Content uploaded by Elizabeth A. Sinclair
Author content
All content in this area was uploaded by Elizabeth A. Sinclair on Oct 24, 2017
Content may be subject to copyright.
ORIGINAL ARTICLE
doi:10.1111/j.1558-5646.2009.00893.x
DNA EVIDENCE FOR NONHYBRID ORIGINS
OF PARTHENOGENESIS IN NATURAL
POPULATIONS OF VERTEBRATES
Elizabeth A. Sinclair,1,2Jennifer B. Pramuk,3Robert L. Bezy,4Keith A. Crandall,5and Jack W. Sites Jr.5
1Botanic Gardens and Parks Authority and Department of Plant Biology, The University of Western Australia, West Perth,
6005, Western Australia, Australia
2E-mail: esinclair@iinet.net.au
3Wildlife Conservation Society/Bronx Zoo, 2300 Southern Blvd. Bronx, New York 10460
4Natural History Museum of Los Angeles County, Los Angeles, California 90007
5Department of Biology and M. L. Bean Life Science Museum, Brigham Young University, Provo, Utah 84602
Received February 16, 2009
Accepted October 16, 2009
Naturally occurring unisexual reproduction has been documented in less than 0.1% of all vertebrate species. Among vertebrates,
true parthenogenesis is known only in squamate reptiles. In all vertebrate cases that have been carefully studied, the clonal
or hemiclonal taxa have originated through hybridization between closely related sexual species. In contrast, parthenogenetic
reproduction has arisen in invertebrates by a variety of mechanisms, including likely cases of “spontaneous” (nonhybrid) origin, a
situation not currently documented in natural populations of vertebrates. Here, we present molecular data from the Neotropical
night lizard genus Lepidophyma that provides evidence of independent nonhybrid origins for diploid unisexual populations
of two species from Costa Rica and Panama. Our mitochondrial and nuclear phylogenies are congruent with respect to the
unisexual taxa. Based on 14 microsatellite loci, heterozygosity (expected from a hybrid origin) is low in Lepidophyma reticulatum
and completely absent in unisexual L. flavimaculatum. The unique value of this system will allow direct comparative studies
between parthenogenetic and sexual lineages in vertebrates, with an enormous potential for this species to be a model system
for understanding the mechanisms of nonhybrid parthenogenesis.
KEY WORDS: Lepidophyma, lizard, microsatellites, nonhybrid origin, parthenogenesis, phylogenetics.
The evolutionary success of sexual reproduction has been at-
tributed to two factors: sexuals can adapt more rapidly to chang-
ing environments and they are less prone to the accumulation of
deleterious mutations (Kondrashov 1993). Unisexual organisms
present a unique opportunity to test these two hypotheses (Simon
et al. 2003). Unisexual reproduction falls into one of three cat-
egories: hybridogenesis (hemiclonal reproduction where the fe-
male half of the genome is passed intact to the next generation
following fertilization), gynogenesis (reproduction in the pres-
ence of sperm, but no fertilization of the egg), or true partheno-
genesis (no participation or requirement from a male). Hybri-
dogenesis and gynogenesis are known in fish and amphibians,
and all-female clones are dependent to some degree on males of a
closely related bisexual species. Although unisexual invertebrates
have been known for centuries and are relatively well studied
(summaries in Suomalainen et al. 1987; Cuellar 1987), unisexual
vertebrates were not discovered until the mid-twentieth century
(summary in Dawley and Bogart 1989). In subsequent decades,
the number of known unisexual vertebrates has increased consid-
erably and stands at more than 70 “biotypes” of parthenogenetic
reptiles and gynogenetic or hybridogenetic amphibians and fish
(Vrijenhoek et al. 1989; Kearney et al. 2009). Although unisexuals
1
C
2009 The Author(s).
Evolution
ELIZABETH A. SINCLAIR ET AL.
constitute less than 0.1% of all vertebrates, they continue to at-
tract considerable attention, in part due to a fascination with the
strange and novel, but also from a recognition that the excep-
tion provides a vantage point from which to evaluate the norm
(Dawley and Bogart 1989; Vrijenhoek 1994, 1998; Beukeboom
and Vrijenhoek 1998; Barraclough et al. 2003; Kearney et al.
2009).
The critical question of how unisexuals originate has impor-
tant implications regarding the roles of genetic variability and
the evolution of sex (Barton and Charlesworth 1998). Histor-
ically, two competing hypotheses have been proposed for the
origin of natural unisexual populations; interspecific hybridiza-
tion and (spontaneous) mutational hypotheses. To date, all natu-
ral populations of unisexual vertebrate taxa investigated, regard-
less of whether they reproduce by gynogenesis, hybridogenesis,
or parthenogenesis, show patterns of genetic variation consis-
tent with an origin by interspecific hybridization between closely
related species (Dawley and Bogart 1989; Simon et al. 2003;
Korchagin et al. 2007; Kearney et al. 2009). Here, we present
phylogenetic (mitochondrial and nuclear sequences) and allelic
(microsatellite) data that are consistent with a nonhybrid origin
for unisexuals in two species of night lizards (Lepidophyma:Xan-
tusiidae) from Central America.
The Hybrid Origin Hypothesis (HOH) postulates that two
closely related gonochoristic (“bisexual” hereafter) species inter-
breed and successfully produce viable offspring that then repro-
duce parthenogenetically. Vertebrate parthenogens are normally
diploid or triploid and commonly reflect high levels of nuclear
gene diversity due to “fixed heterozygosity” at codominant loci
for which the two hybridizing parental species segregate for al-
ternative alleles (Simon et al. 2003). Despite the high nuclear
gene heterozygosity, parthenogens are generally thought to be
ephemeral over evolutionary time (Moritz 1993; but see Fontaneto
et al. 2007). An alternative origin for parthenogenetic forms
is through the accumulation of spontaneous mutation(s) or the
Mutational Origin Hypothesis (MOH) (Suomalainen et al. 1987;
Bullini 1994; Johnson and Leefe 1999). Little is understood about
the mechanisms of such an origin, but the most obvious difference
is detectable genetic evidence for a single parental species under
the MOH, whereas a unique combination of divergent genotypes
from two distinct parentals is required to support the HOH (Simon
et al. 2003). With adequate sampling, a combination of mitochon-
drial and nuclear markers can be used to discriminate between
these alternatives (Fig. 1).
The best vertebrate candidates for a nonhybrid origin of
parthenogenesis involve unisexual populations in two species of
night lizards (Lepidophyma). The genus extends from northern
Mexico through Central America to Panama (Fig. 2), and currently
18 species are recognized (Bezy and Camarillo 2002; Canseco-
M´
arquez et al. 2008). In Lepidophyma reticulatum (Costa Rica)
HOH MOH
nDNA mtDNA
Species C (parthenogen)
parental haplotype 1
Species C (parthenogen)
parental haplotype 2
Speci es A
Speci es B
Outgroup
A
B
C
D
Species D
Speci es B
Speci es A
Species C (parthenogen)
Species D
Outgroup
Speci es B
Speci es A
Speci es C (part henog en)
Speci es D
Outgroup
Speci es B
Species A
Speci es C (part henog en)
Species D
Outgroup
Figure 1. Hypothetical phylogenies under a hybrid (HOH) and
mutational (MOH) origin for parthenogens. (A) HOH for a mito-
chondrial phylogeny; the HOH predicts phylogenetic recovery of
mitochondrial haplotypes for the parthenogen with the bisexual
descendent of the female parental species. (B) HOH for a nuclear
phylogeny; nuclear genes for which alternative alleles segregate
in different hybridizing bisexual species are expected to be ex-
pressed as fixed heterozygous genotypes in parthenogens. Fur-
ther, phylogenetic reconstruction should “split” these fixed hybrid
heterozygotes into alternative clades (Moritz 1987; Sites et al.
1990), each of which is recovered with the descendants of ei-
ther the maternal or paternal ancestors, and the maternal clade
should match that recovered in the mtDNA analysis. This test will
be strongest if the nuclear alleles are synapomorphic characters
in the phylogeny, because this eliminates the possibility that het-
erozygous genotypes in the parthenogens represent shared an-
cestral polymorphisms (Murphy et al. 2000). A multivariate anal-
ysis of microsatellite genotypes from parthenogens should form
a distinct cluster that is intermediate to those representing the
two bisexual species. (C) MOH for a mitochondrial phylogeny; the
MOH predicts an mtDNA phylogeny in which the parthenogens
are recovered with their sister bisexual species. (D) MOH for a
nuclear phylogeny; the nuclear gene tree should recover a sim-
ilar or identical topology to the mtDNA tree. Any nuclear het-
erozygotes in the parthenogens would reflect the idiosyncratic
history of mutations shared with the single bisexual ancestral lin-
eage, or alleles of mutational origin that define clonal diversity in
parthenogens. Clustering of multilocus microsatellite genotypes
from parthenogens should form one group embedded within the
single parental species.
all known individuals are female; L. flavimaculatum populations
in Panama and south-eastern Costa Rica lack males, whereas
northern populations contain both sexes and appear to be bisex-
ual. Limited available data from morphology, karyotypes, and
allozymes (summaries in Bezy 1972; Bezy and Sites 1987; Bezy
and Camarillo 2002) do not provide evidence for a hybrid ori-
gin of the all-female populations of L. flavimaculatum. To date,
all unisexual populations studied are diploid (except one 2N/3N
2EVOLUTION 2009
DNA EVIDENCE FOR NONHYBRID ORIGINS OF PARTHENOGENESIS
Figure 2. Map of Mexico and Central America showing the global distribution of Lepidophyma species. Sampling localities are indicated
by black filled circles (Appendix S1); species not sampled (red crosses), and parthenogens (blue stars); distribution of parthenogenetic
forms (hatched area); region of sympatry between L. reticulatum and bisexual L. flavimaculatum (unshaded blue).
mosaic individual; Bezy 1972). The earlier studies, however, were
limited in taxon sampling and resolution of the genetic charac-
ters. Here, we test the HOH and MOH for the unisexual species
L. reticulatum and unisexual populations of L. flavimaculatum
using phylogenetic and population genetic methods, including
dense population sampling, high-resolution molecular data, and
sufficient taxonomic sampling to include all likely parental taxa.
Materials and Methods
SAMPLE COLLECTION
Fourteen of the 18 recognized Lepidophyma species (Bezy and
Camarillo 2002; Canseco-M´
arquez et al. 2008) were collected
for DNA sequencing and subsequent phylogenetic and popula-
tion genetic analyses (Appendices S1 and S2). Four species not
included in this study (L. lineri, L. chicoasense, L. tarascae,and
the recently described L. cuicateca) have extremely restricted dis-
tributions in southern Mexico and are either very rare or extinct
(Bezy and Camarillo 2002; Canseco-M´
arquez et al. 2008). Sev-
eral other species also have very localized distributions and are
extremely difficult to collect; out of necessity, five species are rep-
resented here by single tissue samples, and two species each are
represented by two and three samples, respectively. The collec-
tion of individuals used here represents over 30 years of fieldwork
by RLB and colleagues. Sequences from five species of the sister
genus Xantusia (Vicario et al. 2003) were used as outgroups: (X.
bolsonae, X. extorris, X. gilberti, X. riversiana,andX. vigilis).
PCR AND PHYLOGENETIC ANALYSES
Two mitochondrial genes were amplified: the complete Cy-
tochrome b(Cyt b; 1143bp) and partial 12S rRNA (12S; 924bp).
Cyt bwas amplified in two fragments using the primer combi-
nations L14724 and CB3 and F1 and RD. The 12S region was
amplified using 12StPhe and 12Se. Polymerase chain reaction
(PCR) conditions are the same as those described in Sinclair et al.
(2004). Eight nuclear regions (alpha-enolase [a-Enol; 238bp],
oocyte maturation factor [C-mos; 495bp], proopiomelanocortin-A
gene [POMC; 516bp], brain-derived neurotrophic factor [BDNF;
708bp], glyceraldehyde-3-phosphate dehydrogenase [Gapdh, in-
tron 11; 305bp], neurotrophin-3 gene [NT-3; 528bp], recombina-
tion activating-1 gene [RAG-1; 840bp], and Pinin [PNN; 932bp])
were also amplified using standard PCR conditions. Details of the
primer sequences and source are given in Appendix S3. All se-
quences are available through GenBank (accession numbers given
in Appendix S1).
PCR cycling was performed in a 9600 thermocycler
(PerkinElmer), and 2 μl of PCR product was run on 1% agarose
gels stained with ethidium bromide and viewed under UV light.
All PCR products were cleaned using Millipore plates. Sequenc-
ing reactions were performed in 4 μl volumes, using the ABI Big-
dye Ready-Reaction kit, following the standard cycle sequencing
protocol. Double-stranded sequences were generated using an
ABI3730XL automated DNA sequencer. Sequences were edited
and aligned using Sequencher (Gene Codes Corp.) and checked
by eye. Alignment gaps were used to maximize codon identity and
minimize the number of insertions or deletions. For the coding
EVOLUTION 2009 3
ELIZABETH A. SINCLAIR ET AL.
(exon) regions, gaps represented a gain/loss of complete amino
acids. These gaps were usually fixed within species. All align-
ments were relatively unambiguous as sequences were from two
closely related sister genera. Reading frames for mitochondrial
and nuclear protein-coding regions were checked to guard against
pseudogene amplification (Song et al. 2008).
Phylogeny reconstruction for 2067bp of mtDNA sequence
was performed using Bayesian Inference (Huelsenbeck and
Ronquist 2001) and maximum likelihood (ML) in PAUP∗
(Swofford 2002). The program ModelTest 3.07 (Posada and Cran-
dall 1998) was used to select the “best-fit” model of evolution for
the combined mitochondrial dataset using the AIC criterion. For
the Bayesian analysis, four independent searches were performed
with three million generations each and four incrementally heated
Markov chains were used to enable a more thorough search of
the parameter space. “Burnin” plots were examined and the ini-
tial 50,000 replicates were excluded from subsequent analysis. A
50% majority-rule consensus tree was generated using the remain-
ing replicates for each of the searches in PAUP∗(Swofford 2002).
The percentage of samples recovering a particular clade was taken
as that clade’s posterior probability (Huelsenbeck and Ronquist
2001). ML runs were performed on a 68-node cluster running
Debian Linux each with two Intel Xeon quad core processors
(E5345) (Intel Corp, Santa Clara, CA) at Brigham Young Univer-
sity. The nuclear gene sequences (4562bp) were analyzed sepa-
rately using equally weighted Maximum Parsimony in PAUP∗and
Bayesian analysis incorporating the “best-fit” model of evolution
for each gene region. In the parsimony analyses, gaps were coded
as a fifth character state, as they were deemed informative. Each
gap was treated as a single event. One thousand search replicates
were performed using the TBR search algorithm, and nodal sup-
port was assessed using 1000 bootstrap replicates. A Shimodaira–
Hasegawa test (S–H test; Shimodaira and Hasegawa 1999) was
executed in PAUP∗(using a RELL distribution with 1000 repli-
cates and separate tests executing the mtDNA and nDNA datasets)
to test whether each of the combined mtDNA and nuclear trees
were significantly different. The phylogenies (and shared nuclear
sequences) were then used to identify potential paternal species
under an HOH.
We tested for recombination in each nuclear gene region
using the software RDP3 (Martin et al. 2005). The program ex-
amines nucleotide sequence alignments and attempts to identify
recombinant sequences and breakpoints using 10 published re-
combination detection methods with a range of performance, in-
cluding approaches well suited for our observed levels of genetic
diversity (Posada and Crandall 2001).
MICROSATELLITE LOCI
Fourteen di- and tetra-nucleotide microsatellite markers were am-
plified in taxa that were both phylogenetically and geographically
proximal to L. flavimaculatum and L. reticulatum (Appendix S2)
using conditions described in Sinclair et al. (2006). By avoiding
distantly related, divergent Lepidophyma species in this analysis,
we reduce the possibility of spurious allelic similarity due to ho-
moplasy. We used GenePop-on-the-web version 3.4 (Raymond
and Rousset 1995) to test for Hardy–Weinberg equilibrium and
linkage disequilibrium for five samples for which at least eight
individuals were genotyped. These were the bisexual populations
of L. flavimaculatum (Rancho Grande, n=11; Tapezco, n=9),
the unisexual population of L. flavimaculatum (Escobal, n=9),
L. reticulatum (Las Cruces, n=9), and L. sylvaticum (San Luis
Potosi, n=10). Hardy–Weinberg exact tests were performed for
each locus where possible, and a Holm–Bonferroni method was
used to correct for multiple tests (Holm 1979). A principal co-
ordinate analysis (PCA) was performed in GenAlEx version 6.1
(Peakall and Smouse 2006) to visually represent the relative de-
gree of genetic similarity among individuals and the distinction,
if any, among sampled populations/species.
Results
PHYLOGENETIC RELATIONSHIPS
The mitochondrial tree recovered all species of Lepidophyma rep-
resented by more than one terminal as clades (ML bootstrap =
96–100; Bayesian PP =1.0), with the exception of L. sylvaticum,
which is paraphyletic relative to L. micropholis (Fig. 3A). All
bisexual and unisexual L. flavimaculatum terminals were recov-
ered as a clade sister to the unisexual L. reticulatum (both well-
supported). Both parthenogens are in a highly derived position in
the tree, and both show a low level of within-taxon mtDNA vari-
ation (<2.0% sequence divergence). The nuclear tree (Fig. 3B)
also recovered the L. flavimaculatum (including bisexual and uni-
sexual terminals) and L. reticulatum clades as sister taxa (MP
bootstrap =99; Bayesian PP =1.0). The S–H tests indicated
there were significant differences between the mitochondrial and
nuclear gene trees (mtDNA, P<0.001, nDNA, P=0.051).
However, the differences involved several poorly sampled taxa
whose phylogenetic position was not well-resolved (L. occulor,
L. pajapanense, and L. mayae). In both trees, the unisexual and
bisexual L. flavimaculatum terminals formed a clade, sister to the
unisexual species L. reticulatum, with L. lipetzi comprising the
sister species to this group (L. lipetzi (L. reticulatum (L. flavimac-
ulatum-bisexual +L. flavimaculatum-unisexual))), (mtDNA ML
bootstrap =99; nuclear; MP bootstrap =100). This topology is
consistent with phylogenetic predictions under an MOH for nu-
clear and mitochondrial sequences (see Fig. 1C,D), but not the
HOH.
Of all nuclear genes sequenced, the Gapdh region was the
most variable at the species level, and proved to be a good marker
for detection of hybridization in Xantusia (Leavitt et al. 2007).
4EVOLUTION 2009
DNA EVIDENCE FOR NONHYBRID ORIGINS OF PARTHENOGENESIS
50 changes
dontomasi 20
dontomasi 23
flavimaculatum 1066
radula 21500
gaigeae 4051
gaigeae 233
gaigeae 4090
pajapanense 542
lipetzi 3792
lowei 3729
lowei 3734
mayae 1234
mayae 1233
occulor 48499
smithii 1057
smithii 5537
reticulatum 6319*
reticulatum 6320*
reticulatum 6318*
X. vigilis
X. riversiana
X. gilberti
X. extorris
X. bolsonae
tuxtlae 5526
tuxtlae 5528
flavimaculatum 1214
micropholis 532
micropholis 1188
micropholis 5546
sylvaticum 1121
sylvaticum 4076
sylvaticum 5544
sylvaticum 5545
sylvaticum 275
flavimaculatum 1072
flavimaculatum 182
flavimaculatum 1017*
flavimaculatum 5285.1*
flavimaculatum 168
flavimaculatum 923
flavimaculatum 30
flavimaculatum 999
flavimaculatum 992
flavimaculatum 4135
flavimaculatum 13355
flavimaculatum 13200
flavimaculatum 11570
100/1.00
100/
1.00
100/1.00
67/
1.00
70/
0.87
55/
0.67
-/
0.53
100/1.00
96/1.00
94/1.00
100/1.00
92/1.00
98/1.00
72/
0.91
100/1.00
-/0.55
100/
1.00
100/
1.00
99/1.00
100/1.00
-/0.65
100/
1.00
100/
1.00
100/1.00
100/1.00
100/1.00
100/1.00
99/
1.00
89/1.00
99/1.00
99/1.00
-/0.75
AB
dontomasi 20
dontomasi 23
radula 21500
lowei 3729
lowei 3731
smithii 5537
reticulatum 6317*
reticulatum 6319*
10 changes
flavimaculatum 1066
flavimaculatum 1072
flavimaculatum 30
flavimaculatum 923
flavimaculatum 13355
flavimaculatum 168
flavimaculatum 999
flavimaculatum 4135
flavimaculatum 13200
flavimaculatum 1018*
flavimaculatum 5285.1*
sylvaticum 5542
sylvaticum 1155
sylvaticum 275
sylvaticum 4076
micropholis 1188
micropholis 5546
micropholis 532
sylvaticum 5543
X. bolsonae
X. extorris
X. riversiana
reticulatum 6320*
X. vigilis
X. gilberti
pajapanense 542
occulor 48499
mayae 1233
mayae 1234
lipetzi 3792
smithii 1057
tuxtlae 5526
tuxtlae 5528
gaigeae 4056
gaigeae 233
gaigeae 4090
flavimaculatum 1214
flavimaculatum 5288.2
94/1.00
99/1.00
86/1.00
96/
1.00
100/1.00
93/0.99
99/1.00
98/
1.00
100/
1.00
100/
1.00
100/1.00
100/1.00
80/1.00
91/
1.00
83/1.00
99/
1.00
100/1.00
81/1.00
79/1.00
82/0.95
98/1.00
99/1.00
95/0.59
100/1.00
-/0. 94
Figure 3. Phylogeny reconstructions for Lepidophyma. (A) Combined mitochondrial phylogeny based on 2067bp of Cyt band 12S, -Llk =
−17771.684; the Akaike information criterion (AIC) implemented in ModelTest, selected the General Time Reversible model plus gamma
distribution rate heterogeneity (GTR+I+G). The parameters were base frequencies A =0.3403, C =0.2989, G =0.1324, and T =0.2284,
transition rates =(A–G) 9.9071 and (C–T) 12.5819, transversion rates =(A–C) 2.1624, (A–T) 1.7805, (C–G) 0.8232, (G–T) 1.000, gamma
distribution shape parameter (G) =0.9070, and the proportion of invariable sites (I) =0.3872. Nodal support =ML bootstrap/Bayes PP.
(B) combined nuclear sequences (4562bp; 28 equally most parsimonious trees of 629 steps, nodal support =MP bootstrap/Bayes PP;
∗indicates parthenogenic individuals; terminals are identified by species name and sample number (Appendix S1).
In Lepidophyma, these sequences detected heterozygosity as
evidenced by two nucleotides (double peaks in the electrophero-
grams; Brumfield et al. 2003) at seven nucleotide positions in
the bisexual species L. sylvaticum from Mexico. The heterozy-
gous positions were not fixed in L. sylvaticum; only four of the
10 alternative alleles were found in the same or nearby popula-
tions. No heterozygous individuals were observed among any of
the unisexual or bisexual terminals for L. flavimaculatum (n=
13) or L. reticulatum (n=3) at any of these or other nucleotide
positions, consistent with an MOH. In fact, all L. flavimaculatum
sequences from unisexual and bisexual individuals were identical
to each other, and with a single nucleotide substitution from uni-
sexual L. reticulatum (also identical to each other). This pattern
was almost identical for all eight nuclear gene regions. Although
these results may be an artifact of limited sample sizes, they are
confirmed by a more extensive microsatellite dataset summarized
below. Furthermore, for clonal species of hybrid origin, multiple
heterozygosities are normally found even with a sample size of
one, and large samples show the heterozygosity to be fixed, with
very few exceptions.
MICROSATELLITE VARIATION
For the microsatellite data (Appendix S2), we found no evidence
of fixed heterozygosity in the L. flavimaculatum unisexuals from
Panama; all 14 loci were homozygous. Average heterozygosity
levels ranged from 25.5% to 30.8% for the two bisexual L. flavi-
maculatum populations from Costa Rica, with no deviation from
Hardy–Weinberg ratios (Table 1). The spatial arrangement of mul-
tilocus genotypes showed extensive among-population variation
within L. flavimaculatum (Fig. 4). All unisexual L. flavimacula-
tum form a single cluster, together with the most geographically
proximal bisexual populations of El Tigre and Tapezco, Costa
EVOLUTION 2009 5
ELIZABETH A. SINCLAIR ET AL.
Tab l e 1 . Summary of observed heterozygosity levels (%) for each microsatellite locus.
Species N
LflA102
LflA104
LflA107
LflB102
LflC101
LflC102
LflC104
LflC105
LflC109
LflC110
LflC112
LflD103
LflD120
LflD124
Overall
L. flavimaculatum (bi) 35 17.1∗24.2∗2.9∗17.1∗22.9∗8.6∗34.4∗45.7∗42.910.0∗11.4∗67.6∗62.9∗38.2∗29.0
Rancho Grande (bi) 11 0.040.00.0 27.30.00.0 63.654.581.80.00.0 81.854.527.330.8
Tapezco (bi) 9 0.012.50.0 22.2 44.40.0 11.133.322.211.10.0 66.7 100.033.3∗25.5
Panama (uni) 11 0.00.00.00.00.00.00.00.00.00.00.00.00.00.00.0
L. spp3794/5793 2 100.0 100.00.00.0 50.00.00.050.0 50.00.050.0 50.00.0 100.039.3
L. reticulatum (uni) 12 0.00.09.1 63.60.0∗0.0∗0.072.7 36.411.1∗9.18.3∗60.00.0∗19.3
Las Cruces (uni) 9 0.00.00.0 62.50.00.00.075.0 12.50.0∗12.50.0∗42.90.0∗14.7
L. lipetzi 10.050.00.00.00.00.0 50.0 100.0 50.0– 50.0 100.0– 0.033.3
L. sylvaticum
San Luis Potosi (bi) 10 70.070.0 30.0 40.00.0∗80.0 20.030.0∗0.040.080.0 22.2 100.077.850.8
∗Deficit in heterozygosity.
Species or population: bi, bisexual; uni, unisexual.
Figure 4. Principal coordinate analysis of multilocus geno-
types for the L. flavimaculatum/L. reticulatum group. The two
parthenogens form discrete clusters, one for L. reticulatum (solid
circles) and a second for unisexual L. flavimaculatum (solid di-
amonds), which are the geographically closest samples from El
Tigre and Tapezco (gray diamonds) and well within the spread of
all bisexual L. flavimaculatum (open diamonds); L. lipetzi (open
circle).
Rica. Our most interior Costa Rican sample from Rancho Grande
formed a separate cluster, and contains private alleles not found
elsewhere in bisexual or unisexual L. flavimaculatum or in L. retic-
ulatum: A104: (144, 148), B102: (258), C104 (186, 191, 192),
D120 (328; Table 2). Within L. flavimaculatum, alleles present
in the unisexual population from Panama are all present in bisex-
ual populations from Costa Rica, with the exception of three rare
alleles segregating in Belize (locus D103–227, freq =0. 059; lo-
cus D124–256, freq =0.015), and Guatemala (locus D103–227,
freq =0.103). All L. reticulatum clustered loosely together, but
differences between the two populations are evident.
Heterozygosity levels varied across microsatellite loci for L.
reticulatum (0.0–75.0%, Table 1). The overall heterozygosity was
lower in unisexual L. reticulatum (14.7–19.3%) than bisexual L.
flavimaculatum (25.5–30.8%). Five of eight polymorphic loci had
a significant deficit in heterozygotes (Table 1). Two populations
were sampled within L. reticulatum—Las Cruces (n=9) and Rin-
con (n=3), and despite small samples there were fixed differences
at seven loci whereas the other seven loci were heterozygous but
shared common alleles between the localities. The pattern of vari-
ation in L. reticulatum is also inconsistent with hybridization for
example, five of eight individuals from the Las Cruces sample are
heterozygous at locus B102 (260/262); allele 262 is unique to the
population, suggesting that heterozygosity at this locus is best ex-
plained by postorigin mutation rather than hybridization. A large
number of alleles are shared with L. flavimaculatum bisexuals
6EVOLUTION 2009
DNA EVIDENCE FOR NONHYBRID ORIGINS OF PARTHENOGENESIS
Tab l e 2 . Summary of alleles within L. flavimaculatum, L. reticulatum, and L. mayae. Alleles shared between unisexuals and bisexuals are in bold.
Locus L. reticulatum Het L. flavimaculatum Het L. flavimaculatum bisexual Het L. flavimaculatum bisexual Het L. mayae bisexual Het
unisexual (Las Cruces, unisexual Costa Rica northern Central America Guatemala
Costa Rica) (Escobal, Panama)
A102 195 0.0 195 0.0 189, 195 0.0 195, 197, 199, 201, 205 42.9 185, 187, 189 20.0
A104 124 0.0 136 0.0 126, 130, 136, 140, 144, 148 23.1 124, 128, 134, 140 26.6 124, 130, 136,
137, 140
80.0
A107 148, 157 9.1 152 0.0 152 0.0 146, 150, 152, 154 14.3 135, 142, 152,
157,159, 167
40.0
B102 252, 254, 256, 260, 262 63.6 256 0.0 248, 250, 256, 258 21.4 248, 254, 260, 270, 272, 274 42.9 227, 242 0.0
C101 176, 183, 187 0.0 171 0.0 155, 163, 167, 171, 175 14.3 163, 167, 171, 175, 179, 183 42.9 143, 164, 176 20.0
C102 209, 213 0.0 197 0.0 197, 217 7.1 197, 213, 226 14.3 205, 226, 230,
234, 239
40.0
C104 148 0.0 176 0.0 176, 180, 183, 186, 188, 191, 192 37.0 156, 176, 180, 183, 180 16.7 151, 254 25.0
C105 299, 307, 311, 315, 323,
397
72.7 299, 303 0.0 287, 290, 295, 299, 303 42.9 290, 295, 299, 303, 306,
307, 311
57.1 273, 356, 372,
384, 389, 399
40.0
C109 266, 275, 279, 287, 291,
303, 311
36.4 275 0.0 257, 275, 279, 283, 287 42.8 271, 275, 279, 283 42.9 234 0.0
C110 178, 183, 191, 216, 237 11.1 220, 225, 229 0.0 179, 220, 225, 229, 233 7.1 200, 205, 216, 220, 225 33.3 183, 203, 208,
211, 224
80.0
C112 210, 227 9.1 206 0.0 206 0.0 201, 206, 214, 219, 221, 230 57.1 203, 219, 222 20.0
D103 155, 207, 211, 219, 227,
231
8.3 211, 223, 227 0.0 203, 207, 211, 215, 219, 223, 231,
235
64.3 203, 207, 219, 223, 227,
231, 235
66.7 163, 169 0.0
D120 295, 299, 303 60.0 283, 287, 291, 307,
312
0.0 271, 275, 279, 283, 287, 291, 295,
299, 303, 307, 312, 316, 320,
328
64.3 279, 287, 291, 307, 312,
316, 320
66.7
–
–
D124 152, 244, 248, 252, 256 0.0 248, 252, 256 0.0 224, 228, 232, 236, 240, 244, 248,
252
25.0 224, 232, 236, 244, 248,
256, 260
33.0
–
–
EVOLUTION 2009 7
ELIZABETH A. SINCLAIR ET AL.
(25/48 =52%) and eight alleles are shared with L. flavimacu-
latum unisexuals. The non-flavimaculatum alleles present in L.
reticulatum are shared with some of the other bisexual species
(data not shown). Ten private alleles from seven loci are exclu-
sive to L. reticulatum. Nine alleles across six loci are shared
between bisexual and unisexual populations of L. flavimaculatum
and at least one of the L. reticulatum populations. This pattern
supports a spontaneous origin of both unisexuals from a bisex-
ual L. flavimaculatum-like ancestor, but also reflects subsequent
divergence of the unisexuals. There was no evidence that the ge-
netic variation present in the unisexuals could be split into two
separate genomes inherited from divergent parental species or
populations.
Discussion
HYBRID VERSUS NONHYBRID ORIGIN
Both the mtDNA tree and the nuclear tree consistently recover
the same clade: (L. lipetzi (L. reticulatum (L. flavimaculatum-
unisexual +L. flavimaculatum-bisexual)). This same topology is
also recovered (Bayesian PP =1.0, and ML bootstrap values ≥93)
from a much larger dataset (nine nuclear and four mitochondrial
genes, 7186 bp) in a phylogenetic study of the family Xantusiidae
(J. Pramuk, R. Bezy, B. Noonan, E. Sinclair, K. de Queiroz, and
J. Sites, unpubl. ms) and indicates that parthenogenesis most likely
evolved twice in the genus Lepidophyma: the first origin involving
the unisexual species L. reticulatum (Pacific Costa Rica), and
a later origin within L. flavimaculatum involving the unisexual
populations in Panama and southeastern Costa Rica.
The unisexual population of L. flavimaculatum in Panama is
not distinguishable from the bisexual populations that range from
northern Costa Rica to Guatemala by any markers in the mitochon-
drial and nuclear genes sequenced, but they do differ in that the
unisexual samples are homozygous at all 14 microsatellite loci.
To our knowledge, this is the first report of a complete lack of het-
erozygosity using microsatellite markers for any unisexual lizard,
and does not conform to expectations of the HOH. The Asher Ef-
fect (Asher 1970) or “decay” of heterozygosity to homozygosity
has been used to explain a complete absence of heterozygosity in
hybrids. It can occur under some forms of meiosis in partheno-
genetic species (under HOH), making it impossible to identify the
two parental species and hence to differentiate between an MOH
and HOH. However, if loss of alleles under the Asher Effect is ran-
dom with respect to parental origin, parthenogens of hybrid origin
should be fixed for alternative alleles from each of the parental
species at different (unlinked) nuclear loci, and therefore be in
strong linkage disequilibrium due to the association of “frozen”
alternative homozygous genotypes. This loss (in L. flavimacula-
tum unisexuals) via the decay mechanisms described by Asher
(1970) seems unlikely as we do not detect linkage disequilibrium
(see below). In such rapidly evolving markers as microsatellites,
high levels of heterozygosity are expected under an HOH. How-
ever, our findings are consistent with all expectations of the MOH
for a recent origin of unisexual L. flavimaculatum.Wefindnoevi-
dence of genomes derived from two genetically divergent parental
populations, as would be expected under the HOH. This result is
also in a stark contrast to studies in the unisexual lizard Darevskia
unisexualis (Tokarskaya et al. 2004; Badaeva et al. 2008), where
variation at a single (GATA)n microsatellite locus has arisen via
germline and somatic mutations in an unstable locus.
The picture for L. reticulatum is more complex as it involves
greater genetic variation, consistent with an earlier origin for this
unisexual species. It differs at between 103 and 125 mitochondrial
nucleotides (5–6% sequence divergence) and six nuclear substi-
tutions from L. flavimaculatum, but is consistently placed sister
to L. flavimaculatum. At the 14 microsatellite loci examined, a
total of 47 alleles are present in L. reticulatum, eight of which are
unique and presumably represent mutations that occurred within
this species. There are 24 alleles shared with L. flavimaculatum,
and these occur widely, rather than being restricted to any one
geographic population. The remaining 14 microsatellite alleles
are shared among other Lepidophyma species, and L. reticula-
tum does not share an unusually high number of alleles with any
one species other than L. flavimaculatum. The overall observed
heterozygosity for the Las Cruces population of L. reticulatum
(14.7%) is about half that of the bisexual populations of L. flavi-
maculatum (Table 1).
Given evidence for two spontaneous origins, the most ob-
vious difference between the two parthenogens is in variation
across the microsatellite loci; a complete absence of heterozygos-
ity within unisexual L. flavimaculatum from Panama compared
with 30.8% and 25.5% for the two bisexual populations of L.
flavimaculatum, and 14.7% for the unisexual L. reticulatum.This
is not attributed to reduced heterozygosity due to cross-species
amplification, as microsatellite loci for a single population of a
more distantly related species, L. sylvaticum, are highly polymor-
phic (overall heterozygosity =51.7%) for the same set of loci and
PCR amplification conditions.
DECAY OF HETEROZYGOSITY AND LOSS
OF HYBRID SIGNAL?
Heterozygosity in parthenogenetic individuals of hybrid origin
can be eliminated either by a terminal fusion in the absence of
recombination or the inhibition of meiosis II (Fig. 2 in Asher
1970), and will result in a homozygous individual. A second
mechanism, ameiotic recombination (Omilian et al. 2006), has
been shown to result in a low rate (<1.2%) of spontaneous loss of
heterozygosity (LOH) in asexual Daphnia lineages. This process
is expected to eliminate heterozygosity much faster than mutation
can replenish it, causing clonal lineages to lose allelic variation
8EVOLUTION 2009
DNA EVIDENCE FOR NONHYBRID ORIGINS OF PARTHENOGENESIS
over time (Omilian et al. 2006). The less likely third alternative
is biased gene conversion; a mechanism demonstrated for the
ribosomal DNA repeat complex in the hybrid unisexual gecko
Heteronotia binoei (Hillis et al. 1990). Yet, this parthenogen still
retained a signature-fixed heterozygosity at diagnostic allozyme
loci characteristic of vertebrate parthenogens.
The decay of heterozygosity and loss of a hybrid-origin ge-
netic signature by any of the above mechanisms could explain
our results if these processes were biased to preferentially elim-
inate one set of paternal alleles, but for several reasons we think
this is unlikely. First, we have no evidence for recombination in
nuclear sequences (see Methods), although the power of the tests
may be limited by low sequence divergence (<5.0% across the
genus). Second, if the 14 microsatellite loci used in this study
are unlinked, and maternal and paternal alleles assort randomly
at each locus in meiosis, the probability of losing only paternal
alleles would be 1/(2)14. The biased loss of all paternal alleles (by
elimination or conversion) might still be possible at a higher prob-
ability if most of the microsatellite markers are strongly linked.
The loci used in this study have not been mapped and sample
sizes generally preclude estimates of linkage disequilibrium, but
for the five largest samples, there was no evidence of linkage dise-
quilibrium for any combination of polymorphic loci (not shown).
The fixed homozygosity for unisexual populations of L. flavi-
maculatum at all microsatellite loci, the absence of evidence for
a paternal lineage in nuclear sequences, and the lack of evidence
for any mechanism of heterozygosity decay all favor the hypoth-
esis of a relatively recent mutational origin for these unisexual
populations.
The patterns of variation in the unisexual L. reticulatum also
suggest that none of the three heterozygosity-eliminating pro-
cesses have operated in this species, for the reasons just given.
However, unlike the completely homozygous unisexual popula-
tions of L. flavimaculatum, varying levels of heterozygosity are
observed at the microsatellite loci in L. reticulatum.Thetwo
unisexuals share about half of their alleles, many of the non-
flavimaculatum alleles present in L. reticulatum are shared with
other bisexual species, suggesting either that polymorphisms have
persisted through several speciation events, or that allelic homo-
plasy is common. Of special interest are the 10 private alleles
(from seven loci) that are exclusive to L. reticulatum.Twomutu-
ally compatible explanations for this pattern are that some alleles
are present in other species but were missed by limited sam-
pling, and/or that some have originated via subsequent mutation
events as clones have diversified. The absence of evidence for a
paternal lineage in this species, and the strong support in both
gene trees for its phylogenetic position as the sister species of
the (L. flavimaculatum – bisexual +L. flavimaculatum – uni-
sexual) clade, leads us to conclude that this species was derived
via spontaneous mutational origin prior to the origin of unisex-
ual L. flavimaculatum. This interpretation is consistent with the
clonal diversity observed in the microsatellite loci, and the modest
mtDNA and nuclear sequence divergence within this clade rela-
tive to the near absence of sequence divergence in the bisexual–
unisexual clade of L. flavimaculatum (Fig. 3).
TWO ORIGINS OR A REVERSAL?
We are interpreting our phylogenetic results as two independent
origins of unisexuality, but from a strict parsimony perspective,
the alternative is that there was a single loss of sexual reproduction
in the ancestor to the (L. reticulatum +L. flavimaculatum)clade,
followed by a reversal of this transition back to sexual reproduc-
tion in the ancestor of the bisexual L. flavimaculatum populations.
Another possibility is the origin of a “parthenogenesis mutation”
and its passage as a polymorphism through the L. reticulatum–
L. flavimaculatum speciation event, followed by fixation of the
mutation in L. reticulatum and its retention as a polymorphism in
L. flavimaculatum. These complex scenarios are consistent with
the following two observations. First, all populations of L. flav-
imaculatum have low levels of geographic divergence compared
to other species of Lepidophyma, suggesting that the species may
represent a recent range expansion of a unisexual lineage into
Central America, with a bisexual reversal in the northern part
of the range. Second, L. flavimaculatum and L.reticulatum are
largely allopatric, but a sympatric contact between bisexual L.
flavimaculatum and unisexual L. reticulatum does exist in the
Volcan Arenal region of Costa Rica; this provides potential for
exchange of genes between these two species, therefore retaining
polymorphism.
However, neither of these alternatives can satisfactorily ex-
plain the observed patterns of variation in all other markers, in the
context of the geographic distributions of all populations of both
species, without additional assumptions, including ad hoc shifts in
geographic range. An ecological observation of interest is that, in
contrast to the majority of organisms of hybrid origin which often
are associated with environments heavily influenced by Pleis-
tocene glacial cycles and characterized by weak biotic interac-
tions (Kearney 2005), the unisexual Lepidophyma show neither.
Both species are found in lowland tropical forests in lower Central
America where the role of Pleistocene climatic cycles was likely
restricted to elevational compression, and the species-rich ecosys-
tems are characterized by strong complex biotic interactions. If
the arguments of Kearney (2005) are correct, then Lepidophyma
represents an example in which parthenogenesis does not serve
to stabilize hybrid genotypes (which boost genetic variability and
thereby provide an advantage in recolonization), and the depau-
perate genetic diversity of these unisexuals must be considered in
another ecological context.
The presently available data cannot be used to resolve
the sequence of events within Lepidophyma. It is plausible to
EVOLUTION 2009 9
ELIZABETH A. SINCLAIR ET AL.
speculate on alternative hypotheses for the origin of unisexuality
in Lepidophyma as these become starting points for further
examination. The uniqueness of this system will allow direct
comparative studies between parthenogenetic and sexual lineages
in vertebrates. Limited information is available on the success
of Lepidophyma in captivity, however, there is an enormous
potential to develop this species into a model system for
understanding the mechanisms of naturally occurring nonhybrid
vertebrate parthenogenesis.
Conclusions
The interpretations of the genetic patterns described here, com-
pleteness of the taxonomic sampling, and the previous allozyme,
chromosome, and morphological data (summarized in Bezy 1989;
Bezy and Camarillo 2002) lead us to a strong inference that the
all-female populations of L. flavimaculatum and L. reticulatum
have been twice derived independently from L. flavimaculatum-
like populations, and that these derivations did not evolve via
hybridization. The microsatellite results presented here for uni-
sexual L. flavimaculatum are unique in that this is the first re-
port of a complete lack of heterozygosity in any naturally oc-
curring population of unisexual lizards. We have no evidence for
the mechanisms that might explain these transitions, but mech-
anisms recently described for aphids by Delmotte et al. (2001)
might be relevant. These are either a complete spontaneous loss
of males and of sexual reproduction through mutations along the
pathways leading to production of unisexual forms, or the deriva-
tion of unisexual lineages resulting from “contagious” transmis-
sion of alleles favoring parthenogenesis. There is some evidence
supporting the latter in the gradual “phasing out” (decrease in
frequency) of males in Costa Rican populations of L. flavimac-
ulatum near the Nicaraguan border and the contact of L. flavi-
maculatum with the unisexual L.reticulatum in this transitional
region.
The genetic architecture for microsatellites in the two unisex-
ual Lepidophyma stands in a stark contrast to that for allozymes
in the well-studied parthenogenetic species of lizards of the genus
Aspidoscelis. These latter are of hybrid origins having complete
sets of alleles from each parental species, resulting in heterozy-
gosity that is fixed and that is among the highest reported in
vertebrates (summaries in Moritz et al. 1989; Reeder et al. 2002).
The maintenance of fixed heterozygosity in parthenogenetic As-
pidoscelis is a result of their meiotic mechanism that involves
premeiotic endomitosis (Cuellar 1971). Whatever mechanism is
responsible for the presumed parthenogenesis in Lepidophyma, it
has resulted in a complete lack of microsatellite heterozygosity in
unisexual L. flavimaculatum, and low levels of heterozygosity in
the older L. reticulatum, presumably resulting from the gradual ac-
cumulation of mutations. Regardless of the mechanism involved,
it appears that stabilization of hybrid genotypes (Kearny 2005)
has not played a role in the transition from sexual to unisexual
reproduction in these tropical lizards.
The apparent lack of hybridization in the origin of unisex-
ual L. flavimaculatum bears importantly on the constraints of the
evolution of parthenogenesis in vertebrates. In the “balance hy-
pothesis” of Moritz et al. (1989), the chances of a hybridization
event resulting in the founding of a parthenogenetic lineage de-
pend on the hybrid-mediated disruption of meiosis resulting in
the production of unreduced oocytes, but with minimal reduction
in fecundity or viability of the hybrid. In the case of unisexual
L. flavimaculatum, it appears that the production of unreduced
oocytes likely occurs via mutation in genes regulating meiosis.
Further, the confounding effects of polyploidy appear to be ab-
sent as chromosomal, allozyme, and microsatellite data contain
no evidence of triploidy. The one exception is an unexplained
diploid/triploid mosaic in one L. flavimaculatum from Panama,
far removed from any known males (Bezy 1972).
The recent reports of “spontaneous” parthenogenetic events
in captive squamates might offer clues about MOH mechanisms;
automictic restitution leads to increased homozygosity, and in
the case of the Komodo dragon (Varanus komodoensis; with a
ZW sex-determining mechanism), all-male offspring (Watts et al.
2006). Although not conclusive, the case in the Burmese python
(Python molurus) appears to involve a clonal mechanism (Groot
et al. 2003). This example and our data for the night lizards
suggest that, contrary to all previous reports, nonhybrid routes to
parthenogenesis do occasionally succeed in vertebrates.
ACKNOWLEDGMENTS
Thanks to the many individuals who have aided with the fieldwork secur-
ing Lepidophyma tissues:K.Bezy,K.Bolles,J.Caldwell,J.Camarillo-R.,
J. Campbell, L. Canseco M´
arquez, J. Dixon, P. Elias, C. Lieb, J. McCranie,
L. Vitt, R. Vogt, B. Shaffer, J. Seems, E. Smith, W. Van Devender, L.
Wilson, and the members of the 1977 Tapezco Earth Watch Expedition.
Tissues were generously loaned by the Museo de Zoologia of the Univer-
sidad Nacional Aut´
onoma de M´
exico (MZFC), University of California
at Berkeley Museum of Vertebrate Zoology (MVZ), Collection of Ver-
tebrates of the University of Texas at Arlington (UTA-R), and the U.S.
National Museum of Natural History (USNM). Thanks to H. Shull and
R. Scholl for laboratory support. Thanks to the three anonymous review-
ers for their constructive comments on how to improve this manuscript.
This project was funded by the Los Angeles County Museum of Natural
History Foundation, Ralph J. Weiler Foundation, Brigham Young Uni-
versity, and the National Science Foundation DEB 0132227 to J.W. Sites,
Jr., and R.L. Bezy; and EF 0334966 to J. W. Sites, Jr.; Cricosaura tissue
collection was funded by NSF award DEB 9318642 to J. B. Losos, A.
Larson, and K. de Queiroz.
LITERATURE CITED
Asher, J. H. 1970. Parthenogenesis and genetic variability. II. One-locus mod-
els for various diploid populations. Genetics 66:369–391.
Barraclough, T. G., C. W. J. Birky, and A. Burt. 2003. Diversification in sexual
and asexual organisms. Evolution 57:2166–2172.
10 EVOLUTION 2009
DNA EVIDENCE FOR NONHYBRID ORIGINS OF PARTHENOGENESIS
Barton, N. H., and B. Charlesworth. 1998. Why sex and recombination?
Science 281:1986–1990.
Beukeboom, L., and R. C. Vrijenhoek. 1998. Evolutionary genetics and
ecology of sperm-dependent parthenogenesis. J. Evol. Biol. 11:755–
782.
Bezy, R. L. 1972. Karyotypic variation and evolution of the lizards in the
family Xantusiidae. Contrib. Sci. LACM 27:1–29.
———. 1989. Morphological differentiation in unisexual and bisexual Xan-
tusiid lizards of the genus Lepidophyma in Central America. Herp.
Monogr. 3:61–80.
Bezy, R. L., and J. L. Camarillo. 2002. Systematics of xantusiid lizards of the
genus Lepidophyma. Contrib. Sci. LACM 493:1–41.
Bezy, R. L., and J. W. Sites. 1987. A preliminary study of allozyme evolution
in the lizard family Xantusiidae. Herpetologica 43:280–292.
Badaeva, T. N., D. N. Malysheva, V. I. Korchagin, and A. P. Ryskov.
2008. Genetic variation and De Novo mutations in the partheno-
genetic Caucasian rock lizard Darevksia unisexualis. PLoS ONE 3(7):
e2730.
Brumfield, R. T., P. Beerli, D. A. Nickerson, and S. V. Edwards. 2003. The
utility of single nucleotide polymorphisms in inferences of population
history. Trends Ecol. Evol. 18:249–256.
Bullini, L. 1994. Origin and evolution of animal hybrid species. Trends Ecol.
Evol. 9:422–426.
Canseco-M´
arquez, L., G. Guti´
errez-Mayen, and A. A. Mendoza-Hern´
andez.
2008. A new species of night-lizard of the genus Lepidophyma (Squa-
mata: Xantusiidae) from the Cuicatl´
an Valley, Oaxaca, M´
exico. Zootaxa
1750:59–67.
Cuellar, O. 1971. Reproduction and the mechanism of meiotic restitution in the
parthenogenetic lizard Cnemidophorus uniparens. J. Morph. 133:139–
165.
———. 1987. The evolution of parthenogenesis: a historical perspective.
Pp. 43–104. in P. B. Moens ed. Meiosis. Academic Press, New York.
Dawley, R. M., and J. P. Bogart. 1989. Evolution and ecology of unisexual
vertebrates. New York State Museum, New York.
Delmotte, F., N. Leterme, J. Bonhomme, C. Rispe, and J.-C. Simon. 2001.
Multiple routes to asexuality in an aphid species. Proc. R. Soc. Lond. B
268:2291–2299.
Fontaneto, D., E. A. Herniou, C. Boschetti, M. Caprioli, G. Melone, C. Ricci,
and T. G. Barraclough. 2007. Independently evolving species in asexual
rotifers. PLoS Biol. 5:0914–0921.
Groot, T. V. M., E. Bruins, and J. A. J. Breeuwer. 2003. Molecular genetic
evidence for parthenogenesis in the Burmese Python, Python molurus
bivittatus. Heredity 90:130–135.
Hillis, D. M., C. Moritz, C. A. Porter, and R. J. Baker. 1990. Evidence
for biased gene conversion in concerted evolution of ribosomal DNA.
Science 251:308–310.
Holm, S. 1979. A simple sequentially rejective multiple test procedure. Scand.
J. Stat. 6:65–70.
Huelsenbeck, J. P., and F. Ronquist. 2001. MRBAYES: Bayesian inference of
phylogenetic trees. Bioinformatics 17:754–755.
Johnson, S. G., and W. R. Leefe. 1999. Clonal diversity and polyphyletic
origins of a hybrid and spontaneous parthenogenetic Campeloma (Gas-
tropoda: Viviparidae). J. Evol. Biol. 12:1056–1068.
Kearney, M. 2005. Hybridization, glaciation and geographical parthenogene-
sis. Trends Ecol. Evol. 20:495–502.
Kearney, M., M. K. Fujita, and J. Ridenour. 2009. Lost sex in the reptiles:
constraints and correlations. Pp. 447–474. in I. Sch¨
on, K. Martens and
P. van Dijk, eds. Lost sex: the evolutionary biology of parthogenesis.
Sprinter Scientific, Dordrecht, Netherlands.
Kondrashov, A. S. 1993. Classification of hypotheses on the advantage of
amphimixis. J. Hered. 84:372–387.
Korchagin,V.I.,T.N.Badaeva,O.N.Tokarskaya,I.A.Mar-
tirosyan, I. S. Darevsky, and A. P. Ryskov. 2007. Molecular char-
acterization of allelic variants of (GATA)n microsatellite oci in
parthenogenetic lizards Darevskia unisexualis (Lacertidae). Gene 392:
126–133.
Leavitt, D. H., R. L. Bezy, K. A. Crandall, and J. W. Sites. 2007. Multi-
locus DNA sequence data reveal a history of deep cryptic vicariance
and habitat-driven convergence in the desert night lizard Xantusia vig-
ilis species complex (Squamata: Xantusiidae). Mol. Ecol. 16:4455–
4481.
Martin, D. P., C. Williamson, and D. Posada. 2005. RDP2: recombination de-
tection and analysis from sequence alignments. Bioinformatics 21:260–
262.
Moritz, C. 1987. Parthenogenesis in the tropical gekkonid lizard, Nactus
arnouxii (Sauria: Gekkonidae). Evolution 41:1252–1266.
———. 1993. The origin and evolution of parthenogenesis in the Heteronotia
binoei complex-synthesis. Genetica 90:269–280.
Moritz, C., W. M. Brown, L. D. Densmore, J. W. Wright, D. Vyas, S.
Donnellan, M. Adams, and P. Baverstock. 1989. Genetic diversity and
the dynamics of hybrid parthenogenesis in Cnemidophorus (Teiidae)
and Heteronotia (Gekkonidae), Pp. 87–112. in R. M. Dawley and J. P.
Bogart, eds. Evolution and ecology of unisexual vertebrates. New York
State Museum, New York.
Murphy, R. W., J. Fu, R. D. MacCulloch, I. S. Darevsky, and L. A.
Kupriyanova. 2000. A fine line between sex and unisexuality: the phylo-
genetic constraints on parthenogenesis in lacertid lizards. Zool. J. Linn.
Soc. 130:527–549.
Omilian, A. R., M. E. A. Cristescu, J. L. Dudycha, and M. Lynch. 2006.
Amneiotic recombination in asexual lineages of Daphnia. Proc. Nat.
Acad. Sci. USA 103:18638–18643.
Peakall, R., and P. E. Smouse. 2006. Genalex 6: genetic analysis in Excel.
Population genetic software for teaching and research. Mol. Ecol. Notes
6:288–295.
Posada, D., and K. A. Crandall. 1998. Modeltest: testing the model of DNA
substitution. Bioinformatics 14:817–818.
———. 2001. Evaluation of methods for detecting recombination from DNA
sequences: computer simulations. Proc. Natl. Acad. Sci. USA 98:13757–
13762.
Raymond, M., and F. Rousset. 1995. GENEPOP (version 1.2): population
genetics software for exact tests and ecumenicism. J. Hered. 86:248–
249.
Reeder, T. W., C. J. Cole, and H. C. Dessauer. 2002. Phylogenetic relationships
of whiptail lizards of the genus Cnemidophorus (Squamata: Teiidae): a
test of monophyly, revaluation of karyotypic evolution, and review of
hybrid origins. Am. Mus. Nov. 3365:1–61.
Shimodaira, H., and M. Hasegawa. 1999. Multiple comparisons of log like-
lihoods with applications to phylogenetic inference. Mol. Biol. Evol.
16:1114–1116.
Simon, J.-C., F. Delmotte, C. Rispe, and T. Crease. 2003. Phylogenetic re-
lationships between parthenogens and their sexual relatives: the possi-
ble routes to parthenogenesis in animals. Biol. J. Linn. Soc. 79:151–
163.
Sinclair, E. A., R. L. Bezy, J. Camarillo, K. Bolles, K. A. Crandall, and
J. W. Sites, Jr. 2004. Testing species boundaries in an ancient species
complex with deep phylogeographic history: genus Xantusia (Squamata:
Xantusiidae). Am. Nat. 164:396–414.
Sinclair, E. A., R. Scholl, R. L. Bezy, K. A. Crandall, and J. W.
Sites. 2006. Isolation and characterization of di- and tetranucleotide
microsatellite loci in the yellow-spotted night lizard Lepidophyma
flavimaculatum (Squamata: Xantusiidae). Mol. Ecol. Notes 6:233–
236.
EVOLUTION 2009 11
ELIZABETH A. SINCLAIR ET AL.
Sites, J. W.J., D. M. Peccinini-Seale, C. Moritz, J. W. Wright, and W. P. Brown.
1990. The evolutionary history of parthenogenetic Cnemidophorus lem-
niscatus (Sauria, Teiidae) I. Evidence for a hybrid origin. Evolution
44:906–921.
Song, H., J. E. Buhay, M. F. Whiting, and K. A. Crandall. 2008. Many
species in one: DNA barcoding overestimates the number of species
when nuclear mitochondrial pseudogenes are coamplified. Proc. Natl.
Acad. Sci. USA 105:13486–13491.
Suomalainen, E., A. Saura, and J. Lokki. 1987. Cytology and the evolution of
parthenogenesis. CRC Press, Boca Raton, FL.
Swofford, D. L. 2002. PAUP∗: Phylogenetic analysis using parsimony
(∗and other methods). Ver. 4.0b10. Sinauer Associates, Sunderland,
MA.
Tokarskaya, O. N., I. A. Martirosyan, T. N. Badaeva, D. N. Malysheva, V. I.
Korchagin, I. S. Darevsky, F. D. Danielyan, and A. P. Ryskov. 2004. In-
stability of (GATA)nmicrosatellite loci in the parthenogenetic Caucasian
rock lizard Darevskia unisexualis (Lacerticae). Mol. Genet. Genom.
270:509–513.
Vicario, S., A. Caccone, and J. Gauthier. 2003. Xantusiid ‘night’ lizards: a puz-
zling phylogenetic problem revisited using likelihood-based Bayesian
methods on mtDNA sequences. Mol. Phylogenet. Evol. 26:243–261.
Vrijenhoek, R. C. 1994. Unisexual fish: models for studying ecology and
evolution. Ann. Rev. Ecol. Syst. 25:71–96.
———. 1998. Animal clones and diversity. BioScience 48:617–628.
Vrijenhoek, R. C., R. M. Dawley, C. J. Cole, and J. P. Bogart. 1989. Pp. 19–
23 in R. M. Dawley and J. P. Bogart, eds. Evolution and ecology of
unisexual vertebrates. New York State Museum, New York.
Watts,P. C., K. R. Buley, S. Sanderson, W. Boardman, C. Claudio Ciofi, and R.
Gibson. 2006. Parthenogenesis in Komodo dragons. Nature 444:1021–
1022.
Associate Editor: J. Huelsenbeck
Supporting Information
The following supporting information is available for this article:
Appendix S1. List of taxa Lepidophyma sampled and outgroup species, with location, collection numbers, and Genbank accession
numbers.
Appendix S2. List of Lepidophyma individuals, their locations, museum voucher numbers, genotypes for 14 microsatellite loci,
and bases at seven nucleotide positions that were heterozygous in L. sylvaticum.
Appendix S3. List of 10 mitochondrial and nuclear genes, primer sequences, and source used in this study.
Supporting Information may be found in the online version of this article.
(This link will take you to the article abstract).
Please note: Wiley-Blackwell is not responsible for the content or functionality of any supporting information supplied by the
authors. Any queries (other than missing material) should be directed to the corresponding author for the article.
12 EVOLUTION 2009
