Molecular Phylogenetics and Evolution 91 (2015) 56–67
Contents lists available at ScienceDirect
Molecular Phylogenetics and Evolution
journal homepage: www.elsevier.com/locate/ympev
Multilocus sequence data reveal dozens of putative cryptic species
in a radiation of endemic Californian mygalomorph spiders
(Araneae, Mygalomorphae, Nemesiidae) q
Dean H. Leavitt a,⇑, James Starrett a, Michael F. Westphal b, Marshal Hedin a
a
b
Department of Biology, San Diego State University, San Diego, CA 92182, United States
Bureau of Land Management, Hollister Field Office, Hollister, CA 95023, United States
a r t i c l e
i n f o
Article history:
Received 6 March 2015
Revised 11 May 2015
Accepted 19 May 2015
Available online 27 May 2015
Keywords:
DNA barcoding
Cryptic species
Genealogical concordance
Species tree
Sympatry
a b s t r a c t
We use mitochondrial and multi-locus nuclear DNA sequence data to infer both species boundaries and
species relationships within California nemesiid spiders. Higher-level phylogenetic data show that the
California radiation is monophyletic and distantly related to European members of the genus
Brachythele. As such, we consider all California nemesiid taxa to belong to the genus Calisoga
Chamberlin, 1937. Rather than find support for one or two taxa as previously hypothesized, genetic data
reveal Calisoga to be a species-rich radiation of spiders, including perhaps dozens of species. This conclusion is supported by multiple mitochondrial barcoding analyses, and also independent analyses of
nuclear data that reveal general genealogical congruence. We discovered three instances of sympatry,
and genetic data indicate reproductive isolation when in sympatry. An examination of female reproductive morphology does not reveal species-specific characters, and observed male morphological differences for a subset of putative species are subtle. Our coalescent species tree analysis of putative
species lays the groundwork for future research on the taxonomy and biogeographic history of this
remarkable endemic radiation.
Ó 2015 Elsevier Inc. All rights reserved.
1. Introduction
Mygalomorph spiders, the radiation that includes tarantulas
and trapdoors spiders, are ancient and diverse. Because of general
dispersal limitation and high levels of regional endemism, several
research groups from different continents have developed mygalomorphs as models for studies of species delimitation, cryptic speciation, and historical biogeography (e.g., Bond and Stockman,
2008; Satler et al., 2013; Hedin et al., 2013; Hendrixson et al.,
2013; Castalanelli et al., 2014; Hamilton et al., 2014; Opatova
and Arnedo, 2014a). Still, many features of mygalomorph biology
make them challenging to study. These spiders are largely subterranean and spend much of their lives out of sight; also, much of
their diversity is obfuscated by conserved morphology across both
species and higher taxonomic groups. Raven (1985) used morphology to revise taxonomy and contributed greatly to our understanding of higher-level mygalomorph phylogenetic relationships.
Recent phylogenetic studies using DNA sequence data have
q
This paper has been recommended for acceptance by Marcos Perez-Losada.
⇑ Corresponding author.
E-mail address: [email protected] (D.H. Leavitt).
http://dx.doi.org/10.1016/j.ympev.2015.05.016
1055-7903/Ó 2015 Elsevier Inc. All rights reserved.
recovered support for some previous morphology-based hypotheses, but have also demonstrated the limitations of morphological
data and revealed some surprising relationships (Hedin and
Bond, 2006; Bond et al., 2012; Starrett et al., 2013; Opatova
et al., 2013; Opatova and Arnedo, 2014a).
In addition to improving our understanding of deeper phylogenetic relationships, DNA sequence data have played a crucial role
in uncovering and recognizing species-level diversity in mygalomorphs. Recent genetic studies have revealed ancient but morphologically cryptic species diversity in various mygalomorph groups
(Starrett and Hedin, 2007; Stockman and Bond, 2007; Hedin and
Carlson, 2011; Graham et al., 2015). Thus far, most genetic studies
have relied principally on one or two fragments of mitochondrial
DNA. Theory has long emphasized the shortcomings of relying on
a single marker (Maddison, 1997), and empirical data increasingly
demonstrate that inferences based on mtDNA alone can be misleading (e.g., McGuire et al., 2007). For example, the mitochondrial
genome is particularly prone to asymmetric introgression across
species boundaries (Schwenk et al., 2008), leading to both misunderstanding species boundaries and gene tree/species tree incongruence if mtDNA is the sole locus employed (e.g., Harrington
and Near, 2012). The need for accompanying nuclear markers is
D.H. Leavitt et al. / Molecular Phylogenetics and Evolution 91 (2015) 56–67
generally well appreciated, and while nuclear DNA data have
become standard in many taxonomic groups, thus far only a few
recent studies of mygalomorphs have included multiple nuclear
loci (e.g., Satler et al., 2013; Hedin et al., 2013; Opatova and
Arnedo, 2014b).
The mygalomorph family Nemesiidae has a global distribution,
with particularly high generic and species diversity in Australia
and South America (Raven, 1985; Goloboff, 1995; World Spider
Catalog, 2015). While recent phylogenetic data reveal uncertainty
in the composition of the family (Bond et al., 2012), North
American nemesiid diversity is known to be depauperate relative
to other continental regions. Mexico is home to Mexentypesa chiapas (Raven, 1987) and at least one undescribed species (Hedin and
Bond, 2006). Other North American representatives include five
nominal species from California: Brachythele longitarsis Simon
1891, B. anomala Schenkel 1950, Calisoga theveneti Simon 1891,
Calisoga sacra Chamberlin 1937 and Calisoga centronetha
Chamberlin and Ivie, 1939 (World Spider Catalog, 2015).
In his unpublished dissertation, Bentzien (1976) found
California Brachythele to be morphologically distinct from
European members of that genus and assigned all California species to the genus Calisoga. These spiders are robust, aggressive spiders that reside in subterranean burrows with ‘‘simple’’ open
entrances, similar to certain tarantulas. Based on statistical analyses of morphology, Bentzien (1976) hypothesized that there are
only two species of Calisoga in California: the common, widespread
‘‘C. longitarsis’’ and the smaller, rare C. theveneti. However, he also
considered the possibility that spiders identified as C. theveneti
are merely ‘‘C. longitarsis’’ males that mature at an earlier age,
and that all populations represented a single species. Despite the
fact that these results were never published, many arachnologists
have come to use the combination ‘‘Calisoga longitarsis’’ to refer
to all Californian nemesiids (e.g., Ubick et al., 2005). California
nemesiids are one of the more commonly encountered mygalomorphs in central and northern California, particularly during the
fall when wandering males enter houses in the San Francisco Bay
area. That the binomial most commonly applied to this taxon –
‘‘Calisoga longitarsis’’ – remains unpublished and technically
unavailable (World Spider Catalog, 2015) underscores the need
for systematic attention. In this study we employ multi-locus
DNA sequence data to examine the phylogenetic relationship of
California nemesiids relative to Old World Brachythele and to infer
both species boundaries and phylogenetic relationships within the
California radiation. This robust phylogenetic hypothesis provides
the framework for future biogeographic studies and formal morphological taxonomic revision.
2. Methods
57
the known distribution both to the north and the southwest. We
attempted to collect 3–5 individuals per locality, although some
localities yielded fewer individuals. In total, this study includes
genetic data for 197 California nemesiid specimens from 106 distinct localities (Fig. 1). Whole spiders were vouchered and deposited in the San Diego State University Terrestrial Arthropods
Collection, and typically one leg was cryopreserved for subsequent
genetic work. DNA was extracted from ethanol-preserved tissue
using the Qiagen DNeasy kit.
2.2. Molecular protocols
For the higher-level phylogenetic analysis of Nemesiidae, we
downloaded published data from Genbank for 28S and elongation
factor 1 c (EF-1c), nuclear loci that are widely used in phylogenetic
analyses of mygalomorphs (Hedin and Bond, 2006; Ayoub et al.,
2007; Bond et al., 2012; Opatova et al., 2013). We also included
additional data from transcriptomes and supplementary PCR reactions (see Supplementary Material, which also includes 28S and
EF-1c primers). All PCR products were purified with Millipore
Multiscreen PCR filter plates and Sanger sequenced in both directions by Macrogen, USA.
Using a published ‘‘Calisoga longistarsis’’ mitochondrial genome
(Masta and Boore, 2008), we modified universal primers to amplify
a region of the cytochrome oxidase I gene (COI) comprising over
1200 base pairs (bp). In addition to the primers used in amplification, we sometimes used internal sequencing primers
(Supplementary Material). The COI fragment was targeted for all
Californian samples included in the study. Additionally, an approximately 1000 bp mitochondrial region that contained partial 12S,
val-tRNA and 16S regions was sequenced for a subset of
Californian specimens.
Based on preliminary analyses of mitochondrial data and geographic distribution, we collected nuclear data for forty individuals. We amplified the internal transcribed spacer (ITS) units 1
and 2 using published primers (Ji et al., 2003). Additional nuclear
loci were developed from genomic resources generated specifically
for Californian samples using comparative transcriptomic data (see
Thomson et al., 2010) derived from next-generation sequencing. A
transcriptome was characterized for a spider from Panoche Hills
(Fresno Co.), and compared to an already published transcriptome
from Jasper Ridge (San Mateo Co., SRX652491; Bond et al., 2014).
Methodological details (including phasing of data) are similar to
those used in Derkarabetian and Hedin (2014) and are further outlined in the Supplementary Material. From this effort we selected
seven exonic loci that amplified consistently across a geographic
test panel. We performed PCR reactions with typical reagents,
and primers are available in Supplementary Material. Purified
nuclear loci were Sanger sequenced using forward and reverse primers by Macrogen USA.
2.1. Taxon sampling and data collection
2.3. Higher-level alignment and phylogenetic analyses
We tested the distinctiveness of Californian nemesiids relative
to the Palearctic Brachythele by performing molecular phylogenetic
analyses of available nemesiid taxa. Using published data and
newly sampled spiders, we compiled a data set including multiple
outgroup taxa representing the families Cyrtaucheniidae and
Ctenizidae as well as 21 nemesiids and allied taxa
(Microstigmatidae and ‘‘Cyrtaucheniidae’’). This sample also
included multiple representatives of both European Brachythele
and Californian spiders currently referred to as ‘‘Calisoga longitarsis’’ (see Supplementary Material for specimen information).
To investigate species diversity within Californian nemesiids,
we sampled spiders from throughout the known distribution in
central and northern California (following Bentzien, 1976). Our
sampling effort also included original collections that extended
An initial alignment of EF-1c was performed in Sequencher and
manually adjusted after translating codons into amino acids. To
align the ribosomal 28S fragment, we used the option Q-INS-i in
MAFFT v7 (Katoh et al., 2005) as it considers RNA secondary structure during alignment. The default settings include a gap-opening
penalty of 1.53 and the offset value equal to 0.0. We used the program GBLOCKS v0.91b (Castresana, 2000) to identify and exclude
regions that could not be reliably aligned. Settings for this analysis
allowed gap positions within the final blocks as well as smaller
final blocks. EF-1c and 28S data were concatenated for phylogenetic analyses. The program jModeltest (Posada, 2008) was used
to aid in selection of nucleotide models for these and subsequent
phylogenetic analyses. We performed a maximum-likelihood
58
D.H. Leavitt et al. / Molecular Phylogenetics and Evolution 91 (2015) 56–67
Fig. 1. Map of central and northern California showing collecting localities included in this study. Colors and clade names correspond to ten major biogeographic clades
recovered in mitochondrial gene tree analyses. Localities where divergent mitochondrial and nuclear lineages were found in sympatry are indicated in white. Black x’s
designate collection locations for CAS male specimens. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this
article.)
analysis using the RAxML-HPC BlackBox 8.0.24 (Stamatakis et al.,
2008) on the CIPRES Science Gateway (Miller et al., 2011). A
GTRgamma model was used for each of four partitions: three for
corresponding codon positions within EF-1c and a fourth for the
28S fragment. The sufficient number of bootstrap replicates was
automatically detected by the analysis. We consider bootstrap values >70 as evidence for high support (Hillis and Bull, 1993; Alfaro
et al., 2003). For a Bayesian BEAST analysis, we used the same
partitioning scheme and nucleotide models (GTR + G), except
HKY + G models were used for the three EF-1c partitions to achieve
convergence in certain estimated parameters. This analysis
employed a lognormal relaxed clock for each of the two gene
regions and a tree prior of birth–death with incomplete sampling.
The BEAST analysis was run for 100 million generations saving
every 4000th. We used TRACER v.1.5 (Rambaut and Drummond,
2007) to plot likelihood scores in order to assess convergence, to
D.H. Leavitt et al. / Molecular Phylogenetics and Evolution 91 (2015) 56–67
identify the appropriate burn-in length, and to verify that all effective sample size (ESS) values were well over 200.
2.4. Mitochondrial gene tree, species delimitation and DNA barcoding
Mitochondrial COI data were translated into amino acids to verify an open reading frame. With no insertions or deletions, the COI
alignment was trivial and performed in Sequencher. For the ribosomal 12S/16S data, we performed analyses with MAFFT and
GBLOCKS v0.91b with the same settings as outlined above. To infer
the mitochondrial gene tree, we concatenated the two fragments.
Similar to higher-level phylogenetic analyses, likelihood and
Bayesian analyses contained four partitions: three for the COI
codon positions and a fourth for the ribosomal fragment. We performed maximum-likelihood analysis in RAxML with the model
GTRgamma for each partition and automatic bootstrapping. For
the BEAST analysis, the use of HKY + G models for each of the four
partitions was necessary to achieve convergence. A strict clock was
used for the COI fragment and a lognormal clock for the 12S fragment, with the 12S rate estimated relative to the COI rate of 1.0
substitution per site. We used a birth–death and incomplete sampling tree prior.
We used three different approaches to DNA barcoding to identify putative species limits. The first method was the traditional
barcoding approach (Hebert et al., 2003a, 2003b) where genetic
distances are calculated and a threshold value that designates distinct species status is assigned. The divergences among relevant
sister clades recovered from the BEAST mtDNA gene tree were calculated using the uncorrected p-distance similar to recently published barcoding studies of mygalomorphs (Castalanelli et al.,
2014; Hamilton et al., 2014). To be able to directly compare the
results of this system to the aforementioned studies, we identified
putative species using the threshold values of both 9.5%
(Castalanelli et al., 2014) and 5% (Hamilton et al., 2014). The 9.5%
value was used to characterize species diversity of mygalomorphs
within the Pilbara region of Australia (the majority of which
belonged to the family Nemesiidae), and the 5% cutoff was used
for species discovery within the North American tarantula genus
Aphonopelma sampled primarily from the United States.
The third barcoding approach used was the Automatic Barcode
Gap Discovery (ABGD) method (Puillandre et al., 2012), implemented on the online ABGD server (http://www.abi.snv.jussieu.
fr/public/abgd/). This analysis uses the ordered, ranked genetic distances from COI sequences to identify marked shifts from low
intraspecific distances to higher interspecific distances. These
observed transitions in genetic distances represent barcode gaps.
Gap detection is an iterative process that assigns sampled individuals into increasingly partitioned groups. We used the same
parameters employed in the tarantula study (Hamilton et al.,
2014) for comparative purposes, as follows: Pmin = 0.0001,
Pmax = 0.200, Steps = 10, X = 1, Nb bins = 20. Hamilton et al.
(2014) found that species delimitation with this approach appears
to be fairly robust to varying parameter settings.
2.5. Nuclear gene genealogical concordance
Species delimitation using a single mitochondrial marker can be
misled due to various processes, including both incomplete lineage
sorting and introgression. Additionally, deep mitochondrial structure is not always accompanied with correspondingly deep nuclear
differentiation (Ogden and Thorpe, 2002), and dispersal-limited
taxa (e.g., mygalomorph spiders) in particular may show spurious
mtDNA breaks (Irwin, 2002). On the other hand, genealogical concordance among multiple sampled loci can provide convincing evidence for genetic boundaries (Avise and Ball, 1990; Kuo and Avise,
2005), and thus multi-locus data can validate the presence of
59
genetically distinctive but morphologically cryptic lineages
(Satler et al., 2013).
We evaluated nuclear data separately from mitochondrial evidence to provide an independent test of putative species identified
from mtDNA barcoding. We performed a species tree analysis
where each of the forty individuals with sampled nuclear data
was treated as a distinct OTU. Identified mtDNA barcode species
are considered validated by the nuclear species tree analysis when
individuals are recovered in a clade with other members of that
species (when multiple individuals sampled), with non-trivial
branch lengths separating a clade/individual from their respective
sister clade. Species tree inference was performed using *BEAST
with BEAST 1.8. Because initial analyses using more complex GTR
models failed to converge, all eight phased nuclear loci were analyzed with HKY + G models (Grummer et al., 2014) and a lognormal
relaxed clock. The ITS rate was set to 1.0 and the rates of other loci
were estimated relative to that of ITS. An inverse gamma shape was
used for the species population mean prior, and an exponential
shape was used for the Yule birth–death prior. Analyses were run
for 200 million generations with every 5000th generation saved.
2.6. Combined species tree analyses
We performed a combined species tree analysis of the nuclear
and mitochondrial data similar to the *BEAST analysis of the
nuclear-only data. OTUs for this analysis were mtDNA barcode species identified at the 9.5% threshold. In addition, in three instances
individuals from divergent clades just under this threshold (9%)
were treated as distinct taxa (species Q was divided into Q1 and
Q2, H into clades H and I, and F into F1 and F2). Mitochondrial barcode species that lacked nuclear data were excluded in this analysis. As with other *BEAST analyses of these data, all loci were
analyzed with an HKY + G model of sequence evolution. The COI
fragment was partitioned by codon position, and a strict clock
was employed for this gene region alone. Other gene regions used
lognormal relaxed clocks whose rates were estimated relative to
that of COI. Other priors were the same as the nuclear-only analysis, with analyses run for 200 million generations with every
5000th generation saved.
2.7. Morphological data
The morphology of adult male pedipalps and first legs (which
possess mating spurs) is often used as a primary character for species delimitation in mygalomorph spiders, including nemesiids
(e.g., Goloboff, 1995; Decae and Cardoso, 2005; Decae et al.,
2014). To assess qualitative morphological divergence across
Californian populations, adult males (N = 9) were borrowed from
the California Academy of Sciences (the same males studied by
Bentzien, 1976), and augmented with males (N = 9) from our
own collections. Also, mature spermathecal organs for all adult
females included in our molecular study were dissected and examined (N = 103). Specimens were imaged using a Visionary Digital
BK Plus system (http://www.visionarydigital.com), including a
Canon 5D digital camera, Infinity Optics Long Distance
Microscope, P-51 camera controller, and FX2 lighting system.
Individual images were combined into a composite image using
Helicon Focus V5.3, and then edited using Adobe Photoshop CS6.
3. Results
3.1. Higher-level phylogenetic placement
The EF-1c data included for phylogenetic analysis consists of
924 aligned base pairs. The MAFFT alignment of the 28S fragment
60
D.H. Leavitt et al. / Molecular Phylogenetics and Evolution 91 (2015) 56–67
resulted in 1964 aligned bp, and the GBLOCKS analysis reduced
this length to 1812 bp. All sampled taxa were represented by the
EF-1c fragment; three individuals lacked the 28S fragment but
were closely related to taxa with complete data. Of the total
2736 aligned characters, 665 were variable and 440 were parsimony informative.
Maximum-likelihood and Bayesian BEAST analyses recovered
similar topologies, differing only in the placement of the clade containing Hermacha sp. and the undescribed taxon from Ngome
Forest, South Africa. The topology recovered from the BEAST analysis is reported in Fig. 2. The sampled Californian spiders are recovered with high support in a clade containing representatives of two
Mediterranean genera, Iberesia and Ambylocarenum. In the RAxML
analysis, the Ambylocarenum sp. sample from Spain is recovered
as sister to Californian spiders with a high support (BS = 94); this
relationship is also recovered in the Bayesian BEAST analysis but
with lower support (PP = 0.66). The European Brachythele is weakly
recovered as sister to Fufius sp., and this clade is recovered as sister
to the aforementioned Iberesia + Amblyocarenum + Californian
clade, also with weak support. These genera are separated by
branch lengths that are long relative to crown Nemesiidae (Fig. 2).
3.2. Mitochondrial gene tree and DNA barcoding
The COI matrix included data for 194 spiders that we sampled
(Supplementary Material), plus COI and ribosomal mtDNA from a
published mitochondrial genome (Masta and Boore, 2008). The
aligned length of the COI fragment was 1263 base pairs; 560 positions were variable, of which 515 were parsimony informative. The
mitochondrial ribosomal fragment was obtained for 55 spiders.
The mtDNA ribosomal alignment after MAFFT and GBLOCKS
analyses consisted of 986 bp; 519 characters were variable with
433 of those being parsimony informative.
Likelihood and Bayesian analyses recovered similar gene tree
topologies; the Bayesian BEAST topology is depicted in Fig. 3. The
partitioned HKY + G model with the strict molecular clock as
employed in BEAST indicates that deepest corrected divergences
among sampled California specimens are 22%. The mtDNA gene
tree is composed of divergent, geographically cohesive clades that
are allopatric or parapatric to neighboring clades with few exceptions. We arbitrarily name ten major clades with reference to their
geographical distribution (Fig. 1, Supplementary Material). These
clades include: Bay Area (locations 1–29), Sacramento Valley
(30–36), Northern Sierra (37–53), Calaveras (50, 54), Tuolumne
(55–58), Coulterville (59–63), Central Sierra (60, 64–74), Southern
Sierra (75–78), Monterey (1, 79–101), and Panoche (l02–106).
We found sympatry of divergent mitochondrial lineages at
three localities (Fig. 1). Bay Area and Monterey clades were found
at locality 1 (Coalinga Road), Northern Sierra and Calaveras clades
at locality 50 (west of Sheep Ranch), and Coulterville and Central
Sierra clades at locality 61 (west of Coulterville). At almost all other
localities with multiple sampled individuals, COI sequences were
either identical or minimally divergent. An exception is locality
45 (east of Mount Aukum), where the two sampled COI sequences
were just over 6% divergent.
Using the traditional barcoding approach with a 9.5% threshold,
26 putative species are identified (Fig. 3, Supplementary Material).
Most of the major geographic clades previously referenced are
composed of multiple species at this cutoff, but Tuolumne,
Calaveras and Coulterville clades each correspond to a single species (Fig. 3). Using the 5% threshold (similar to Hamilton et al.,
2014), the number of species identified more than doubles to 59
species (Fig. 3, Supplementary Material). Of the aforementioned
Fig. 2. Maximum clade credibility tree from the BEAST analysis of concatenated EF-1c and 28S data. Clade of sampled Nemesiidae and allied taxa is indicated. Support values
at nodes represent Bayesian posterior probabilities first, followed by ML bootstrap values.
61
D.H. Leavitt et al. / Molecular Phylogenetics and Evolution 91 (2015) 56–67
Fig. 3. Maximum clade credibility tree from the mtDNA BEAST analysis inferred from concatenated COI and ribosomal fragments. Nodal support is shown, with Bayesian
posterior probabilities first and ML bootstraps second. For simplicity, individual haplotypes have been collapsed into clades corresponding to putative species at the 5%
threshold. Clades are also labeled according to recovered species at the 9.5% threshold. The sampled localities comprising each clade are listed. Major geographic clades are
also labeled (see Fig. 1).
Table 1
Nuclear loci summary information.
Primer combo
Annotation
B1_B2
D1_D3
D6_D7
E10_E11
F9_F10
G9_G10
H1_H2
CAS18sF1 CAS5p8sB1d
Orthologous to
Orthologous to
Orthologous to
Orthologous to
Orthologous to
Orthologous to
Orthologous to
ITS ribosomal
Ixodes
Ixodes
Ixodes
Ixodes
Ixodes
Ixodes
Ixodes
ISCW003568
ISCW012830
ISCW013467
ISCW019432
ISCW020841
ISCW024803
ISCW024894
(conserved hypothetical protein)
(60S ribosomal protein L3)
(conserved hypothetical protein)
(ccaat enhanced binding protein)
(zinc finger protein)
(60S ribosomal protein L23)
(acetylcholinesterase)
# Seq.
Length bp
# Inform.
# Uninform
40
38
39
40
37
40
40
34
945
924
876
603
578
496
585
992
44
87
38
25
56
36
41
69
3
10
9
10
14
7
7
16
Notes: # Seq. = the number of individuals sequenced, # Inform. = the number of parsimony informative sites, # Uninform = the number of variable but parsimony uninformative sites.
geographic clades, only Coulterville corresponds to a single species
at this cutoff. The Automatic Barcode Gap Discovery (ABGD)
method results in a much higher number of putative species. A
minimum of 95 species are inferred with this approach
(Supplementary Material), with most of these restricted to a single
sampled geographic location.
3.3. Species validation using nuclear-only data
Gene annotation and patterns of nuclear gene variation are
summarized in Table 1. The eight nuclear loci totaled 5999 bp with
393 parsimony informative sites. The nuclear-only *BEAST analysis
with forty individuals treated as OTUs recovers a topology that displays a great deal of topological concordance with the mtDNA gene
tree, both in terms of species membership as well as major clade
membership (Fig. 4). Using the 9.5% threshold, 23 of 26 candidate
species were sampled for nuclear data. Of those 23, ten species had
nuclear data collected for multiple individuals. In nine out of ten
instances individuals belonging to a single species were recovered
together in a clade. The exception was species H, as one member of
H was not recovered with the other two (Fig. 4). However, these
spiders belong to two divergent mitochondrial clades that are near
the cutoff threshold (9.1%).
At the three localities with sympatric divergent mitochondrial
clades, a specimen from each clade was sampled for nuclear data.
In these three instances all eight nuclear loci were informative in
regards to distinguishing species membership (Fig. 4). In each case
not only did the combined nuclear data substantiate the mtDNA
clade membership, but also each of the nuclear loci individually
supported the mitochondrial species membership.
In every case the species identified at both the 9.5% and 5% cutoffs are recovered in the same major clades with the nuclear-only
species tree as they are with the mitochondrial gene tree.
Relationships among major clades from the nuclear species tree
and the mtDNA gene tree show a large degree of congruence, but
there are some differences. However, in none of these cases is a
62
D.H. Leavitt et al. / Molecular Phylogenetics and Evolution 91 (2015) 56–67
Fig. 4. Maximum clade credibility tree from the *BEAST analysis of nuclear data with Bayesian posterior probabilities at each node. Individuals are labeled according to
putative species at the 9.5% threshold, major geographic clade and voucher number. Sympatric specimens are indicated by a square (locality 50), a circle (locality 61) and a
star (locality 1). Inset is adult female from Hare Canyon Road (Monterey County).
strongly supported node (0.95 PP or above) from the nuclear species tree in conflict with a strongly supported node in the mitochondrial gene tree.
3.4. Combined data species tree
The species tree from the combined nuclear and mitochondrial
data is depicted in Fig. 5. Nodes that are in common with the
nuclear-only species tree either maintain high support values or
generally received increased support, indicating that the mitochondrial gene tree has not been misled by introgression and can
appropriately be analyzed with the nuclear loci in a combined species tree analysis. In the recovered phylogeny, the three northernmost clades (Sacramento Valley, North Sierra and Bay Area) are
recovered as sister to southerly-distributed clades. Among these
southern clades, the Central Sierra is sister to and deeply divergent
from a clade containing the Tuolumne, Coulterville, Calaveras,
Southern Sierra, Monterey and Panoche clades. Relationships
among these clades are not well supported except for the sister
relationship between the Monterey and Panoche clades, both
distributed west of the Central Valley.
(Supplementary Material) and an obviously smaller percentage of
putative molecular species. This sample includes the CAS specimens (without DNA data), most of which could be confidently
placed into larger geographic clades based on geographic location
(Supplementary Material, Fig. 1). Comparison of palpal and leg I
clasper morphology among individuals from primary clades
indicates limited morphological divergence, with only very subtle
differences in spination, bulb shape, and relative length of leg
elements (Fig. 6). Obviously, an important task will now be targeted
geographic sampling of adult males, informed by the barcoding
results. Our sample of females with mature spermathecal organs
is comprehensive, including all primary geographic clades, and all
26 conservative barcode species (Supplementary Material).
However, female variation showed no obvious trends relative to
clade or species membership, with morphological asymmetry often
apparent within individuals, and variation among individuals from
a single location often greater than divergence among individuals
from different primary clades (Supplementary Material).
4. Discussion
4.1. Taxonomic validity and composition of Calisoga
3.5. Morphological variation
Our sample of adult males is geographically biased, as males are
only available for five of the ten primary geographic clades
The combined results of our higher-level phylogenetic study
and our population-level sampling confirm that California nemesiid diversity consists of a single radiation that is divergent from
D.H. Leavitt et al. / Molecular Phylogenetics and Evolution 91 (2015) 56–67
63
Fig. 5. Maximum clade credibility tree from the *BEAST analysis of the combined mitochondrial and nDNA data with Bayesian posterior probabilities at each node. Putative
species were designated based on the 9.5% threshold, and the comprising localities are listed. Major geographic clades are also labeled. Sympatric localities are indicated by a
square (locality 50), a circle (locality 61) and a star (locality 1). Scale bar represents species tree branch lengths estimated using a rate of 1 substitution/site for the COI
fragment.
(and not sister to) European Brachythele and should instead be
placed in the distinct genus Calisoga Chamberlin 1937. This conclusion was also reached by Bentzien (1976) in his unpublished dissertation based on observed morphological variation. The long
branches separating European Brachythele from Calisoga indicate
ancient divergences.
While the five named taxa from California were originally allocated to three different families, Raven (1985) transferred them all
to Nemesiidae. He further recognized that they belonged in the
genus Calisoga rather than Brachythele, but he neglected to transfer
Brachythele longitarsis Simon 1891 and B. anomala Schenkel 1950
(Ubick and Ledford, 2005). Thus, we recommend taxonomic
changes to make the two following names available.
Calisoga longitarsis (Simon 1891) comb. n.
Brachythele longitarsis Simon 1891.
Type locality: ‘‘California meridionalis’’.
Remarks: The type of longitarsis was reported as being collected
in ‘‘southern California’’. The type specimen is a mature male and is
located in the M.N.H.N. in Paris. Bentzien (1976) examined the
specimen and provided measurements, a description, and partial
illustrations. In the ‘‘1890’’ paper describing a female specimen
of longitarsis (published after the 1891 description of the type),
Simon lists a number of other regions where longitarsis specimens
occur; some of these are clearly in error (e.g., Idaho and Texas).
While the specific geographic origin of the type of longitarsis is currently unknown, it can be reasonably assumed that this species is
represented in our comprehensive sampling (or is at least closely
related to sampled spiders). Due to their morphological similarity
to the type, the majority of the populations sampled in this study
(including the those in the higher-level analysis) have historically
been considered longitarsis.
Calisoga anomala (Schenkel 1950) comb. n.
Brachythele anomala Schenkel 1950.
Type locality: Wald um Guerneville und Monte Rio am Russian
River.
Remarks: The type specimen of C. anomala is an immature animal in poor condition (Bentzien, 1976). However, the type locality
is perhaps the most specific of named Calisoga: forest around
Guerneville and Monte Rio along the Russian River (Sonoma
County, CA).
In addition to the two taxa above, the following species names
are also available for the genus Calisoga: C. sacra Chamberlin, 1937
from Sacramento, CA, the type of genus; C. theveneti (Simon, 1891)
from Mariposa, CA; and C. centronetha (Chamberlin and Ivie, 1939),
from Mayfield (Palo Alto), CA. Thus, five names are available for
California nemesiid diversity. In his unpublished dissertation,
Bentzien (1976) considered Calisoga longitarsis, C. sacra, C. centronetha, and C. anomala as synonyms.
With at least 26 molecular species and five available names,
taxonomy within Calisoga clearly needs further attention. The
small body size and the geographic distribution of ‘‘species C’’ indicate that it should tentatively be considered C. theveneti. The sampled population that is closest to the type locality of C. sacra
(Sacramento) is Deer Valley Rd. and belongs to ‘‘species Q1’’. The
64
D.H. Leavitt et al. / Molecular Phylogenetics and Evolution 91 (2015) 56–67
Bay_Loc21_MY1514
noSier_Loc44
_GMY187
CentSier_
Loc65_MY4603
Coulter_Loc60
_MY 4658
Fig. 6. Digital images of palpal and leg I clasper morphology from male spiders representing four primary geographic clades. Spiders are labeled with primary clade, locality
and voucher number. All views prolateral, scale bars = 1 mm. Arrows point to minor differences in spination and relative leg length observed in the sample.
type locality of C. anomala falls within the distribution of ‘‘species
S’’, and the type of C. centronetha falls within that of ‘‘species U’’.
Unfortunately, the type locality of C. longitarsis is currently
unknown; should its origin be determined, this older name would
have priority if synonymous with one of the three latter named
taxa.
4.2. Molecular species diversity
Given the results of other molecular studies of California mygalomorph spiders (e.g., Bond and Stockman, 2008; Hedin et al.,
2013; Satler et al., 2013), it is not surprising that the
morphology-based ‘‘two species’’ hypothesis of species diversity
within Calisoga is an underestimate. Using the conservative 9.5%
cutoff used in the Pilbara study (Castalanelli et al., 2014), we find
a similar number of molecular species per sampled spider: about
1 species per 7 spiders sampled (7.5 in this study, 7 [1134 spiders/161 lineages] in Castalanelli et al., 2014). Even within the
more conspicuous genus Aphonopelma, mitochondrial barcoding
identified previously unknown species diversity with perhaps nine
morphologically cryptic species recovered (Hamilton et al., 2014).
Of course, the number of ‘‘real species’’ represented in this study
remains unknown. How much divergence is needed to correctly
identify distinct species is a looming question in the DNA barcoding literature (Hamilton et al., 2014) and is likely taxon specific. In
this study system, the 9.5% threshold does not appear to result in
an overestimate of putative species, as the nuclear data support
genetic distinctiveness at this cutoff. This divergence threshold
D.H. Leavitt et al. / Molecular Phylogenetics and Evolution 91 (2015) 56–67
may in fact be conservative, and in a few instances the
nuclear-only topology led us to treat clades just under the threshold (9%) as distinct OTUs.
On the other hand, the approaches of species delimitation using
methods such as ABGD and GMYC (Pons et al., 2006), which do not
use an arbitrary threshold, have been shown to return unrealistic
numbers of species in other mygalomorph studies (e.g., Hamilton
et al., 2014). In a study of the California trapdoor spider species
Aliatypus thompsoni, bGMYC identifies 61–79 species at a 95% confidence interval with mitochondrial data, whereas the authors prefer a conservative three species hypothesis based on analyses of
multilocus variation (Satler et al., 2013). In the current study,
ABGD infers a minimum of 95 species from 106 sampled locations
– almost one species per locality. The overestimation of species
using these methods is due to structured genetic patterns emerging from the viscous nature of mygalomorph populations (e.g.,
Hedin et al., in press). The biology of burrow-dwelling mygalomorphs may be particularly prone to generating extreme mitochondrial structuring. In species not characterized by balloon
dispersal, spiderlings generally establish burrows in favorable substrate near the maternal burrow. Female mygalomorphs usually
remain in the same burrow throughout their entire lives. Males,
on the other hand, are known to disperse from their burrow once
they reach reproductive maturity as they seek opportunities to
mate with receptive females. Female philopatry and male-based
dispersal could create situations where mitochondrial data overestimate isolation and the number of distinct lineages.
The boundary between intraspecific variation and interspecific
variation appears to be better defined in Aphonopelma tarantulas
(Hamilton et al., 2014) than in Calisoga. Some tarantula clades
show patterns of shallow intraspecific mtDNA divergence, indicative of either recent population expansion or a nonviscous genetic
structure (Hamilton et al., 2011). This includes a recent phylogeographic analysis of tarantulas from central California (Wilson et al.,
2013) where the sampled area overlaps a great deal with the area
sampled in this study. Because of some ecological similarities to
tarantulas, we expected Calisoga to share similar phylogeographic
patterns. However, the extreme level of mitochondrial structuring
observed within Calisoga is more similar to that seen in trapdoor
spiders (Satler et al., 2013; Stockman and Bond, 2007; Cooper
et al., 2011) and turret spiders (Starrett and Hedin, 2007; Hedin
et al., 2013). In these systems, mtDNA clades are narrowly distributed, often to a single sampled locality, and genetic divergence
to neighboring clades can be surprisingly deep. Identifying the
boundary between intraspecific phylogeographic structure versus
species boundaries is challenging in low-dispersing, spatially viscous organisms (Bond and Stockman, 2008; Keith and Hedin,
2012; Schuchert, 2014). In an example from California, mitochondrial lineages of the slender salamander Batrachoseps attenuatus
are narrowly distributed, leading to the suggestion that this taxon
comprises 39 extant lineages (Highton, 2014), whereas the initial
researchers that generated the data (Martínez-Solano et al.,
2007) consider this taxon to represent a single species with deep
phylogeographic structure.
Because species boundaries based on mitochondrial data can be
misled by processes like introgression, incomplete lineage sorting
and/or phylogeographic structure unaccompanied by nuclear differentiation, the need for nuclear data is well appreciated. Given
the biology of these spiders, the possibility for discordance between
the mitochondrial genome and the rest of the genome seemed considerable. But rather than reveal potential pitfalls of a
mitochondrial-only analysis, the nuclear-only species tree analysis
provided strong validation for the mtDNA-identified species-level
diversity. Our nuclear-only approach to validating mitochondrial
species boundaries took a qualitative approach. We tested whether
individuals belonging to the same mtDNA-identified species were
65
recovered in a clade using nuclear-only data, and we evaluated
whether the branch lengths separating different species or individuals were consistent with distinct species status. We did not perform statistical evaluations of branch lengths, but rather
considered proportional length relative to overall tree length.
Methods have been developed to statistically test species boundaries using species trees. For example, Bayes factor delimitation
(BFD) (Grummer et al., 2014) evaluates likelihood scores of different species models that either split or group individuals (or reassigns them), and a related approach has been developed with the
program DISSECT (Jones et al., 2015) that allows more flexibility
in testing split or lumped arrangements. The use of DISSECT
appears to have the potential to over-split species with phylogeographic structure (Jones et al., 2015), however, which is a concern
that likely extends to methods like BFD as well. Where the cutoff
between interspecific structure and species boundaries lies in low
dispersing organisms will continue to be an interesting question
as relative information content increases with genomic datasets.
The observed concordance of nuclear and mitochondrial lineage
boundaries provides convincing verification that Calisoga contains
not one or two species but likely over two dozen. Particularly compelling evidence that the lineages recovered in this study represent
‘‘good species’’ is the discovery of three instances of sympatry, each
involving different species pairs (Figs. 1, 4 and 5). In all three
instances, reproductive isolation appears to be quite high with no
detected evidence for gene flow among the sampled nuclear loci.
Reproductive isolation in sympatry is considered the most convincing evidence of distinct species boundaries (Mayr, 1942), and
has been observed in morphologically cryptic diatoms (Amato
et al., 2007), lichens (Leavitt et al., 2011), earthworms (Donnelly
et al., 2013) and lizards (Barley et al., 2013). Certainly the extent
of sympatry among Calisoga species is greater than that shown
here since we sampled relatively few spiders per locality.
Combining future fine-scale geographic sampling with larger sample sizes, to discover and study additional cases of sympatry, will
constitute one of the strongest tests of species limits in Calisoga.
4.3. Phylogenetic relationships and taxonomy
Many of the phylogenetic relationships recovered from the
combined data species tree analysis are well supported, including
those deepest in the tree (Fig. 5). Additional data is needed to solidify relationships among lineages within the Northern Sierra clade
as well as those among the geographically restricted clades in
the central Sierran region, however. As seen in the mtDNA gene
tree and nuclear-only species tree, the combined phylogeny shows
that sympatrically occurring species are distantly related, with two
of the three instances spanning the deepest divergence within
Calisoga (Fig. 5).
Among the Australian mygalomorphs from the Pilbara, most
sampled males were morphologically distinct at the 9.5% threshold
(Castalanelli et al., 2014). The exception was the Aname mellosa
complex, where Castalanelli et al. recovered ten lineages at that
threshold that were not obviously morphologically distinguishable.
To what extent are the molecularly identified species of Calisoga
morphologically identifiable? Surveying the morphology of female
genitalic characters, we observed no obvious species-specific or
clade specific morphology, even though variation among females
was clearly evident (Supplementary Material). An inspection of
male morphology reveals limited morphological divergence, with
only subtle differentiation at the 9.5% cutoff (Fig. 6). Much more
extensive sampling of male specimens is needed to establish
whether these minor differences are taxonomically useful.
Bentzien (1976) considered most of the Californian populations
to belong to a single widespread species, but he did observe morphological variation among these populations, including differing
66
D.H. Leavitt et al. / Molecular Phylogenetics and Evolution 91 (2015) 56–67
average body sizes. While our sample sizes are small, we also
observed some size variation among different populations.
Additional sampling is needed to determine whether differently
sized populations specifically mentioned by Bentzien (1976) represent distinct species. For example, the males he observed at Indian
Valley, Sierra Co. were among the smallest in his study. The
Calisoga specimens we found from the northern Sierra region were
often very large, however, suggesting the Indian Valley spiders
may be a distinct, unsampled species. Additionally, Bentzien found
both small (<7 mm) and large mature males (>10 mm) at Hopland
Field Station, Mendocino Co., which suggests the possibility of
sympatric species. Analyses of morphology have helped shed light
on the taxonomic status of the Nemesiidae genera Ambylocarenum
(Decae et al., 2014) and Iberesia (Decae and Cardoso, 2005); a
molecular perspective will undoubtedly provide additional clarification by revealing genetically unique lineages in these genera as
they have in Calisoga.
4.4. Conclusions
The results of our study reveal that the California genus Calisoga
is distinct from the European genus Brachythele, and that Calisoga
represents a species rich radiation, rather than just one or two species. A correct understanding of species boundaries is requisite for
understanding evolutionary history as well as ecological relationships (e.g., Pante et al., 2015). Additionally, the genetic data provides an important perspective in conservation planning (e.g.,
Hedin, 2015) and habitat preservation (Moritz and Faith, 1998),
and the inclusion of endemic arthropods has proven to be of high
utility (Redak, 2000; Bond et al., 2006; Arnedo and Ferrández,
2007; Harvey et al., 2011). Species discovery in organisms that
are narrowly distributed is an iterative process (Hedin et al.,
2013), and further geographic sampling of Calisoga is needed to
test hypotheses derived from this study and to reveal additional
diversity. While much remains to be done regarding systematic
work in this genus, this study sets the stage for investigating the
temporal and spatial aspects of diversification of Calisoga in the
geologically dynamic region of central and northern California.
Other recent biogeographic studies of California mygalomorphs
have revealed both shared patterns with other organisms as well
as genuine surprises (Hedin et al., 2013).
Funding
This work was supported by a Theodore Roosevelt Memorial
Grant from the American Museum of Natural History awarded to
Dean Leavitt and funding from the Bureau of Land Management.
Acknowledgments
For help in specimen collection we would like to thank Bob
Thomson, Levi Gray, Peter Scott, Anna Bennett, Eileen Bennett,
Chris Hamilton, Jordan Satler, Erika Garcia, Robin Hedin, Bob
Keith, Steven Thomas, Ryan Fitch, Dave Carlson, Axel Schönhofer,
and Lars Hedin. Jason Bond provided assembled transcriptomes
for the Jasper Ridge Calisoga, and Pionothele. Charles Griswold
and Darrell Ubick facilitated the specimen loan from the
California Academy of Sciences. Cheryl Hayashi provided lab space
and other resources for RNA extraction work.
Supplementary material
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.ympev.2015.05.
016.
References
Alfaro, M.E., Zoller, S., Lutzoni, F., 2003. Bayes or bootstrap? A simulation study
comparing the performance of Bayesian Markov chain Monte Carlo sampling
and bootstrapping in assessing phylogenetic confidence. Mol. Biol. Evol. 20,
255–266.
Amato, A., Kooistra, W.H., Levialdi Ghiron, J.H., Mann, D.G., Pröschold, T., Montresor,
M., 2007. Reproductive isolation among sympatric cryptic species in marine
diatoms. Protist 158, 193–207.
Arnedo, M.A., Ferrández, M.A., 2007. Mitochondrial markers reveal deep population
subdivision in the European protected spider Macrothele calpeiana (Walckenaer,
1805) (Araneae, Hexathelidae). Conserv. Genet. 8, 1147–1162.
Avise, J.C., Ball, R.M., 1990. Principles of genealogical concordance in species
concepts and biological taxonomy. Oxford Surv. Evol. Biol. 7, 45–67.
Ayoub, N.A., Garb, J.E., Hedin, M., Hayashi, C.Y., 2007. Utility of the nuclear proteincoding gene, elongation factor-1 gamma (EF-1c) for spider systematics,
emphasizing family level relationships of tarantulas and their kin (Aranaea:
Mygalomorphae). Mol. Phylogenet. Evol. 42, 394–409.
Barley, A.J., White, J., Diesmos, A.C., Brown, R.M., 2013. The challenge of species
delimitation at the extremes: diversification without morphological change in
Philippine sun skinks. Evolution 67, 3556–3572.
Bentzien, M.M., 1976. The Biosystematics of the Spider Genus Brachythele Ausserer
(Araneidae: Dipluridae). PhD Dissertation, UC Berkeley, CA.
Bond, J.E., Stockman, A.K., 2008. An integrative method for delimiting cohesion
species: finding the population-species interface in a group of Californian
trapdoor spiders with extreme genetic divergence and geographic structuring.
Syst. Biol. 57, 628–646.
Bond, J.E., Beamer, D.A., Lamb, T., Hedin, M., 2006. Combining genetic and geospatial
analyses to infer population extinction in mygalomorph spiders endemic to the
Los Angeles region. Anim. Conserv. 9, 145–157.
Bond, J.E., Hendrixson, B.E., Hamilton, C.A., Hedin, M., 2012. A reconsideration of the
classification of the spider infraorder Mygalomorphae based on three nuclear
genes and morphology (Arachnida: Araneae). PLoS One 7, e38753.
Bond, J.E., Garrison, N.L., Hamilton, C.A., Godwin, R.L., Hedin, M., Agnarsson, I., 2014.
Phylogenomics resolves a spider backbone phylogeny and rejects a prevailing
paradigm for orb web evolution. Curr. Biol. 24, 1765–1771.
Castalanelli, M.A., Teale, R., Rix, M.G., Kennington, W.J., Harvey, M.S., 2014.
Barcoding of mygalomorph spiders (Araneae: Mygalomorphae) in the Pilbara
bioregion of Western Australia reveals a highly diverse biota. Invertebr. Syst. 28,
375–385.
Castresana, J., 2000. Selection of conserved blocks from multiple alignments for
their use in phylogenetic analysis. Mol. Phylogenet. Evol. 17, 540–552.
Chamberlin, R.V., Ivie, W., 1939. New tarantulas from the southwestern states. Bull.
Univ. Utah 29, 1–17.
Cooper, S.J., Harvey, M.S., Saint, K.M., Main, B.Y., 2011. Deep phylogeographic
structuring of populations of the trapdoor spider Moggridgea tingle (Migidae)
from southwestern Australia: evidence for long-term refugia within refugia.
Mol. Ecol. 20, 3219–3236.
Decae, A., Cardoso, P., 2005. Iberesia, a new genus of trapdoor spiders (Araneae,
Nemesiidae) from Portugal and Spain. Rev. Ibérica Aracnol. 12, 3–11.
Decae, A., Colombo, M., Manunza, B., 2014. Species diversity in the supposedly
monotypic genus Amblyocarenum Simon, 1892, with the description of a new
species from Sardinia (Araneae, Mygalomorphae, Cyrtaucheniidae).
Arachnology 16, 228–240.
Derkarabetian, S., Hedin, M., 2014. Integrative taxonomy and species delimitation in
harvestmen: A revision of the western North American genus Sclerobunus
(Opiliones: Laniatores: Travunioidea). PLoS One 9, e104982.
Donnelly, R.K., Harper, G.L., Morgan, A.J., Orozco-Terwengel, P., Pinto-Juma, G.A.,
Bruford, M.W., 2013. Nuclear DNA recapitulates the cryptic mitochondrial
lineages of Lumbricus rubellus and suggests the existence of cryptic species in an
ecotoxological soil sentinel. Biol. J. Linn. Soc. 110, 780–795.
Goloboff, P.A., 1995. A revision of the South American spiders of the family
Nemesiidae (Araneae, Mygalomorphae). Part I: species from Peru, Chile,
Argentina, and Uruguay. Bull. Am. Mus. Nat. Hist. 224, 1–189.
Graham, M.R., Hendrixson, B.E., Hamilton, C.A., Bond, J.E., 2015. Miocene
extensional tectonics explain ancient patterns of diversification among turretbuilding tarantulas (Aphonopelma mojave group) in the Mojave and Sonoran
deserts. J. Biogeogr. 42, 1052–1065.
Grummer, J.A., Bryson Jr., R.W., Reeder, T.W., 2014. Species delimitation using Bayes
factors: simulations and application to the Sceloporus scalaris species group
(Squamata: Phrynosomatidae). Syst. Biol. 63, 119–133.
Hamilton, C.A., Formanowicz, D.R., Bond, J.E., 2011. Species delimitation and
phylogeography
of Aphonopelma hentzi
(Araneae, Mygalomorphae,
Theraphosidae): cryptic diversity in North American tarantulas. PLoS One 6,
e26207.
Hamilton, C.A., Hendrixson, B.E., Brewer, M.S., Bond, J.E., 2014. An evaluation of
sampling effects on multiple DNA barcoding methods leads to an integrative
approach for delimiting species: a case study of the North American tarantula
genus Aphonopelma (Araneae, Mygalomorphae, Theraphosidae). Mol.
Phylogenet. Evol. 71, 79–93.
Harrington, R.C., Near, T.J., 2012. Phylogenetic and coalescent strategies of
species delimitation in snubnose darters (Percidae: Etheostoma). Syst. Biol.
61, 63–79.
Harvey, M.S., Rix, M.G., Framenau, V.W., Hamilton, Z.R., Johnson, M.S., Teale, R.J.,
Humphreys, G., Humphreys, W.F., 2011. Protecting the innocent: studying
D.H. Leavitt et al. / Molecular Phylogenetics and Evolution 91 (2015) 56–67
short-range endemic taxa enhances conservation outcomes. Invertebr. Syst. 25,
1–10.
Hebert, P., Cywinska, A., Ball, S., DeWaard, J., 2003a. Biological identifications
through DNA barcodes. Philos. Trans. R. Soc. Lond. B 270, 313–321.
Hebert, P., Ratnasingham, S., de Waard, J.R., 2003b. Barcoding animal life:
cytochrome c oxidase subunit 1 divergences among closely related species.
Philos. Trans. R. Soc. B 270, S96–S99.
Hedin, M., 2015. High-stakes species delimitation in eyeless cave spiders (Cicurina,
Dictynidae, Araneae) from central Texas. Mol. Ecol. 24, 346–361.
Hedin, M., Bond, J.E., 2006. Molecular phylogenetics of the spider infraorder
Mygalomorphae using nuclear rRNA genes (18S and 28S): conflict and
agreement with the current system of classification. Mol. Phylogenet. Evol.
41, 454–471.
Hedin, M., Carlson, D., 2011. A new trapdoor spider species from the southern Coast
Ranges of California (Mygalomorphae, Antrodiaetidae, Aliatypus coylei, nov sp),
including consideration of mitochondrial phylogeographic structuring. Zootaxa
2963, 55–68.
Hedin, M., Starrett, J., Hayashi, C., 2013. Crossing the uncrossable: novel trans-valley
biogeographic patterns revealed in the genetic history of low dispersal
mygalomorph spiders (Antrodiaetidae, Antrodiaetus) from California. Mol.
Ecol. 22, 508–526.
Hedin, M., Carlson, D., Coyle, F., in press. Sky island diversification meets the
multispecies coalescent: divergence in the spruce-fir moss spider (Microhexura
montivaga) on the highest peaks of southern Appalachia. Mol. Ecol.
Hendrixson, B.E., DeRussy, B.M., Hamilton, C.A., Bond, J.E., 2013. An exploration of
species boundaries in turret-building tarantulas of the Mojave Desert (Araneae,
Mygalomorphae, Theraphosidae, Aphonopelma). Mol. Phylogenet. Evol. 66, 327–
340.
Highton, R., 2014. Detecting cryptic species in phylogeographic studies: Speciation
in the California Slender Salamander, Batrachoseps attenuatus. Mol. Phylogenet.
Evol. 71, 127–141.
Hillis, D.M., Bull, J.J., 1993. An empirical test of bootstrapping as a method for
assessing confidence in phylogenetic analysis. Syst. Biol. 42, 182–192.
Irwin, D.E., 2002. Phylogeographic breaks without geographic barriers to gene flow.
Evolution 56, 2383–2394.
Ji, Y.J., Zhang, D.X., He, L.J., 2003. Evolutionary conservation and versatility of a new
set of primers for amplifying the ribosomal internal transcribed spacer regions
in insects and other invertebrates. Mol. Ecol. Notes 3, 581–585.
Jones, G., Aydin, Z., Oxelman, B., 2015. DISSECT: an assignment-free Bayesian
discovery method for species delimitation under the multispecies coalescent.
Bioinformatics 31, 991–998.
Katoh, K., Kuma, K.-I., Toh, H., Miyata, T., 2005. MAFFT version 5: improvement in
accuracy of multiple sequence alignment. Nucleic Acids Res. 33, 511–518.
Keith, R., Hedin, M., 2012. Extreme mitochondrial population subdivision in
southern Appalachian paleoendemic spiders (Araneae: Hypochilidae:
Hypochilus), with implications for species delimitation. J. Arachnol. 40, 167–181.
Kuo, C.H., Avise, J.C., 2005. Phylogeographic breaks in low-dispersal species: the
emergence of concordance across gene trees. Genetica 124, 179–186.
Leavitt, S.D., Fankhauser, J.D., Leavitt, D.H., Porter, L.D., Johnson, L.A., St Clair, L.L.,
2011. Complex patterns of speciation in cosmopolitan ‘‘rock posy’’ lichens–
Discovering and delimiting cryptic fungal species in the lichen-forming
Rhizoplaca melanophthalma species-complex (Lecanoraceae, Ascomycota). Mol.
Phylogenet. Evol. 59, 587–602.
Maddison, W.P., 1997. Gene trees in species trees. Syst. Biol. 46, 523–536.
Martínez-Solano, I., Jockusch, E.L., Wake, D.B., 2007. Extreme population
subdivision throughout a continuous range: phylogeography of Batrachoseps
attenuatus (Caudata: Plethodontidae) in western North America. Mol. Ecol. 16,
4335–4355.
Masta, S.E., Boore, J.L., 2008. Parallel evolution of truncated transfer RNA genes in
arachnid mitochondrial genomes. Mol. Phylogenet. Evol. 25, 949–959.
Mayr, E., 1942. Systematics and the Origin of Species. Columbia University Press,
New York, New York.
McGuire, J.A., Linkem, C.W., Koo, M.S., Hutchison, D.W., Lappin, A.K., Orange, D.I.,
Lemos-Espinal, J.A., Riddle, B.R., Jaeger, J.R., 2007. Mitochondrial introgression
and incomplete lineage sorting through space and time: phylogenetics of
crotaphytid lizards. Evolution 61, 2879–2897.
Miller, M.A., Holder, M.T., Vos, R., Midford, P.R., Liebowitz, T., Chan, L., Hoover, P.,
Warnow, T., 2011. The CIPRES Portals. CIPRES. <http://www.phylo.
org/sub_sections/portal>.
67
Moritz, C., Faith, D.P., 1998. Comparative phylogeography and the identification of
genetically divergent areas for conservation. Mol. Ecol. 7, 419–429.
Ogden, R., Thorpe, R.S., 2002. Molecular evidence for ecological speciation in
tropical habitats. Proc. Natl. Acad. Sci. 99, 13612–13615.
Opatova, V., Arnedo, M.A., 2014a. From Gondwana to Europe: inferring the origins
of Mediterranean Macrothele spiders (Araneae: Hexathelidae) and the limits of
the family Hexathelidae. Invertebr. Syst. 28, 361–374.
Opatova, V., Arnedo, M.A., 2014b. Spiders on a hot volcanic roof: colonisation
pathways and phylogeography of the Canary Islands endemic trap-door spider
Titanidiops canariensis (Araneae, Idiopidae). PLoS One 9, e115078.
Opatova, V., Bond, J.E., Arnedo, M.A., 2013. Ancient origins of the Mediterranean
trap-door spiders of the family Ctenizidae (Araneae, Mygalomorphae). Mol.
Phylogenet. Evol. 69, 1135–1145.
Pante, E., Puillandre, N., Viricel, A., Arnaud-Haond, S., Aurelle, D., Castelin, M.,
Chenuil, A., Destombe, C., Forcioli, D., Valero, M., Viard, F., Samadi, S., 2015.
Species are hypotheses: avoid connectivity assessments based on pillars of
sand. Mol. Ecol. 24, 525–544.
Pons, J., Barraclough, T.G., Gomez-Zurita, J., Cardoso, A., Duran, D.P., Hazell, S.,
Kamoun, S., Sumlin, W.D., Vogler, A.P., 2006. Sequence-based species
delimitation for the DNA taxonomy of undescribed insects. Syst. Biol. 55,
595–609.
Posada, D., 2008. JModelTest: phylogenetic model averaging. Mol. Biol. Evol. 25,
1253–1256.
Puillandre, N., Lambert, A., Brouillet, S., Achaz, G., 2012. ABGD, Automatic Barcode
Gap Discovery for primary species delimitation. Mol. Ecol. 21, 1864–1877.
Rambaut, A., Drummond, A.J., 2007. Tracer v1.5. <http://beast.bio.ed.ac.uk/Tracer>
Raven, R.J., 1985. The spider infraorder Mygalomorphae (Araneae): cladistics and
systematics. Bull. Am. Mus. Nat. Hist. 182, 1–180.
Raven, R.J., 1987. A new mygalomorph spider genus from Mexico (Nemesiinae,
Nemesiidae, Arachnida). J. Arachnol. 14, 357–362.
Redak, R.A., 2000. Arthropods and multispecies habitat conservation plans: are we
missing something? Environ. Manage. 26, 97–107.
Satler, J.D., Carstens, B.C., Hedin, M., 2013. Multilocus species delimitation in a
complex of morphologically conserved trapdoor spiders (Mygalomorphae,
Antrodiaetidae, Aliatypus). Syst. Biol. 62, 805–823.
Schuchert, P., 2014. High genetic diversity in the hydroid Plumularia setacea: a
multitude of cryptic species or extensive population subdivision? Mol.
Phylogenet. Evol. 76, 1–9.
Schwenk, K., Brede, N., Streit, B., 2008. Introduction. Extent, processes and
evolutionary impact of interspecific hybridization in animals. Philos. Trans. R.
Soc. B 363, 2805–2811.
Simon, E., 1891. Etudes arachnologiques. 23e Mémoire. XXXVIII. Descriptions
d’espèces et de genres nouveaux de la famille des Aviculariidae. Ann. Soc.
Entomol. France 60, 300–312.
Stamatakis, A., Hoover, P., Rougemont, J., 2008. A rapid bootstrap algorithm for the
RAxML web-servers. Syst. Biol. 75, 758–771.
Starrett, J., Hedin, H., 2007. Multilocus genealogies reveal multiple cryptic species
and biogeographic complexity in the California turret spider Antrodiaetus riversi
(Mygalomorphae, Antrodieatidae). Mol. Ecol. 16, 583–604.
Starrett, J., Hedin, M., Ayoub, N., Hayashi, C.Y., 2013. Hemocyanin gene family
evolution in spiders (Araneae), with implications for phylogenetic relationships
and divergence times in the infraorder Mygalomorphae. Gene 524, 175–186.
Stockman, A.K., Bond, J.E., 2007. Delimiting cohesion species: extreme population
structuring and the role of ecological interchangeability. Mol. Ecol. 16, 3374–
3392.
Thomson, R.C., Wang, I.J., Johnson, J.R., 2010. Genome-enabled development of DNA
markers for ecology, evolution and conservation. Mol. Ecol. 19, 2184–2195.
Ubick, D., Ledford, J.M., 2005. Nemesiidae. In: Ubik, D., Paquin, P., Cushing, P.E., Roth,
V. (Eds.), Spiders of North America: An Identification Manual. American
Arachnological Society, pp. 52–53.
Ubick, D., Paquin, P., Cushing, P.E., Roth, V. (Eds.), 2005. Spiders of North America:
An Identification Manual. American Arachnological Society, 377 pages.
Wilson, J.S., Gunnell, C.F., Wahl, D.B., Pitts, J.P., 2013. Testing the species limits of the
tarantulas (Araneae: Theraphosidae) endemic to California’s Southern Coast
Ranges, USA. Insect Conserv. Divers. 6, 365–371.
World Spider Catalog, 2015. World Spider Catalog. Natural History Museum Bern.
<http://wsc.nmbe.ch>, version 16 (accessed 02.02.15).