were signifi-
cantly higher in sexual than in parthenogenetic species
(Student’s t-test: t=3.369; df =6; P=0.015).
Intraspecific splits occurred in M. helleri,M. hrabei,
M. ticinensis, and M. tirolensis (Fig. 2). In M. ticinensis and
M. tirolensis, these splits were congruent with reproductive
mode and ploidy level, respectively. In M. helleri, the two
resulting clades were partly congruent with differing chro-
mosome number (Fig. 2 and Fig. 3). In M. hrabei, the two
clades corresponded to sampling locations from Austria
(KNB, VIE) and the Czech Republic (BRN; see Table 1).
Chromosome numbers
We found predominantly mitotic chromosomes in female
gonads and meiotic chromosomes in male gonads, except
in two male specimens (each one M. helleri and
M. lehnhoferi), where both meiotic and mitotic chromo-
somes were found. In most sampling locations of the
three sexual species (M. helleri,M. hrabei, and M. lehnho-
feri), we consistently counted 2n =52 chromosomes in
females and n=26 bivalents in males. The assembled
karyotypes of these species revealed that they differed in
the numbers of metacentric, submetacentric, and acrocen-
tric chromosomes, and hence also in the fundamental
number of arms (FN; Table 1, Fig. S1). In three locations
of M. helleri, however, we found only 2n =50 in females
and n=25 bivalents in males (HWD and RAX), and
2n =51 chromosomes and n=25 bivalents in one male
from GIE (Fig. 3; Table 1). Unfortunately, no unambigu-
ously countable chromosome spread could be produced
in female M. helleri specimens from GIE and RAX. All
M. helleri specimens with 52 chromosomes fell into one
well-supported mitochondrial subclade, including some
specimens from locations RAX (all) and HWD (six indi-
viduals), which had only 50 chromosomes. The remaining
individuals from HWD (five individuals) and GIE corre-
sponded to the basal subclade within M. helleri (Fig. 3),
having 50 chromosomes.
M. fuscistylis
M. pallida
M. lehnhoferi
M. hrabei
M. helleri
M. tirolensis
M. ticinensis
M. engiadina
M. sp. A
M. inermis
Lepismachilis y-signata
relative genome size
N = 17
M. glacialis
M. mesolcinensis
N = 22
N = 14
N = 10
N = 16
N = 13
N = 12
N = 10
N = 10
N = 13
N = 42
N = 9
N = 6
0.1
*
*
*
*
*
*
*
*
*
*
*
*
*
**
*
*
*
*
*
*
*
*
*
*
*
*
46
46+B
50
51
52
54
56
75
78
Number of
chromosomes
0 0.5 1 1.5 2 2.5 3
N = 8
N = 7
Figure 2. Bayesian phylogeny based on 222
mitochondrial CO1 sequences from 12 Eastern-
Alpine Machilis species and Lepismachilis y-
signata as out-group. All nodes supported by
posterior probabilities higher than 0.95 are
indicated by stars. Red branches highlight
parthenogenetic species. Species included in
the phylogeny but excluded from other
analyses have gray branches. Mean relative
genome size and chromosome number are
given for all species and major intraspecific
clades. Diploid and triploid chromosome
numbers are indicated by circled and squared
symbols, respectively. The number of
individuals used to calculate average genome
size values (N) is given to the right side of bars.
Error bars indicate standard deviation (SD).
6ª2014 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd.
Genomic Variation in Machilis Jumping Bristletails M. Gassner et al.
In asexual species, chromosome number ranged from
2n =46 to 2n =3x =78 (Fig. 2, Table 1). In M. fuscisty-
lis, we detected intraspecific variation with 2n =54
chromosomes across four locations (OBG, FOT, MAR,
and SRK) but consistently 2n =56 chromosomes in speci-
mens from HIN (Fig. 2 and Fig. S1). In all M. pallida
specimens, we counted 2n =3x =78 chromosomes. In
M. tirolensis, most sampling locations harbored exclusively
diploid specimens (2n =50 chromosomes), but at two
locations (LDL and LDC), only triploids with 2n =
3x =75 chromosomes were found (Fig. S1). We consis-
tently found 2n =50 chromosomes in specimens of
M. engiadina.
In M. ticinensis, both males and females were found at
location MIR, while exclusively females were found at the
other five sampling locations (see Table 1), possibly indi-
cating a pattern of geographic parthenogenesis. In the
sexual population, specimens had 2n =46 chromosomes
plus a varying number of supernumerary elements, that
is, B chromosomes, representing 2n =46 (N=3),
2n =46 +1B (N=1), 2n =46 +4B (N=3), and
2n =46 +5B (N=1) in females, and n=23 (N=2)
bivalents in male meiotic spreads (online supplementary
material Fig. S1 and Tab. S1). In contrast, parthenoge-
netic specimens consistently had 2n =46 chromosomes
and no B chromosomes. In specimens from the sexual
population, mean chromosome size was significantly
larger compared with parthenogenetic specimens (Mann–
Whitney U-test; U =0, P=0.003, N
sex
=5, x
sex
=4.26
0.38 lm, N
parth
=9, x
parth
=3.33 0.12 lm).
Table 2. Genetic P-distances (in %) within (SD) and among species. Rows 1–3 are sexual species, rows 4–8 (light gray background) are parthe-
nogenetic species, and rows 9–12 (dark gray background) are species that were included in the phylogeny but not in other analyses due to low
sample size.
S. no Species
Mean within
group distance
Pairwise mean net between distance\standard deviation
123456789101112
1Machilis helleri 0.9 (0.2) 1.03 1.04 1.37 1.45 1.45 1.36 1.32 1.61 1.18 1.35 1.29
2Machilis hrabei 1.0 (0.3) 7.61 0.57 1.51 1.49 1.48 1.44 1.48 1.49 1.37 1.53 1.36
3Machilis lehnhoferi 0.6 (0.2) 7.38 3.04 1.48 1.43 1.55 1.35 1.47 1.63 1.34 1.53 1.44
4Machilis ticinensis 0.5 (0.2) 12.48 14.56 13.78 1.49 1.50 1.59 0.89 1.53 1.47 1.13 1.51
5Machilis distincta 0.1 (0.0) 13.55 13.97 13.46 14.52 1.52 1.33 1.49 1.40 1.51 1.48 1.66
6Machilis fuscistylis 0.0 (0.0) 13.51 14.31 14.81 13.99 14.60 1.55 1.55 1.57 1.48 1.45 1.51
7Machilis pallida 0.1 (0.1) 11.96 12.99 12.06 15.93 12.75 15.00 1.55 1.59 1.42 1.53 1.49
8Machilis tirolensis 0.4 (0.2) 12.49 14.44 14.17 5.95 14.77 15.10 15.20 1.52 1.41 1.16 1.48
9Machilis glacialis 0.0 (0.0) 15.67 14.91 15.93 14.38 13.60 15.22 15.40 13.80 1.49 1.59 1.62
10 Machilis inermis 0.5 (0.2) 10.15 12.75 12.25 13.54 14.20 14.08 12.65 12.69 14.27 1.40 1.32
11 Machilis mesolcinensis 0.3 (0.2) 12.23 14.59 14.53 9.07 14.39 13.70 14.63 9.37 14.94 13.37 1.50
12 Machilis spA 0.0 (0.0) 11.78 12.14 12.56 14.68 16.15 14.65 13.98 14.51 16.22 11.57 15.08
50
51
52
Number of
chromosomes
UST
1.45 ± 0.01
ADM EIS
RAX
HWD
*
*
*
POL
CZE
AUT
0
40
300
260
340
0 40 80 120 160
km
0.05
GER
CZE
ITA
SUI
AUT
SLO
N
1.48 ± 0.02 1.54 ± 0.06
1.53 ± 0.04
GIE
1.60 ± 0.01
1.59 ± 0.02
Figure 3. Geographic locations of sampled
Machilis helleri populations with corresponding
number of chromosomes and mean relative
genome size (SD). Below, a simplified
drawing of the corresponding branch of the
mitochondrial phylogeny is displayed. All nodes
supported by posterior probabilities higher
than 0.95 are indicated by stars.
ª2014 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd. 7
M. Gassner et al.Genomic Variation in Machilis Jumping Bristletails
Chromosome base numbers (x) significantly increased
with elevation across all species (Pearson’s r:r=0.503,
P<0.001; Fig. 4A). However, sexual species did not con-
tribute to this trend as no correlation was found among
populations of sexual species alone (Pearson’s r:
r=0.024, P=0.919; Fig. 4A). In contrast, when only
populations from parthenogenetic species were consid-
ered, the correlation was strongest (Pearson’s r: r=0.785,
P<0.001; Fig. 4A). No such correlation was found
within any of the single species (data not shown).
Relative genome size measurements
The relative genome sizes (2C) ranged from 1.43 (M. hel-
leri) to 2.12 (M. ticinensis) in diploids and from 2.26
(M. pallida) to 2.57 (M. tirolensis) in triploids (Fig. 2 and
Fig. 5). This approximately corresponds to C values rang-
ing from 2.21 pg to 3.59 pg in diploids and 3.82 pg to
4.35 pg in triploids. Even when taking into account
potential bias introduced by DAPI staining and using a
plant as an internal standard, these values are higher than
mean C values reported for most insect orders (except
Orthopthera), and a similarly high C value has been
reported for the closely related insect order Zygentoma
(Thermobia domestica:C=3.09 pg; Gregory 2014).
Among sexual species, relative genome size (2C) varied
significantly in overall (Kruskal–Wallis ANOVA:
H=66.388, df =2, P<0.001) and multiple pairwise
comparisons (Dunn’s test: M. hrabei vs. helleri:
Q=8.086, P<0.05; M. hrabei vs. lehnhoferi:Q=3.466,
P<0.05; M. lehnhoferi vs. helleri:Q=4.629, P<0.05).
In M. hrabei, 2C values varied significantly among the
two mitochondrial subclades (Student’s t-test: t=5.33,
df =22, P<0.001), with specimens from the Czech
Republic (BRN) having larger genomes than specimens
from Austria. Intraspecific variation (as measured by stan-
dard deviations of relative genome size within each spe-
cies) was significantly higher in sexual than in
parthenogenetic species when M. ticinensis was excluded
(Student’s t-test: t=2.448, df =5, P
one-tailed
=0.029;
see Fig. 5).
In the parthenogenetic species M. fuscistylis, specimens
with 56 chromosomes had significantly larger relative gen-
ome sizes than specimens with 54 chromosomes (Stu-
dent’s t-test: t=3.919, df =11, P=0.002). In triploid
individuals of M. tirolensis, relative genome size was 1.58-
fold compared with conspecific diploids, thus corroborat-
ing their triploid status. In the potentially geographic
parthenogenetic species M. ticinensis, relative genome size
was significantly larger in the sexual population compared
with parthenogenetic specimens (Mann–Whitney U-test:
U=0, N
parth.
=13, N
sex
=10, P<0.001). Within the
sexual M. ticinensis population, relative genome size
increased significantly with the number of B chromo-
somes (Pearson’s r:r=0.939, P<0.001). Nevertheless,
relative genome size of sexual M. ticinensis specimens
without B chromosomes (2n =46) was significantly larger
than relative genome size in parthenogenetic specimens
(Student’s t-test: t=25.211, df =16, P<0.001; Fig. 5).
Monoploid genome sizes (Cx) decreased significantly
with increasing elevation across all species (Pearson’s r:
r=0.460, P<0.001; Fig. 4B). Sexual species did not
contribute to this trend, as no correlation was found
among populations from sexual species alone (Pearson’s
r = –0.369; P = 0.008
r = –0.360; P = 0.109
r = –0.678; P < 0.001
r = –0.460; P < 0.001
r = 0.126; P = 0.588
r = –0.893; P < 0.001
r = 0.503; P < 0.001
r = –0.024; P = 0.919
r = 0.785; P < 0.001
Monoploid genome size (Cx) Monoploid genome size (Cx) Chromosome base number (x)
1.1
1.0
0.9
0.8
0.7
1.1
1.0
0.9
0.8
0.7
22 23 24 25 26 27 28 29
0 500 1000 1500 2000 2500 3000 3500
0 500 1000 1500 2000 2500 3000 3500
29
28
27
26
25
24
23
22
Chromosome base number (x)
Elevation (m a.s.l.)
Elevation (m a.s.l.)
(A)
(B)
(C)
Figure 4. Correlation between elevation and chromosome base
number (A), elevation and monoploid genome size (B), and
chromosome base number and monoploid genome size (C). Sexual
populations are represented by black dots, parthenogenetic
populations by gray circles. Regression lines and correlation
coefficients including P-values are given for all populations together
(black line), just sexual populations (black dotted line), and just
parthenogenetic populations (gray dashed line).
8ª2014 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd.
Genomic Variation in Machilis Jumping Bristletails M. Gassner et al.
r: r=0.126, P=0.588; Fig. 4B). Among parthenogenetic
species, however, the negative correlation of Cx and eleva-
tion was strongest (Pearson’s r: r=0.893, P<0.001;
Fig. 4B). Similarly, a considerable negative correlation of
Cx and chromosome base number (x) was found among
parthenogenetic (Pearson’s r: r=0.678, P<0.001;
Fig. 4C) but not among sexual species (Pearson’s r:
r=0.360, P=0.109; Fig. 4C). When populations from
all species were considered, this correlation was very weak
(Pearson’s r: r=0.369, P<0.001; Fig. 4C).
Discussion
In this study, we investigated the link between parthe-
nogenesis and polyploidy in Alpine representatives of
the jumping-bristletail genus Machilis by generating rel-
ative genome-size estimates and karyotypes from geo-
graphically representative samples of three sexual and
five presumably parthenogenetic species. By mapping
relative genome size and chromosome numbers onto a
mitochondrial phylogeny, we tackled the questions
raised in the introduction and identified promising tra-
jectories for future research.
Multiple origin of parthenogenesis
As this is the first phylogenetic framework in the genus
Machilis, no ‘a priori’ hypotheses concerning species rela-
tionships and evolution of parthenogenesis were available.
Assuming that a reversal from parthenogenetic to sexual
reproduction is unlikely following Dollo’s law (Gould
1970), the nonmonophyly of parthenogenetic species indi-
cated that asexuality originated at least five times indepen-
dently (see Fig. 2). Multiple evolution of parthenogenesis
within a genus has been demonstrated in other animal
groups such as stick insects (Schwander and Crespi 2009)
and lizards (Manriquez-Moran et al. 2014), and these
groups offer exciting opportunities for studying evolution-
ary scenarios leading to parthenogenesis.
In Alpine Machilis species, multiple origins of parthe-
nogenesis may have been facilitated by Pleistocene glacia-
tion cycles, which are thought to have increased the
incidence of asexual reproduction in plants and animals
(H€orandl 2009). In fact, persistence on so-called nunataks
(i.e., rocky peaks towering above the ice shield) has been
suggested in two alpine Machilis species (M. fuscistylis:
Janetschek 1956; M. pallida: Wachter et al. 2012) and may
be linked to the onset of parthenogenesis in these species.
In contrast, inner-Alpine persistence of glaciations at lower
elevations is unlikely. In species occurring at lower eleva-
tions (e.g., M. engiadina,M. ticinensis, and M. tirolensis),
parthenogenesis is thus more likely to have evolved during
recolonization of the inner-Alpine area following persis-
tence in peripheral refugia. However, in the absence of a
robust time calibration, dating speciation events and
origins of parthenogenesis remain elusive. In line with the
concept of geographic parthenogenesis (Vandel 1928;
H€
orandl 2009), parthenogenetic species included in our
study mainly occurred in central parts of the Alps, whereas
sexual species were mostly scattered close to the margin of
the Alps and in surrounding lowland areas.
Incidence of polyploidy
Polyploidy was found in two parthenogenetic species. In
M. tirolensis, populations with both, 50 and 75 chromo-
somes were identified. Because relative genome size in the
latter was on average 1.58 times higher than in the former,
the latter’s triploid status is strongly supported. The second
case of triploidy is hypothesized in M. pallida (2n =
3x =78) based on the 1.5 ratio in chromosome number
compared with sexual species. However, the direct sexual
ancestor of M. pallida is unknown and possibly extinct,
and thus, accurate inference of a phylogenetically unbiased
genome size ratio is not possible. Still, a ratio of 1.34 was
calculated using the average relative genome size across
sexual species included in this study. Deviation from the
1.6
2.0
2.4
2.2
1.8
1.4
2.6
M. helleri
M. hrabei
M. lehnhoferi
M. ticinensis
M. fuscistylis
M. engiadina
M. tirolensis
M. pallida
M. ticinensis
M. tirolensis
Relative genome size (2C)
Figure 5. Relative genome size plotted for 191 Machilis individuals
for which chromosome number has been determined as well.
Individual measurements are grouped according to species affiliation
and, in M. ticinensis and M. tirolensis, split according to reproductive
mode (sexual/parthenogenetic) and ploidy level (diploid/triploid),
respectively. Dark and light gray backgrounds indicate sexual and
parthenogenetic mode of reproduction, respectively. Diploid
individuals are represented by filled circles and triploids by empty
squares. Chromosome number is coded using the same colors as in
Fig. 2. Within groups, random horizontal scatter was applied to
individuals for better visualization.
ª2014 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd. 9
M. Gassner et al.Genomic Variation in Machilis Jumping Bristletails
expected 1.5 ratio might be explained by phylogenetic dis-
tance between M. pallida and these sexual taxa.
Since the other three parthenogenetic species were dip-
loid, polyploidy is not always coupled with parthenogene-
sis in Machilis. Similarly, parthenogenetic diploids have
been found in numerous other arthropod taxa, for exam-
ple, brine shrimps (Maccari et al. 2013), weevils (Stenberg
et al. 2003), and stick insects (Schwander and Crespi
2009). Our results are thus in line with the hypothesis,
that in animals, polyploidy emerges secondarily, after the
establishment of diploid parthenogenetic populations
(Suomaleinen et al. 1987).
Opportunities for studying genome
evolution in the genus Machilis
In this study, we found substantial variation in chromo-
some number and relative genome size among and within
Machilis species. In the following, we highlight three
peculiarities in our data that emphasize the potential of
the genus Machilis as a study system for investigating the
role of chromosomal rearrangements and genome size
alterations in evolutionary diversification.
Genome downsizing
Monoploid genome size (Cx) was negatively correlated
with chromosome base number (x) among species
included in this study, and this correlation was strongest
among parthenogenetic species. Moreover, while Cx
decreased with elevation, chromosome base number
tended to increase with elevation, and this effect was also
stronger among parthenogenetic species. We cautiously
interpret this as a signal of genome downsizing along an
altitudinal gradient that is possibly stronger in partheno-
genetic than in sexual species. Downsizing of nuclear
DNA content is a known phenomenon in polyploid
plants, where higher ploidy levels tend to have smaller C
values than diploids of a given species (Leitch and Ben-
nett 2004). Here, in contrast, we hypothesize that genome
size and chromosome number might be shaped by envi-
ronmental conditions and that parthenogenetic Machilis
species might be more prone to this process than sexual
congeners.
To date, specific mechanisms involved in genome size
reductions are largely unclear. However, Hessen et al.
(2010) proposed that reductions in genome size in
eukaryotes may be explained by the need for streamlined
genomes in environments selecting for rapid growth due
to low food availability. As in alpine environments (i.e.,
above 2000 m a.s.l.) activity periods are short and food
availability is low, we suggest that harsh climatic condi-
tions selected for smaller genomes in alpine Machilis
species. Alternatively, our interpretation might be misled
by the fact that no exclusively alpine sexual species are
included in our study. One reason for this is the unclear
taxonomic situation in the genus Machilis (Dejaco et al.
2012), with several species being poorly delimited and
thus not available for evolutionary research. Also, most
putative Machilis species found in alpine habitats are pre-
sumably parthenogenetic, and one could expect that asex-
ual reproduction itself is a crucial property enabling
persistence in this habitat. Our results uncover a potential
alternative hypothesis by pointing at smaller genomes
(and more and thus smaller chromosomes) in alpine spe-
cies. Future work should aim at validating this correlation
in sexual alpine Machilis species to disentangle the relative
contribution of genome size reduction versus asexual
reproduction to the persistence in alpine environments.
Intraspecific chromosomal rearrangements
We found intraspecific variation in chromosome number
within two species. In the parthenogenetic M. fuscistylis,
one population (HIN) harbored two additional chromo-
somes compared with the other populations. As the fun-
damental number of arms differed between these
subgroups, we excluded the possibility of chromosome
fusion or fission, as both types of rearrangement would
not alter FN. This assumption is further supported by rel-
ative genome size measurements, which confirmed a sig-
nificantly higher 2C value in the HIN population. We
propose that the duplication of one chromosome pair in
HIN (Fig. S1) is more parsimonious than the parallel loss
of one chromosome pair in the other populations. Also,
as both karyotypic groups shared the same CO1 haplo-
type, the novel karyotype must have originated recently.
We stress the importance of chromosomal rearrangements
and karyotypic variation as a potential source of diver-
gence among parthenogenetic lineages, which is in line
with postulates by others (Sunnucks et al. 1998; Blackman
et al. 2000).
Differing chromosome numbers in the sexual species
M. helleri (52 vs. 50 chromosomes, one male with 51)
showed that chromosomal rearrangements are not
restricted to parthenogenetic Machilis species. As the fun-
damental number of arms was congruent across different
karyotypes, the change in chromosome number is best
explained by fission of one or fusion of two chromo-
somes, while the single male with 51 chromosomes may
represent a hybrid karyotype or a chromosomal aberra-
tion. Alternatively, an additional chromosome in males
compared with females could, for example, also indicate
an XX/XYY’ sex determination system. The geographical
spacing of populations with 50 and 52 chromosomes
(Fig. 3) is contrasted by an increase of 10% in relative
10 ª2014 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd.
Genomic Variation in Machilis Jumping Bristletails M. Gassner et al.
genome size along a west–east gradient, resulting in larger
genomes in populations having fewer chromosomes.
Assuming that the ancestral chromosome number was 52,
an increase in relative genome size might be explained as
a by-product of the fusion of two chromosomes. Alterna-
tively, this gradient may represent an independent phe-
nomenon, possibly linked to an ecological cline or
varying retrotransposon activity.
We hypothesize that the change in chromosome num-
ber in M. helleri could have built up a reproductive bar-
rier, thus initiating genetic divergence. The evolutionary
implications differ from those in parthenogenetic species
(e.g., M. fuscistylis), where asexuality itself is a barrier to
gene flow. This becomes evident in our mitochondrial
phylogeny, where karyotypically differing individuals did
not form reciprocally monophyletic groups but showed
some incongruence in the populations RAX and HWD
(Fig. 3), possibly a consequence of limited, ongoing gene
flow between the karyotypic groups. Recently, it has been
shown that chromosomal rearrangements are key factors
promoting speciation during secondary contact in several
species pairs of rodents (Castiglia 2014). Similarly, the
pattern seen in M. helleri could be interpreted as an early
stage of speciation triggered by a chromosomal rearrange-
ment.
Occurrence of B chromosomes
Varying numbers of B chromosomes were found in the
sexual population of M. ticinensis. Interestingly, relative
genome size significantly increased with the number of B
chromosomes within this population, but the significant
difference in relative genome size between sexual and par-
thenogenetic populations could not be explained by their
presence alone. This is consistent with the results of Triv-
ers et al. (2004), who found approx. 35% larger genomes
in plants with reported B chromosomes compared with
plants without B chromosomes. Because B chromosomes
are thought to include a high amount of transposable ele-
ments (Camacho 2005), repeated transposition of mobile
elements to the original chromosome set and subsequent
proliferation could explain increased nuclear DNA con-
tent in the sexual population of M. ticinensis. Studying B
chromosomes is an excellent opportunity to learn more
about the dynamics of evolving genomes (Houben et al.
2014), and thus, M. ticinensis represents a prime study
system for future research targeting the impact of B chro-
mosomes on genome-size evolution.
Acknowledgments
We thank Marianne Magauer, Daniela Pirkebner, and
Barbara Pernfuß for technical assistance with flow cytometry,
Richard Hastik for help with specimen collection, and
Francesco Cicconardi for helpful comments on figures.
Conflict of Interest
None declared.
Data Accessibility
All DNA sequences were deposited at Genbank (acces-
sion numbers KJ501697 –KJ501918), including the final
alignment used in this study (PopSet: 632799655). Indi-
vidual chromosome counts and genome-size measure-
ments are listed in Table S1 of the online supplemental
material.
References
Adolfsson, S., Y. Michalakis, D. Paczesniak, S. N. S. Bode, R.
K. Butlin, D. K. Lamatsch, et al. 2010. Evaluation of
elevated ploidy and asexual reproduction as alternative
explanations for geographic parthenogenesis in Eucypris
virens ostracods. Evolution 64:986–997.
Bell, G. 1982. The masterpiece of nature: the evolution and
genetics of sexuality. Univ. of California Press, Berkeley.
Bengtsson, B. O. 2009. Asex and evolution: a very large-scale
overview. Pp. 1–19 in I. Sch€
on, K. Martens, P. V. Dijk, eds.
Lost sex. Springer, Dordrecht, The Netherlands.
Blackman, R. L., J. M. Spence, and B. B. Normark. 2000. High
diversity of structurally heterozygous karyotypes and rDNA
arrays in parthenogenetic aphids of the genus Trama
(Aphididae: Lachninae). Heredity 84:254–260.
Bode, S. N. S., S. Adolfsson, D. K. Lamatsch, M. J. F. Martins,
O. Schmit, J. Vandekerkhove, et al. 2010. Exceptional
cryptic diversity and multiple origins of parthenogenesis in
a freshwater ostracod. Mol. Phylogenet. Evol. 54:542–552.
Camacho, J. P. M. 2005. B chromosomes. pp. 223–286 in
T. R. Gregory, ed. The evolution of the genome. Elsevier
Academic Press, Burlington, MA, USA.
Castiglia, R. 2014. Sympatric sister species in rodents are more
chromosomally differentiated than allopatric ones:
implications for the role of chromosomal rearrangements in
speciation. Mamm. Rev. 44:1–4.
Choleva, L., and K. Janko. 2013. Rise and persistence of
animal polyploidy: evolutionary constraints and potential.
Cytogenet. Genome Res. 140:151–170.
Comai, L. 2005. The advantages and disadvantages of being
polyploid. Nat. Rev. Genet. 6:836–846.
Cuellar, O. 1977. Animal parthenogenesis. Science 197:837–843.
Darriba, D., G. L. Taboada, R. Doallo, and D. Posada. 2012.
jModelTest 2: more models, new heuristics and parallel
computing. Nat. Methods 9:772.
Dejaco, T., W. Arthofer, H. D. Sheets, K. Moder , B.
Thaler-Knoflach, E. Christian, et al. 2012. A toolbox for
ª2014 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd. 11
M. Gassner et al.Genomic Variation in Machilis Jumping Bristletails
integrative species delimitation in Machilis jumping
bristletails (Microcoryphia: Machilidae). Zoologischer
Anzeiger - J. Comp. Zool. 251:307–316.
Elzinga, J., J. Jokela, and L. N. Shama. 2013. Large variation in
mitochondrial DNA of sexual and parthenogenetic Dahlica
triquetrella (Lepidoptera: Psychidae) shows multiple origins
of parthenogenesis. BMC Evol. Biol. 13:90.
Ghiselli, F., L. Milani, V. Scali, and M. Passamonti. 2007. The
Leptynia hispanica species complex (Insecta Phasmida):
polyploidy, parthenogenesis, hybridization and more. Mol.
Ecol. 16:4256–4268.
Glesener, R. R., and D. Tilman. 1978. Sexuality and the
components of environmental uncertainty: clues from
geographic parthenogenesis in terrestrial animals. Am. Nat.
112:659–673.
Gould, S. J. 1970. Dollo on Dollo’s law: irreversibility and the
status of evolutionary laws. J. Hist. Biol. 3:189–212.
Gregory, T. R. 2014. Animal genome size database. http://
www.genomesize.com
Guindon, S., and O. Gascuel. 2003. A simple, fast, and
accurate algorithm to estimate large phylogenies by
maximum likelihood. Syst. Biol. 52:696–704.
Hessen, D. O., P. D. Jeyasingh, M. Neiman, and L. J. Weider.
2010. Genome streamlining and the elemental costs of
growth. Trends Ecol. Evol. 25:75–80.
H€
orandl, E. 2009. Geographic parthenogenesis: opportunities
for asexuality. Pp. 161–186 in I. Sch€
on, K. Martens and
P. V. Dijk, eds. Lost sex - the evolutionary biology of
parthenogenesis. Springer, Dordrecht.
Houben, A., A. Banaei-Moghaddam, S. Klemme, and
J. Timmis. 2014. Evolution and biology of supernumerary B
chromosomes. Cell. Mol. Life Sci. 71:467–478.
Janetschek, H. 1949. Beitrag zur Kenntnis der Felsenspringer
(Thysanura, Machilidae) Nordtirols. Ver€
offentlichungen des
Museums Ferdinandeum (Innsbruck) 26/29:147–165.
Janetschek, H. 1954.
€
Uber mitteleurop€
aische Felsenspringer
(Ins., Thysanura).
€
Osterreichische Zoologische Zeitschrift
5:281–328.
Janetschek, H. 1956. Das Problem der inneralpinen
Eiszeit€uberdauerung durch Tiere (Ein Beitrag zur Geschichte
der Nivalfauna).
€
Osterreichische Zoologische Zeitschrift
6:421–506.
Johnston, J. S., M. D. Bennett, A. L. Rayburn, D. W.
Galbraith, and H. J. Price. 1999. Reference standards for
determination of DNA content of plant nuclei. Am. J. Bot.
86:609–613.
Judson, O. P., and B. B. Normark. 1996. Ancient asexual
scandals. Trends Ecol. Evol. 11:41–46.
Kearney, M. 2003. Why is sex so unpopular in the Australian
desert? Trends Ecol. Evol. 18:605–607.
Kearney, M. 2005. Hybridization, glaciation and geographical
parthenogenesis. Trends Ecol. Evol. 20:495–502.
Kratochv
ıl, J. 1945. Na
se supinu
sky se zvl
a
stnim z
retelem na
moravsk
a chr
anĕn
a
uzem
ı. Folia Entomologica 8:41–67.
Law, J. H., and B. J. Crespi. 2002. The evolution of geographic
parthenogenesis in Timema walking-sticks. Mol. Ecol.
11:1471–1489.
Leitch, I. J., and M. D. Bennett. 2004. Genome downsizing in
polyploid plants. Biol. J. Linn. Soc. 82:651–663.
Lundmark, M., and A. Saura. 2006. Asexuality alone does not
explain the success of clonal forms in insects with
geographical parthenogenesis. Hereditas 143:23–32.
Mable, B. K. 2004. ‘Why polyploidy is rarer in animals than in
plants’: myths and mechanisms. Biol. J. Linn. Soc.
82:453–466.
Maccari, M., F. Amat, and A. G
omez. 2013. Origin and genetic
diversity of diploid parthenogenetic Artemia in Eurasia.
PLoS ONE 8:e83348.
Maniatsi, S., A. D. Baxevanis, I. Kappas, P. Deligiannidis, A.
Triantafyllidis, S. Papakostas, et al. 2011. Is polyploidy a
persevering accident or an adaptive evolutionary pattern?
The case of the brine shrimp Artemia. Mol. Phylogenet.
Evol. 58:353–364.
Manriquez-Moran, N. L., F. R. Cruz, and R. W. Murphy.
2014. Genetic variation and origin of parthenogenesis in the
Aspidoscelis cozumela complex: evidence from mitochondrial
genes. Zoolog. Sci. 31:14–19.
Maynard Smith, J. 1978. The evolution of sex. Cambridge
Univ. Press, Cambridge, UK.
Otto, S. P., and J. Whitton. 2000. Polyploid incidence and
evolution. Annu. Rev. Genet. 34:401–437.
Palissa, A. 1964. Apterygota. Quelle & Meyer, Leipzig.
Reeves, A. A. J. T. 2000. MicroMeasure for Windows, version
3.3., Free program distributed by the authors over the
Internet from http://www.colostate.edu/Depts/Biology/
MicroMeasure.
Riezler, H. 1941.
€
Uber Machiliden Nordtirols.
Ver€
offentlichungen des Museums Ferdinandeum
(Innsbruck) 19:193–267.
Rinnhofer, L., N. Roura-Pascual, W. Arthofer, T. Dejaco, B.
Thaler-Knoflach, G. Wachter, et al. 2012. Iterative species
distribution modelling and ground validation in endemism
research: an Alpine jumping bristletail example. Biodivers.
Conserv. 21:2845–2863.
Ronquist, F., M. Teslenko, P. van der Mark, D. L. Ayres, A.
Darling, S. H€ohna, et al. 2012. MrBayes 3.2: efficient
bayesian phylogenetic inference and model choice across a
large model space. Syst. Biol. 61:539–542.
Sahara, K., F. Marec, and W. Traut. 1999. TTAGG telomeric
repeats in chromosomes of some insects and other
arthropods. Chromosome Res. 7:449–460.
Sch€
onswetter, P., J. Suda, M. Popp, H. Weiss-Schneeweiss, and
C. Brochmann. 2007. Circumpolar phylogeography of Juncus
biglumis (Juncaceae) inferred from AFLP fingerprints,
cpDNA sequences, nuclear DNA content and chromosome
numbers. Mol. Phylogenet. Evol. 42:92–103.
Schwander, T., and B. J. Crespi. 2009. Multiple direct
transitions from sexual reproduction to apomictic
12 ª2014 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd.
Genomic Variation in Machilis Jumping Bristletails M. Gassner et al.
parthenogenesis in Timema stick insects. Evolution 63:84–
103.
Stenberg, P., M. Lundmark, S. Knutelski, and A. Saura. 2003.
Evolution of clonality and polyploidy in a weevil system.
Mol. Biol. Evol. 20:1626–1632.
Suda, J., H. Weiss-Schneeweiss, A. Tribsch, G. M. Schneeweiss,
P. Travn
ıcek, and P. Sch€onswetter. 2007. Complex
distribution patterns of di-, tetra-, and hexaploid cytotypes
in the European high mountain plant Senecio carniolicus
(Asteraceae). Am. J. Bot. 94:1391–1401.
Sunnucks, P., D. Chisholm, E. Turak, and D. F. Hales. 1998.
Evolution of an ecological trait in parthenogenetic Sitobion
aphids. Heredity 81:638–647.
Suomalainen, E. 1940. Polyploidy in parthenogenetic
Curculionidae. Hereditas 26:51–64.
Suomaleinen, E., A. Saura, and J. Lokki. 1987. Cytology and
evolution in parthenogenesis. CRC Press Inc,
Boca Raton, FL.
Tamura, K., D. Peterson, N. Peterson, G. Stecher, M. Nei, and
S. Kumar. 2011. MEGA5: Molecular evolutionary genetics
analysis using maximum likelihood, evolutionary distance,
and maximum parsimony methods. Mol. Biol. Evol.
28:2731–2739.
Trivers, R., A. Burt, and B. G. Palestis. 2004. B
chromosomes and genome size in flowering plants.
Genome 47:1–8.
Vandel, A. 1928. La parth
enogen
ese g
eographique:
Contribution
al’
etude biologique et cytologique de la
parth
enogen
ese naturelle. Bulletin biologique de la France et
de la Belgique 62:164–281.
Vrijenhoek, R. C., and E. D. Jr Parker. 2009. Geographical
parthenogenesis: General purpose genotypes and frozen niche
variation. Pp. 99–131 in I. Sch€
on, K. Martens and P. V. Dijk,
eds. Lost sex. Springer, Dordrecht, The Netherlands.
Wachter, G. A., W. Arthofer, T. Dejaco, L. J. Rinnhofer, F. M.
Steiner, and B. C. Schlick-Steiner. 2012. Pleistocene survival
on central Alpine nunataks: genetic evidence from the
jumping bristletail Machilis pallida. Mol. Ecol. 21:4983–
4995.
Wygodzinsky, P. W. 1941. Beitr€
age zur Kenntnis der Dipluren
und Thysanuren der Schweiz. Denkschriften der
Schweizerischen Naturforschenden Gesellschaft 74:110–227.
Zhang, L., and H. Lefcort. 1991. The effects of ploidy level on
the thermal distributions of brine shrimp Artemia
parthenogenetica and its ecological implications. Heredity
66:445–452.
Supporting Information
Additional Supporting Information may be found in the
online version of this article:
Figure S1. Karyotypes from selected individuals for each
species included in this study
Table S1. Raw data and sampling information for all
individuals included in this study.
ª2014 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd. 13
M. Gassner et al.Genomic Variation in Machilis Jumping Bristletails
