Biogeographical evidence for common vicariance and rare ...€¦ · hypothesized to have dispersed...

ORIGINALARTICLE

Biogeographical evidence for commonvicariance and rare dispersal in asouthern Appalachian harvestman(Sabaconidae Sabacon cavicolens)Marshal Hedin1 and Maureen McCormack12

1Department of Biology San Diego State

University San Diego CA 92182-4614 USA2Wisconsin State Lab of Hygiene University

of Wisconsin-Madison Madison WI 53706

USA

Correspondence Marshal Hedin Department

of Biology San Diego State University San

Diego CA 92182-4614 USA

E-mail mhedinmailsdsuedu

ABSTRACT

Aim Species or higher taxa that are obviously dispersal-limited but which

occupy large geographical distributions represent a biogeographical paradox

Dispersal must have happened likely under special and infrequent environ-

mental conditions but details have been lost to history The overarching goal

of our research is to understand the details of a lsquocommon vicariance rare dis-

persalrsquo biogeographical history in a widespread but habitat-specialized harvest-

man species (Sabacon cavicolens) with a southern Appalachian centre of

distribution

Location Eastern North America southern Appalachians

Methods We assessed cryptic speciation using mitochondrial and nuclear gene

DNA sequence data testing alternative delimitation hypotheses using multi-

species coalescent analyses We also tested whether riverine barriers are associ-

ated with mitochondrial genealogical structuring focusing on multiple rivers in

the southern Blue Ridge physiographical province Finally we conducted popu-

lation genetic analyses to assess female-based range expansion out of the south-

ern Blue Ridge

Results Genetic analyses suggest a large number of species-level lineages

within S cavicolens although we prefer a more conservative three-species

hypothesis These putative species are geographically cohesive and allopatric

(Ozarks Cumberland Plateau southern Blue Ridge) with the Blue Ridge spe-

cies including multiple divergent mitochondrial haplogroups Several genealogi-

cal breaks in the Blue Ridge species coincide with riverine barriers separating

mostly allopatric mitochondrial lineages Contrasting with evidence for con-

strained gene flow and vicariance two Blue Ridge haplogroups reveal extensive

range expansion both northwards and westwards resulting in the widespread

distribution of closely related haplotypes and occasional sympatry of dispersive

haplotypes

Main Conclusions Hidden beneath the apparently widespread distribution of

a single species is a history of old vicariance separating geographically disjunct

cryptic species How these lineages came to occupy such disparate geographies

is illustrated by dynamics within the Blue Ridge species where both in situ

vicariance and long-distance dispersal have shaped a lsquocommon vicariance rare

dispersalrsquo biogeographical history

Keywords

competitive exclusion cryptic species glacial refugia long-distance dispersal

riverine barrier short-range endemism

ordf 2017 John Wiley amp Sons Ltd httpwileyonlinelibrarycomjournaljbi 1665doi101111jbi12973

Journal of Biogeography (J Biogeogr) (2017) 44 1665ndash1678

INTRODUCTION

Short-range endemic (SRE) taxa defined as species or

higher taxa with naturally small geographical distributions

(Harvey 2002 Harvey et al 2011 Keith amp Hedin 2012)

present a biogeographical paradox SRE taxa are often dis-

tributed as arrays of parapatric or allopatric taxa consis-

tent with a biogeographical history dominated by low

dispersal and vicariance However the larger encompassing

geographical distribution of such allo- or parapatric arrays

implies some level of dispersal during the history of the

lineage (eg Garcıa-Vazquez amp Ribera 2016) An illustra-

tive example is the North American spider genus Hypochi-

lus with five species in the southern Appalachians two

species in the southern Rockies and three species in mon-

tane California All Hypochilus species are habitat-specia-

lized short-range endemics where allopatry prevails even

within montane regions (Hedin 2001) Moreover available

genetic data for individual species suggests extreme popula-

tion genetic fragmentation (Hedin amp Wood 2002 Keith amp

Hedin 2012) Overall vicariance clearly dominates the his-

tory of this SRE taxon But how do we explain the overall

larger geographical distribution in California or in Appala-

chia or in North America Biogeographers are left to infer

historical dispersal based on indirect evidence sometimes

despite a complete lack of evidence for modern-day dis-

persal abilities in such taxa This is a problem with

ancient and rare dispersal ndash dispersal must have happened

likely under special environmental conditions but details

have been lost to history

Biogeographical patterns observed in the harvestman

genus Sabacon seem consistent with a history dominated by

vicariance with occasional rare dispersal Sabacon includes

over 50 described species known from disjunct geographical

centres in the Northern Hemisphere (North America west-

ern Europe eastern Siberia Nepal Korea and Japan see

Scheuroonhofer et al 2013 Martens 2015) Sabacon prefer

moist cool microhabitats in habitats such as caves upland

forests under rocks or woody debris and under rocks near

streams (summarized in Scheuroonhofer et al 2013) It appears

that physiological constraints to cryophilic microhabitats

restrict gene flow and higher level phylogenetic results are

consistent with this assertion Larger phylogenetic clades

correspond to geographical centres of endemism on differ-

ent continents each of these centres with arrays of closely

related SRE taxa occurring in allopatry or parapatry How-

ever interwoven in this vicariance-dominated history is evi-

dence for rare long-distance dispersal Because harvestmen

are not known to be phoretic (Giribet amp Sharma 2015)

this dispersal must have occurred via intrinsic short-range

movements over hundreds or thousands of generations In

Sabacon an apparent pulse of trans-continental dispersal

events is temporally coincident with special environmental

conditions of the lsquoicehousersquo Eocene-Oligocene transition

(Ivany et al 2000 Scheuroonhofer et al 2013) These ancient

dispersal events occurred in multiple lineages and are

inferred based on the combination of geographical distribu-

tion time-calibrated phylogenies and algorithmic biogeo-

graphical analyses that indirectly imply dispersal

(Scheuroonhofer et al 2013)

This paper focuses on a single wide-ranging Sabacon

species (S cavicolens Packard 1884) from the eastern Uni-

ted States This taxon is related to species from southern

Europe with the lineage that ultimately led to S cavicolens

hypothesized to have dispersed to eastern North America

near the Eocene-Oligocene transition (Scheuroonhofer et al

2013) Sabacon cavicolens occurs most commonly in moist

rocky habitats in the richly forested mountains of the

southern Appalachians but also includes more northern

populations in previously glaciated regions (eg Wisconsin

Michigan Maine etc) and disjunct populations in the

Ozark highlands (Fig 1 Shear 1975) We hypothesize that

S cavicolens displays a lsquocommon vicariance rare dispersalrsquo

biogeographical history and as such represents a micro-

cosm of biogeographical patterns observed more broadly

within Sabacon However because the timeframe is more

recent inferred patterns provide more direct and thus

more powerful evidence for biogeographical processes For

example intraspecific phylogeographic data can be used to

more directly infer demographic expansion events such

demographic information is lost in comparisons at deeper

phylogenetic levels

Geographical fragmentation has promoted SRE specia-

tion in many terrestrial lineages that occupy the mountains

of the southern Appalachians Vertebrate examples are

numerous in salamanders (eg Weisrock amp Larson 2006

Kozak amp Wiens 2010 Herman amp Bouzat 2016) whereas

arthropod examples include other harvestmen (Thomas amp

Hedin 2008 Hedin amp Thomas 2010) spiders (Hedin

1997 2001 Hendrixson amp Bond 2005 Keith amp Hedin

2012 Hedin et al 2015) and others Also intraspecific

phylogeographical fragmentation is evident within more

wide-ranging species with many taxa including divergent

populations on distinct massifs separated by lowland river-

ine barriers For example a conspicuous lowland riverine

barrier is the Asheville Basin which includes the French

Broad River This relatively dry lowland is unsuitable for

habitat-specialized upland species and many phylogenetic

and phylogeographic studies have implicated this barrier

(eg Thomas amp Hedin 2008 Kozak amp Wiens 2010 Hedin

et al 2015) More generally Thomas amp Hedin (2008)

hypothesized that multiple riverine barriers in the southern

Blue Ridge physiographical province may act to promote

divergence a pattern also seen in salamanders (Herman amp

Bouzat 2016)

Climatic stability through time and in particular a lack of

glaciation in the southernmost mountains during the Pleis-

tocene has also promoted high diversity in the southern

Blue Ridge However many lineages have ultimately

expanded out of these refugial mountains (reviewed in Soltis

et al 2006 Herman amp Bouzat 2016) with genetic studies

revealing patterns consistent with post-glacial range

Journal of Biogeography 44 1665ndash1678ordf 2017 John Wiley amp Sons Ltd

1666

M Hedin and M McCormack

expansion from a southern Appalachian centre of distribu-

tion These expansion events involve lineages that have

rapidly dispersed northwards to previously glaciated regions

but there is also evidence for expansion westwards and even

southwards (Walker et al 2009 Emerson et al 2010

Herman amp Bouzat 2016)

The research presented here addresses three primary

questions Based on a three-gene concatenated analysis for

a limited sample Scheuroonhofer et al (2013) suggested that S

cavicolens includes multiple cryptic species We use a larger

specimen sample with mitochondrial and five nuclear genes

to formally assess cryptic speciation testing alternative

delimitation hypotheses using multispecies coalescent

(MSC) analyses Finding cryptic species would indicate

more regional endemism than the current single-species

taxonomy implies Second we use a dense geographical

sample from the southern Blue Ridge to explicitly test

whether riverine barriers have resulted in phylogeographical

fragmentation in Sabacon Finally we seek evidence for

range expansion from the southern Blue Ridge Overall our

goal is to understand the details of a lsquocommon vicariance

rare dispersalrsquo biogeographical history at relatively shallow

time-scales

MATERIALS AND METHODS

Taxon sampling and sequence data collection

Specimens were hand-collected or collected with an aspira-

tor from under rocks and woody debris typically along

streams Specimens destined for molecular work were pre-

served in 100 EtOH and stored at 80 degC Genomic

DNA was extracted from leg tissues using the Qiagen

DNeasy kit per manufacturerrsquos instructions the remaining

specimen was retained as a voucher in the SDSU Terres-

trial Arthropods collection The complete sample includes

168 specimens from 88 locations with denser sampling in

the southern Blue Ridge (Fig 1 see Appendix S1 in Sup-

porting Information) The sample covers the known south-

ern western and north-eastern distributional limits of S

cavicolens but lacks north-western samples (eg Michigan

Wisconsin etc Shear 1975)

Figure 1 Map of sampled locations for Sabacon cavicolens Upper left inset shows overall geographical sample highlighting distributionsfor Ozarks Cumberland and southern Blue Ridge genetic clades Main map shows sample locations for mitochondrial subclades within

southern Blue Ridge Sabacon Numbers correspond to geographical locations as in Appendix S1 Circled numbers correspond to siteswith Clinch Mountain-Bullpen syntopy site 74 (square box) includes Clinch Mountain-Cumberland syntopy Approximate location of

relevant regional rivers shown [Colour figure can be viewed at wileyonlinelibrarycom]


1667

Biogeographical history of Appalachian Sabacon

Mitochondrial gene tree and species discovery

analyses

Mitochondrial cytochrome oxidase I (COI) sequences were

generated for the complete specimen sample using PCR pri-

mers and conditions summarized in Appendix S2 A single

COI sequence was also extracted from a transcriptome (see

Appendix S2) Twelve ingroup COI sequences were previ-

ously reported (Scheuroonhofer et al 2013) PCR products were

purified via polyethylene glycol (PEG) precipitation and

Sanger sequenced at the SDSU Microchemical Core Facility

Sequence contigs were assembled and edited using

Sequencher 45 (Gene Codes Corporation Ann Arbor MI

USA) and manually aligned using Geneious Pro 717

(httpwwwgeneiouscom Kearse et al 2012)

A mitochondrial gene tree was reconstructed using maxi-

mum likelihood implemented in the RAxML_GUI

(Stamatakis 2006 2014 Silvestro amp Michalak 2012) Analy-

ses included a thorough bootstrap analysis (1000 bootstrap

replicates) followed by multiple inferences (100) on align-

ments Sequences were partitioned by codon with a

GTR_Gamma model applied to each partition Mitochon-

drial trees were rooted using sequences from the Ozarks lin-

eage following the root placement of Scheuroonhofer et al

(2013) based on Bayesian analyses of concatenated nuclear

and mitochondrial data The mitochondrial RAxML gene

tree was used as input in bPTP species delimitation analyses

implemented on the bPTP server (httpspeciesh-itsorgptp

Zhang et al 2013) PTP implements a model assuming

gene tree branch lengths generated by two independent Pois-

son process classes (within- and among-species substitution

events) to delimit putative species bPTP analyses were run

for 100000 Markov chain Monte Carlo (MCMC) genera-

tions with a thinning of 100 and burn-in of 01

Nuclear species validation analyses

RAxML and bPTP analyses reveal multiple highly divergent

COI lineages perhaps corresponding to cryptic species (see

Results also Scheuroonhofer et al 2013) To further assess this

possibility we subsampled 20 individuals from 19 locations

generating Sanger sequence data for five nuclear genes (see

Appendix S1) These data were supplemented with transcrip-

tome data from a single sample (see Appendix S2) Nuclear

genes included ribosomal 28S and elongation factor one-

alpha (EF1-a) exon used previously in species-level harvest-

man studies (eg Derkarabetian et al 2011 Scheuroonhofer

et al 2013) plus three novel nuclear genes generated from

comparisons of Sabacon transcriptomes (Table 1 see

Appendix S2) PCR conditions and primers are summarized

in Appendix S2 PCR products were purified via PEG precip-

itation or Millipore plates and Sanger sequenced at SDSU or

Macrogen USA All matrices were assembled edited and

aligned manually in Geneious EF1-a data from four indi-

viduals included a single heterozygous site per sequence

which were phased manually Heterozygous nuclear

sequences for the three novel nuclear genes were bioinfor-

matically phased to alleles using the software program Phase

211 (Stephens et al 2001 Stephens amp Donnelly 2003)

SeqPhase (Flot 2010) was first used to convert matrices for

input into Phase with Phase analyses then conducted using

default settings (phase threshold = 90 100 iterations thin-

ning interval = 1 burn-in = 100)

To summarize overall patterns of nuclear gene divergence

a NeighborNet network was constructed in SplitsTree4

(Huson amp Bryant 2006) This network was based on multi-

genic nuclear genetic distances among individuals calculated

using Pofad 105 (Joly amp Bruneau 2006) Pofad analyses

were conducted using both standardized (all individual

matrices given the same weight) and non-standardized matri-

ces (more variable matrices with greater weight) individual

gene distance matrices were calculated using Kimura 2-para-

meter (K2P) distances in mega 606 (Tamura et al 2013)

Bayes factor delimitation (BFD Grummer et al 2014)

was used to statistically distinguish among alternative species

delimitation models This method compares the marginal

likelihoods of competing hypotheses (eg species tree models

with sequences assigned to differing numbers of lineages)

and chooses the model that best explains the data based on

calculation of BFs (Grummer et al 2014) Following COI

bPTP and nuclear Pofad results Sabacon lineages were split

or combined to represent three alternative species

Table 1 Eastern Sabacon gene data

Primer combo

name

Aligned lengthno

sequences

Substitution

model PI sites Nucleotide diversity Annotation

COI 954168 266 0081

28S 114917 HKY+I+G 17 0006

EF1-a 66621 K80+I 18 0010

SAB A1_A2 55131 K80+I 18 0008 Homologous to Ixodes protein ISCW001223

Mid1-interacting protein


Zinc finger protein

SAB B5_B6 50427 K80+I 47 0028 Homologous to Ixodes protein ISCW010225

Conserved hypothetical protein

Note Parsimony informative sites and nucleotide diversity values calculated using mega (Tamura et al 2013)


1668


delimitation hypotheses (Table 2) A beast species tree (in-

ferred using beast 182 Drummond et al 2012) was esti-

mated for each alternative hypothesis using nuclear only

matrices without outgroups beast analyses were per-

formed using 10 million generations with data saved every

1000 generations the first 20 of each run was discarded as

burn-in For each hypothesis three beast replicates were

conducted to ensure convergence further assessed using

effective sample size (ESS) values with Tracer 16 (Rambaut

et al 2014) Substitution models were chosen with Parti-

tionFinder 111 (Lanfear et al 2012) using linked branch

lengths beast models of molecular evolution BIC model

selection and greedy searches (Table 1) Because almost all

nucleotide variation occurred at third positions for nuclear

protein coding genes a single model was applied to each

gene Marginal likelihoods were estimated using path-sam-

pling (PS Lartillot amp Philippe 2006) and stepping-stone (SS

Xie et al 2011) methods with 100 path steps a chain length

of 100000 generations and likelihoods saved every 100 gen-

erations Marginal likelihood estimates (MLE) were averaged

across replicate runs to generate a single PS and SS value for

each hypothesis Bayes factors were calculated by taking the

difference between the log of the best MLE and the log of

other MLEs and multiplying each result by two [ie

2 9 (lnHypA lnHypB)] Following Kass amp Raftery

(1995) BF values exceeding 10 were considered as lsquodecisiversquo

support for a hypothesis

Test for riverine barriers

The significance of mitochondrial phylogeographical breaks

across modern-day riverine barriers was tested using partial

Mantel tests (eg Kozak et al 2006 Lemmon et al 2007)

Based on COI gene tree patterns four analyses were con-

ducted to examine the relationship between genetic (patris-

tic) distance and putative barriers (Little Tennessee

Tuckaseegee French Broad Table 3) while controlling for

geographical distance Using a randomly sampled single hap-

lotype per geographical location phylogenetic patristic dis-

tances were calculated from a COI RAxML tree using the

lsquoadephylorsquopackage (Jombart amp Dray 2008) in R (2014) For

each test a binary river barrier response matrix was assem-

bled with locations on the same versus opposite sides of a

potential barrier scored as 0 versus 1 respectively (see

Table 3) Geographical distance matrices were calculated

using the Geographic Distance Matrix Generator (http

biodiversityinformaticsamnhorgopen_sourcegdmg) Partial

mantel tests were calculated using the lsquoveganrsquo package (Oksa-

nen et al 2015) in R using 10000 permutations to evaluate

significance

Range expansion of mitochondrial subclades

Standard nucleotide summary statistics were calculated for

southern Blue Ridge COI subclades (see Results) including

intra- and intergroup K2P distances number of haplotypes

(h) number of segregating sites (S) and nucleotide diversity

(p) All values were calculated in mega (Tamura et al

2013) Tajimarsquos D (Tajima 1989) and Fursquos Fs (Fu 1997)

were used to detect deviations from demographic equilib-

rium where an excess of low-frequency polymorphisms (cor-

responding to negative values) indicates demographic

expansion (or non-neutral molecular evolution) Significance

was assessed in Arlequin 3522 (Excoffier amp Lischer 2010)

using 50000 coalescent simulations Mismatch distributions

were used to test for spatial expansion (Excoffier 2004

Table 2 Alternative species delimitation hypotheses tested using BFD

Hypothesis Distinct species (total in parentheses)

Average marginal likelihoods

(path samplingstepping stone) Motivation

H1 Ozarks Cumberland Chunky Gal_1

Chunky Gal_2 Joyce Kilmer

northern Georgia Bullpen

Clinch Mountain (8)

70176970191 COI RaxML clades + conservative

interpretation of bPTP

H2 Ozarks Cumberland southern Montane (3) 703605703737 Conservative COI RaxML nuclear POFAD

H3 Ozarks Appalachians (2) 708749708915 Conservative biogeographical hypothesis

Table 3 Results of partial Mantel tests for riverine barriers

CladeBarrier R P-value Details

Within Joyce Kilmer ndash role of little Tennessee 01937 00278 Populations 8ndash20 vs 21ndash25 on opposite sides of barrier

Within Bullpen ndash role of Tuckaseegee 09237 00001 Populations (26 27 31 32) vs (28ndash30 33ndash43) on opposite sides of barrier

BullpenJoyce Kilmer ndash role of little Tennessee 03044 00001 Populations 8ndash20 vs 21ndash43 on opposite sides of barrier

BullpenClinch ndash role of French Broad 08208 00001 Populations (26ndash28 31ndash43) vs (29 30 46ndash80) on opposite sides of barrier

Notes See Fig 1 for map of locations All tests involving Bullpen Clade excluded populations 44ndash50 part of lsquoBullpen expansion cladersquo (ie not

found in southern Blue Ridge)


1669


Excoffier amp Lischer 2010) considering both sum of

squared deviations (SSD) statistics and the raggedness

index (Harpending 1994) with statistical significance

assessed in Arlequin based on parametric bootstraps

(99000 replicates)

Bayesian skyline plots (BSPs) were also used to detect

changes in effective population size potentially associated

with range expansion (Ho amp Shapiro 2011) These analyses

were conducted using beast 183 (Drummond et al 2012)

focusing on three well-sampled Blue Ridge COI subclades

(see Results) Models of molecular evolution were chosen

using PartitionFinder beast analyses were conducted

using a lognormal uncorrelated relaxed clock model piece-

wise-constant Coalescent Bayesian Skyline tree prior (Drum-

mond et al 2005) with MCMC chain lengths (gt 20

million) set to achieve high ESS values A COI clock rate was

specified based on a well calibrated arthropod rate using a

normal prior with a ucldmean of 00178 00019 (Papado-

poulou et al 2010 Table 4) This mitochondrial rate has

provided realistic divergence time estimates in other harvest-

men (DiDomenico amp Hedin 2016) but we acknowledge the

caveats involved in applying a universal rate in this particular

system All beast analyses were replicated to insure consis-

tent results

RESULTS

Unique mitochondrial sequences and unphased nuclear DNA

sequences have been submitted to GenBank (see

Appendix S1) Sequence lengths number of parsimony infor-

mative sites sequence evolution models and diversity statis-

tics are summarized in Table 1

Mitochondrial clades

Three primary geographical clades are recovered in COI

RAxML analyses (Figs 1 amp 2) These include an Ozarks

clade a Cumberland Plateau clade (including a location near

the type locality of S cavicolens) and a broadly distributed

southern Blue Ridge clade with an apparent centre of diversi-

fication in the Blue Ridge of the southern Appalachians

Four subclades are found within the southern Blue Ridge

clade informally named Clinch Mountain northern Georgia

Joyce Kilmer and Bullpen (Figs 1ndash3) pairwise average K2P

distances among these subclades are relatively high ranging

from 6 to 13 (Table 4) Sequences from two additional

southern populations form a grade that we refer to as

Chunky Gal Within the southern Blue Ridge clade geo-

graphically adjacent Joyce Kilmer and Bullpen are COI sister

clades occupying the intermediate geographical region

between disjunct northern Georgia and Clinch Mountain

clades (Fig 1)

Nuclear evidence for cryptic species

Mitochondrial bPTP analyses suggest up to twelve putative

cryptic species within S cavicolens (Figs 2 amp 3) These corre-

spond to the primary clades and subclades outlined above

some of which are further split in bPTP analyses (eg each

Chunky Gal branch corresponds to a putative species etc

Figs 2 amp 3) This high number of distinct mitochondrial lin-

eages is not reflected in the nuclear standardized-distances

Pofad results which more conservatively indicate three

groups congruent with three primary mitochondrial clades

(Ozarks Cumberland Plateau southern Blue Ridge Fig 4a)

Pofad results using non-standardized distances are similar

(not shown) Individuals from different southern Blue Ridge

mitochondrial subclades do not form distinct clusters in

nuclear Pofad networks (Fig 4a)

Three alternative hypotheses were tested with nuclear-only

matrices using BFD (Table 2) Average marginal likelihood

values are highest for the eight species model using both PS

and SS methods with BF values indicating lsquodecisiversquo support

over alternative three-species and two-species hypotheses

(Table 2) Likewise BF values indicate lsquodecisiversquo support for a

three-species hypothesis over the two-species hypothesis

Because the three-species hypothesis is clearly supported by

independent analyses (Figs 2 amp 4a) we prefer this more con-

servative hypothesis the beast tree corresponding to this

result is shown in Fig 4b An argument as to why BFD might

potentially over-split lineages is presented in the Discussion

Riverine barriers

Several genealogical breaks both within and among southern

Blue Ridge COI subclades suggest a possible role for rivers as

Table 4 Nucleotide summary statistics for Blue Ridge COI subclades

COI subclade noGA CG JK CM BP n S p

Northern Georgia 0005 ndash ndash ndash ndash 10 12 00045

Chunky Gal 0112 003 ndash ndash ndash 4 47 00288

Joyce Kilmer 0131 006 0015 ndash ndash 32 57 00144

Clinch Mountain 0082 0117 0128 0028 ndash 61 120 00269

Bullpen 0134 0069 0076 0125 0018 54 91 00177

Bullpen expansion ndash ndash ndash ndash ndash 9 33 00115

Notes Number of base substitutions per site averaged over all sequence pairs within groups (on diagonal italics) and between groups (lower tri-

angle) conducted using K2P model Chunky Gal populations treated as a clade n = Number of sequences S = segregating sites p = nucleotide

diversity All values calculated in mega


1670


barriers to female-based gene flow The Little Tennessee plays

a potential role within the Joyce Kilmer subclade and

between the Joyce Kilmer and Bullpen subclades the Tuck-

aseegee plays a potential role within the Bullpen subclade

the French Broad plays a role between Clinch and Bullpen

subclades (Figs 1ndash3) The influence of all of these potential

barriers is statistically significant using partial Mantel tests

(Table 3)

Syntopy and expansion of mitochondrial lineages

Although allopatry of mitochondrial clades or subclades is

the rule there were five instances where multiple specimens

sampled from a locality carry haplotypes belonging to two

separate COI lineages (Figs 1ndash3) In four instances haplo-

types belonging to Bullpen and Clinch Mountain subclades

are syntopic (Fig 3) Most of these locations are in the

Appalachian Valley and Ridge physiographical province

north-east of the French Broad River with one exception in

south-central Tennessee (Fig 1 site 46) At the remaining

locality (site 74 ndash Cave Springs Recreation Area) members

of Clinch Mountain and Cumberland Plateau clades are syn-

topic It is notable that all instances of syntopy are outside

of the southern Blue Ridge

Qualitative patterns of low genetic divergence over large

geographical distances are consistent with spatial expansion

for Bullpen and Clinch Mountain populations found outside

of the southern Blue Ridge Closely related haplotypes in a

weakly supported Bullpen lsquoexpansionrsquo subclade (Fig 3) are

disjunct from remaining Bullpen locations included in this

geographically dispersed lineage are samples from northern

Alabama and southern Tennessee (sites 45 46) northeastern

Tennessee (sites 47 48 50) and southern Illinois (site 44

Fig 1) Illinois sequences are ~1 divergent from other Bull-

pen sequences whereas the Bullpen subclade is on average

gt 7 divergent from samples in the neighbouring Joyce Kil-

mer subclade (Table 4) This pattern of highly dispersed

samples showing minimal genetic divergence is also found in

the Clinch Mountain subclade although disparate samples

here do not form a cohesive mitochondrial lineage (Fig 3)

A sample from New Hampshire (site 80) is approximately

880 km from related populations but NH haplotypes are

lt 1 divergent from related Clinch Mountain haplotypes

Tajimarsquos D values are not statistically significant whereas

values for three subclades are significant for Fursquos Fs(Table 5) This pattern is potentially consistent with a higher

sensitivity of Fursquos Fs to population expansion (Fu 1997

Ramırez-Soriano et al 2008) Mismatch distributional analy-

ses for the Clinch subclade did not converge Mismatch dis-

tributions for three other subclades are unimodal with non-

significant SSD values indicating observed data consistent

with the spatial expansion model of Arlequin (Table 5)

Figure 2 COI RAxML gene tree rooted

using Ozarks clade Clades or groups namedas in text phylogenetic detail for Bullpen

and Clinch Mountain subclades shown inFig 3 Bootstrap values shown only for

primary clades Arrowhead designates sitewith syntopy black ovals indicate clades or

sequences supported as putative species inbPTP analyses Clades impacted by potential

riverine barriers named Location numbersas in Appendix S1 [Colour figure can be

viewed at wileyonlinelibrarycom]


1671


these subclades show small raggedness values also consistent

with expansion Bayesian Skyline Plots for Clinch Bullpen

and Joyce Kilmer all indicate increases in population effective

sizes (see Appendix S3) with temporally older demographic

increases in Clinch and Bullpen consistent with hypothesized

range expansion

DISCUSSION

Cryptic species

The possibility of cryptic species in S cavicolens was argued

in Scheuroonhofer et al (2013) based on concatenated analyses

of three genes We further examined this possibility using

MSC analyses and formal hypothesis testing and these analy-

ses conservatively support the existence of three distinct spe-

cies in this complex We hypothesize that the Cumberland

clade corresponds to S cavicolens as our sampled Red River

Gorge population is geographically close to the type locality

This implies that both the Ozarks and southern Blue Ridge

lineages need to be described as new taxa which is planned

for a separate manuscript Whether these three separate lin-

eages exhibit morphological divergence remains unclear In

Sabacon male characters are most diagnostic but males for

the S cavicolens complex are rare in collections Shear (1975)

illustrated minor differences in male chelicerae and pedipal-

pal patellas for specimens from Ferne Clyffe IL (Bullpen sub-

clade) Mt Mitchell NC (Clinch Mountain subclade) and

Figure 3 COI RAxML gene tree forBullpen and Clinch Mountain subclades

Bootstrap values shown only for primarylineages Arrowheads designate locations

with syntopy black ovals indicate clades orsequences supported as species in bPTP

analyses Clades impacted by potentialriverine barriers named Location numbers

as in Appendix S1 [Colour figure can beviewed at wileyonlinelibrarycom]


1672


Vermont (Clinch Mountain subclade) Adult males are cur-

rently unknown for the Ozarks and Cumberland clades

(holotype S cavicolens specimen is a female) but given the

differences observed in males within the southern Blue Ridge

clade we predict that more substantive diagnostic differences

will exist in males from these obvious lsquocrypticrsquo species

Nuclear-only BFD results actually support more than

three species consistent with the mitochondrial gene tree

and bPTP results However nuclear Pofad results do not

clearly indicate more than three lineages Although it is

possible that genome-scale (eg RADSeq) or more rapidly

evolving nuclear genes will help resolve additional taxa

within the southern Blue Ridge radiation it may also be

that BFD is over-splitting the complex This potential

over-splitting occurs because of the panmixia assumption

of the MSC model which unrealistically assumes that

Figure 4 (a) NeighborNet network reconstructed using standardized Pofad nuclear distances Colours for specimens from southernBlue Ridge mitochondrial subclades as in Figs 2 amp 3 (b) Nuclear-only beast species tree estimated for three-species hypothesis with

posterior probability values at nodes Tree files from three separate beast runs were combined using LogCombiner and a maximumclade credibility tree (burnin = 3000) was produced using TreeAnnotator [Colour figure can be viewed at wileyonlinelibrarycom]


1673


population genetic structure does not exist within species

(Hedin et al 2015)

The conservative three-species hypothesis implies older

vicariance and mostly exclusive allopatry of geographically

disparate taxa found in the Ozarks on the Cumberland Pla-

teau and in the southern Blue Ridge Many potential barri-

ers to gene flow separate these regions including large rivers

(eg Mississippi and Tennessee Rivers) and landscape barri-

ers (eg low elevation habitats of the Appalachian Valley and

Ridge Province) If we accept the Ozarks root placement as

hypothesized in Scheuroonhofer et al (2013) then the overall

biogeographical pattern in eastern Sabacon is similar to the

WgtEgtW pattern observed in phalangodid harvestmen

(Hedin amp Thomas 2010) where in this case eastern Blue

Ridge lineages have lsquoescapedrsquo the mountains and dispersed

back westwards (and northwards)

Riverine barriers in southern Appalachia

Clear evidence for phylogeographical fragmentation is

observed in the southern Blue Ridge species which is parti-

tioned into divergent mostly allopatric mitochondrial sub-

clades Based on phylogeographical patterns found in the

harvestman Fumontana deprehendor Thomas amp Hedin

(2008) proposed that major rivers and intermontane basins

in the southern Blue Ridge act as barriers to gene flow Riv-

ers as barriers to gene flow have been hypothesized in other

cryophilic harvestman taxa (Richart amp Hedin 2013) The

French Broad Little Tennessee and Tuckaseegee Rivers have

been identified as potential barriers in several harvestmen

(Fumontana Bishopella) beetle (Trechus) and salamander

taxa (Desmosgnathus Plethodon) (Kane et al 1990 Weisrock

amp Larson 2006 Thomas amp Hedin 2008 Kozak amp Wiens

2010) in the southern Blue Ridge In most of these prior

studies barriers were inferred qualitatively with phylogenetic

breaks in gene trees coinciding with proposed barriers Here

we took a quantitative approach and identified the Tuck-

aseegee Little Tennessee and French Broad rivers as signifi-

cant barriers The Hiwassee may also play a vicariance role

for the Northern Georgia subclade although we did not for-

mally test this hypothesis because of sparse geographical

sampling adjacent to the Hiwassee (Fig 1)

Exactly how rivers constrain gene flow both historically

and currently remains unclear This is particularly so given

that Sabacon appear to prefer moist habitats and can

frequently be found living at the edge of smaller order

streams higher in drainage basins It is also clear that speci-

mens in spatially expanding lineages can disperse across very

large rivers (eg Tennessee River) In the southern Blue

Ridge where multiple allopatric lineages show breaks coinci-

dent with riverine barriers rivers may not act as absolute

physical barriers per se but rather may reduce or filter levels

of gene flow We hypothesize that the interaction of compet-

itive exclusion (priority effects Case et al 2005) reflecting

ecological niche conservatism with close congeners on oppo-

sites sides of barriers plus reduced gene flow largely main-

tains lineage boundaries If the southern Blue Ridge

subclades actually represent cryptic species then reproductive

interference (low hybrid fitness) might also play a role in

maintaining lineage boundaries (Case et al 2005)

Expansion out of the Blue Ridge

Multiple analyses provide congruent evidence for demo-

graphic and spatial expansion for several Blue Ridge lineages

although both quantitative and spatial qualitative patterns

are most compelling for Bullpen and Clinch Mountain sub-

clades Both lineages have apparent centres of distribution in

the southern Blue Ridge with expansions westwards and

northwards (Fig 1) Informally our field experience suggests

that population densities are lower outside of the Blue Ridge

with populations more scattered in scarcer suitable micro-

habitats We hypothesize that these lower densities implying

less congeneric competition may have played a role in facili-

tating lineage expansion

Northern distributional limits for eastern Sabacon remain

unknown and future geographical sampling in the far north

is needed to determine if these populations represent the

Clinch andor Bullpen subclades versus the Cumberland

clade or perhaps a novel lineage We hypothesize that Pleis-

tocene glacial cycles have impacted spatial expansion dynam-

ics consistent with an lsquoout of Appalachiarsquo refugial model

observed in other taxa (Church et al 2003 Soltis et al

2006 Walker et al 2009 Emerson et al 2010 Herman amp

Bouzat 2016) Some of our sampled Sabacon populations

come from locations covered by Laurentide ice sheets at Last

Glacial Maximum (LGM) (eg southern New Hampshire)

and many species records from northern states (Wisconsin

Michigan Maine Shear 1975) are from locations covered by

LGM ice Our BSP time estimates for population size

Table 5 Neutrality and mismatch results for Blue Ridge COI subclades

COI subclade D P-value Fs P-value Spatial SSD P-value Raggedness index P-value

Northern Georgia 0121 0583 2377 0069 0009 0867 0022 0968

Joyce Kilmer 0215 0557 11304 0002 0007 0271 0006 0901

Clinch Mountain 0137 0520 14455 0003 ndash ndashBullpen 0507 0352 15248 0001 0004 0800 0006 0745

Bullpen expansion 0268 0409 1096 0223 0034 0474 0086 0595

Notes D = Tajimarsquos D F = Fursquos F All values calculated in Arlequin


1674


increase in Bullpen and Clinch post-date the LGM (see

Appendix S3) but we emphasize that these time estimates

rely upon a COI clock rate that may not strictly apply in

Sabacon

Expansion out of the southern Blue Ridge westwards and

northwards has also been found in the harvestmen genus

Bishopella (Hedin amp Thomas 2010) In Bishopella two

mitochondrial clades that bound the Asheville Basin appear

to have diversified within the southern Blue Ridge and sub-

sequently expanded out onto the Cumberland Plateau Spa-

tial patterns in wood-feeding roaches are also strikingly

similar to Sabacon ndash a chromosomal race distributed mostly

east of the French Broad in montane NC also extends

northeastwards into northern Virginia and a chromosomal

race from the central southern Blue Ridge includes disjunct

populations in northern Alabama and Kentucky (Nalepa

et al 2002) Generally these spatial expansions demonstrate

that habitat-specialized cryophilic arthropods can sometimes

cross large river basins perhaps in the context of competi-

tive release

Range expansion in Sabacon has resulted in the secondary

contact and sympatry of previously allopatric lineages always

outside of the southern Blue Ridge (Fig 1) This sympatry is

almost certainly more common than our limited geographic

and specimen sample indicates Without morphological dif-

ferences to diagnose sympatric lineages both nuclear and

mitochondrial data are needed to understand the nature of

interactions associated with this secondary contact Thus far

we have not sampled sufficient nuclear data at enough syn-

topic sites to understand such interactions Our results reveal

one instance of mitonuclear discordance from south-western

VA (Sample OP658 site 74 Fig 1) placed in the mitochon-

drial Clinch Mountain clade but in the nuclear Cumberland

clade (Figs 3 amp 4) We hypothesize that this discordance

results from mitochondrial introgression of Clinch haplo-

types into the Cumberland species This hypothesis is consis-

tent with the easier movement of organellar genes across

lineage boundaries but inconsistent with theory and empiri-

cal data showing that introgression is typically asymmetrical

in directionality from resident to invading lineages (Currat

et al 2008 Toews amp Brelsford 2012) Future research to

understand nuclear gene flow dynamics for additional cases

of sympatry will provide robust tests of cryptic speciation

hypotheses

We discovered a single lineage of closely related mitochon-

drial haplotypes corresponding to the weakly supported Bull-

pen lsquoexpansionrsquo subclade (sites 44 45 46 47 48 50 Fig 3)

disjunct from remaining Bullpen locations outside of the

southern Blue Ridge (Fig 1) This apparently non-random

pattern might be consistent with a lsquofounder take allrsquo dynamic

including the rapid expansion of founder alleles in a low

competition environment followed by high-density blocking

which precludes other alleles (Fig 1 Waters et al 2013)

Again population densities and congeneric competitive inter-

actions are expected to play important roles in this process

Common vicariance rare dispersal

Eastern Sabacon might well be characterized as lsquofragilersquo ndashspecimens collected from suitable microhabitats and placed

in dry vials at room temperature die within minutes (pers

obs MH MM) How such moisture- and temperature-sen-

sitive organisms can achieve larger geographical distribu-

tions even on different continents seems paradoxical Our

data and analyses for the Sabacon cavicolens complex sheds

light on this paradox We view dynamics uncovered within

the southern Blue Ridge lineage as a microcosm of those

seen more broadly in eastern North America and the

northern Hemisphere where cycles of common vicariance

and uncommon long-distance dispersal interact to shape

biogeographical histories

ACKNOWLEDGEMENTS

We thank Fred Coyle Shahan Derkarabetian Ryan Fawcett

Dalton Hedin Lars Hedin Robin Keith Michael Lowder

Jordan Satler Jeff Shultz Jim Starrett and Steven Thomas

for their help with specimen collection Bill Shear provided

specimens for morphological study Jim Starrett helped with

RNA extractions and Dave Carlson assisted with transcrip-

tome assembly We gratefully thank Steve Perlaky for allow-

ing us to borrow a vehicle while in the field and thank

Fred Coyle and the Abbott family for providing excellent

field accommodations Research was funded by grants to

MM from the American Arachnological Society (Vincent

Roth Fund for Systematic Research) and grants to MH

from the US Fish and Wildlife Service (agreement

401814G013) California State University Program for Edu-

cation and Research in Biotechnology (CSUPERB) and the

NSF (DEB 1354558) Shahan Derkarabetian Casey Richart

and two anonymous reviewers provided comments that

improved the manuscript

REFERENCES

Case TJ Holt RD McPeek MA amp Keitt TH (2005)

The community context of speciesrsquo borders ecological and

evolutionary perspectives Oikos 108 28ndash46Church S Kraus J Mitchell J Church D amp Taylor D

(2003) Evidence for multiple Pleistocene refugia in the

postglacial expansion of the Eastern Tiger Salamander

Ambystoma tigrinum tigrinum Evolution 57 372ndash383Currat M Ruedi M Petit R amp Excoffier L (2008) The

hidden side of invasions massive introgression by local

genes Evolution 62 1908ndash1920Derkarabetian S Ledford J amp Hedin M (2011) Genetic

diversification without obvious genitalic morphological

divergence in harvestmen (Opiliones Laniatores Scler-

obunus robustus) from montane sky islands of western

North America Molecular Phylogenetics and Evolution 61

844ndash853


1675


DiDomenico A amp Hedin M (2016) New species in the

Sitalcina sura species group (Opiliones Laniatores Phalan-

godidae) with evidence for a biogeographic link between

California desert canyons and Arizona sky islands Zoo-

Keys 586 1ndash36Drummond AJ Rambaut A Shapiro B amp Pybus OG

(2005) Bayesian coalescent inference of past population

dynamics from molecular sequences Molecular Biology and

Evolution 22 1185ndash1192Drummond AJ Suchard MA Xie D amp Rambaut A

(2012) Bayesian phylogenetics with BEAUti and the

BEAST 17 Molecular Biology and Evolution 29 1969ndash1973

Emerson KJ Merz CR Catchen JM Hohenlohe PA

Cresko WA Bradshaw WE amp Holzapfel CM (2010)

Resolving postglacial phylogeography using high-through-

put sequencing Proceedings of the National Academy of

Sciences USA 107 16196ndash16200Excoffier L (2004) Patterns of DNA sequence diversity and

genetic structure after a range expansion lessons from the

infinite-island model Molecular Ecology 13 853ndash864Excoffier L amp Lischer HEL (2010) Arlequin suite ver 35

a new series of programs to perform population genetics

analyses under Linux and Windows Molecular Ecology

Resources 10 564ndash567Flot JF (2010) SeqPHASE a web tool for interconverting

PHASE inputoutput files and FASTA sequence align-

ments Molecular Ecology Resources 10 162ndash166Fu YX (1997) Statistical tests of neutrality of mutations

against population growth hitchhiking and background

selection Genetics 147 915ndash925Garcıa-Vazquez D amp Ribera I (2016) The origin of wide-

spread species in a poor dispersing lineage (diving beetle

genus Deronectes) Peer J 4e2514

Giribet G amp Sharma PP (2015) Evolutionary biology of

harvestmen (Arachnida Opiliones) Annual Review of

Entomology 60 157ndash175Grummer JA Bryson RW Jr amp Reeder TW (2014) Spe-

cies delimitation using Bayes factors simulations and

application to the Sceloporus scalaris species group (Squa-

mata Phrynosomatidae) Systematic Biology 63 119ndash133Harpending HC (1994) Signature of ancient population

growth in a low-resolution mitochondrial DNA mismatch

distribution Human Biology 66 591ndash600Harvey MS (2002) Short-range endemism among the Aus-

tralian fauna some examples from non-marine environ-

ments Invertebrate Systematics 16 555ndash570Harvey MS Rix MG Framenau VW Hamilton ZR

Johnson MS Teale RJ Humphreys G amp Humphreys

WF (2011) Protecting the innocent studying short range

endemic taxa enhances conservation outcomes Inverte-

brate Systematics 25 1ndash10Hedin M (1997) Speciational history in a diverse clade of

habitat-specialized spiders (Araneae Nesticidae Nesticus)

inferences from geographic-based sampling Evolution 51

1929ndash1945

Hedin M (2001) Molecular insights into species phylogeny

biogeography and morphological stasis in the ancient spi-

der genus Hypochilus (Araneae Hypochilidae) Molecular

Phylogenetics and Evolution 18 238ndash251Hedin M amp Thomas SM (2010) Molecular systematics of

eastern North American Phalangodidae (Arachnida Opil-

iones Laniatores) demonstrating convergent morphologi-

cal evolution in caves Molecular Phylogenetics and

Evolution 54 107ndash121Hedin M amp Wood DL (2002) Genealogical exclusivity in

geographically proximate populations of Hypochilus thor-

elli Marx (Araneae Hypochilidae) on the Cumberland

Plateau of North America Molecular Ecology 11 1975ndash1988

Hedin M Carlson D amp Coyle F (2015) Sky island diversi-

fication meets the multispecies coalescent ndash divergence in

the spruce-fir moss spider (Microhexura montivaga Ara-

neae Mygalomorphae) on the highest peaks in southern

Appalachia Molecular Ecology 24 3467ndash3484Hendrixson BE amp Bond JE (2005) Testing species bound-

aries in the Antrodiaetus unicolor complex (Araneae Myga-

lomorphae Antrodiaetidae) ldquoparaphylyrdquo and cryptic

diversity Molecular Phylogenetics and Evolution 36 405ndash416

Herman TA amp Bouzat JL (2016) Range-wide phylogeog-

raphy of the four-toed salamander out of Appalachia and

into the glacial aftermath Journal of Biogeography 43

666ndash678Ho SWY amp Shapiro B (2011) Skyline-plot methods for

estimating demographic history from nucleotide

sequences Molecular Ecology Resources 11 423ndash434Huson DH amp Bryant D (2006) Application of phyloge-

netic networks in evolutionary studies Molecular Biology

amp Evolution 23 254ndash267Ivany LC Patterson WP amp Kyger CL (2000) Cooler

winters as a possible cause of mass extinctions at the

EoceneOligocene boundary Nature 407 887ndash890Joly S amp Bruneau A (2006) Incorporating allelic variation

for reconstructing the evolutionary history of organisms

from multiple genes an example from Rosa in North

America Systematic Biology 55 623ndash636Jombart T amp Dray S (2008) Package lsquoadephylorsquo exploratory

analyses for the phylogenetic comparative method Version

11-6 Available at httpscranr-projectorgpackage=ade

phylo

Kane TC Barr TC Jr amp Stratton GE (1990) Genetic

patterns and population structure in Appalachian Trechus

of the vandykei group (Coleoptera Carabidae) Brim-

leyana 16 133ndash150Kass RE amp Raftery AE (1995) Bayes factors Journal of the

American Statistical Association 90 773ndash795Kearse M Moir R Wilson A Stones-Havas S Cheung

M Sturrock S Buxton S Cooper A Markowitz S

Duran C Thierer T Ashton B Mentjies P amp Drum-

mond A (2012) Geneious Basic an integrated and

extendable desktop software platform for the organization


1676


and analysis of sequence data Bioinformatics 28 1647ndash1649

Keith R amp Hedin M (2012) Extreme mitochondrial popu-

lation subdivision in southern Appalachian paleoendemic

spiders (Araneae Hypochilidae Hypochilus) with implica-

tions for species delimitation Journal of Arachnology 40

167ndash181Kozak KH amp Wiens JJ (2010) Niche conservatism drives

elevational diversity patterns in Appalachian salamanders

The American Naturalist 176 40ndash54Kozak KH Blaine RA amp Larson A (2006) Gene lineages

and eastern North American palaeodrainage basins phylo-

geography and speciation in salamanders of the Eurycea

bislineata species complex Molecular Ecology 15 191ndash207Lanfear R Calcott B Ho SY amp Guindon S (2012) Parti-

tionFinder combined selection of partitioning schemes

and substitution models for phylogenetic analyses Molecu-

lar Biology and Evolution 29 1695ndash1701Lartillot N amp Philippe H (2006) Computing Bayes factors

using thermodynamic integration Systematic Biology 55

195ndash207Lemmon EM Lemmon AR amp Cannatella DC (2007)

Geological and climatic forces driving speciation in the

continentally distributed trilling chorus frogs (Pseudacris)

Evolution 61 2086ndash2103Martens J (2015) Sabacon Simon 1879 in the Palaearctic a

survey of new and known species from France Nepal

India China Russia and Japan (Arachnida Opiliones

Sabaconidae) In Hartmann M amp Weipert J (Eds) Bio-

diversitat und Naturaustattung im Himalaya V Erfurt pp

167ndash210Nalepa CA Luykx P Klass K-D amp Deitz LL (2002)

Distribution of karyotypes of the Cryptocercus punctulatus

species complex (Dictyoptera Cryptocercide) in the south-

ern Appalachians relation to habitat and history Annals

Entomological Society of America 95 276ndash287Oksanen J Blanchet FG Kindt R Legendre P Minchin

PR OrsquoHara RB Simpson GL Solymos P Stevens

MHH amp Wagner H (2015) Package lsquoveganrsquo Community

ecology package Version 2-2 Available at httpscranr-

projectorgpackage=vegan

Papadopoulou A Anastasiou I amp Vogler AP (2010)

Revisiting the insect mitochondrial molecular clock the

mid-Aegean trench calibration Molecular Biology and Evo-

lution 27 1659ndash1672R Core Team (2014) R A language and environment for sta-

tistical computing R Foundation for Statistical Computing

Vienna Austria Available at httpwwwR-projectorg

(last accessed Dec 15 2016)

Rambaut A Suchard MA Xie D amp Drummond AJ

(2014) Tracer v1 6 Computer program and documenta-

tion distributed by the author Available at httpbeast

bio ed ac ukTracer

Ramırez-Soriano A Ramos-Onsins SE Rozas J Calafell

F amp Navarro A (2008) Statistical power analysis of neu-

trality tests under demographic expansions contractions

and bottlenecks with recombination Genetics 179 555ndash567

Richart C amp Hedin M (2013) Three new species in the

harvestmen genus Acuclavella (Opiliones Dyspnoi Ischy-

ropsalidoidea) including description of male Acuclavella

quattuor Shear 1986 ZooKeys 311 19ndash68Scheuroonhofer AL McCormack M Tsurusaki N Martens J

amp Hedin M (2013) Molecular phylogeny of the harvest-

men genus Sabacon (Arachnida Opiliones Dyspnoi)

reveals multiple Eocene-Oligocene intercontinental disper-

sal events in the Holarctic Molecular Phylogenetics and

Evolution 66 303ndash315Shear WA (1975) The opilionid genera Sabacon and Tomi-

comerus in America (Opiliones Troguloidea Ischyropsali-

dae) Journal of Arachnology 3 5ndash29Silvestro D amp Michalak I (2012) raxmlGUI a graphical

front-end for RAxML Organisms Diversity amp Evolution

12 335ndash337Soltis DE Morris AB McLachlan S Manos PS amp Sol-

tis PS (2006) Comparative phylogeography of ungla-

ciated eastern North America Molecular Ecology 15

4261ndash4293Stamatakis A (2006) RAxML-VI-HPC maximum likeli-

hood-based phylogenetic analyses with thousands of taxa

and mixed models Bioinformatics 22 2688ndash2690Stamatakis A (2014) RAxML version 8 a tool for phyloge-

netic analysis and post-analysis of large phylogenies Bioin-

formatics 30 1312ndash1313Stephens M amp Donnelly P (2003) A comparison of Baye-

sian methods for haplotype reconstruction from popula-

tion genotype data American Journal of Human Genetics

73 1162ndash1169Stephens M Smith N amp Donnelly P (2001) A new statis-

tical method for haplotype reconstruction from popula-

tion data American Journal of Human Genetics 68 978ndash989

Tajima F (1989) Statistical method for testing the neutral

mutation hypothesis by DNA polymorphism Genetics

123 585ndash595Tamura K Stecher G Peterson D Filipski A amp Kumar

S (2013) MEGA6 molecular evolutionary genetics analy-

sis version 60 Molecular Biology amp Evolution 30 2725ndash2729

Thomas SM amp Hedin M (2008) Multigenic phylogeo-

graphic divergence in the paleoendemic southern Appala-

chian opilionids Fumontana deprehendor Shear (Opiliones

Laniatores Triaenonychidae) Molecular Phylogenetics and

Evolution 46 645ndash658Toews DP amp Brelsford A (2012) The biogeography of

mitochondrial and nuclear discordance in animals Molec-

ular Ecology 21 3907ndash3930Walker MJ Stockman AK Marek PE amp Bond JE

(2009) Pleistocene glacial refugia in the Appalachian

Mountains and coastal plain evidence from a unique

mitochondrial phylogeographic pattern in the millipede

genus Narceus BMC Evolutionary Biology 9 25


1677


Waters JM Fraser CI amp Hewitt GM (2013) Founder

takes all density-dependent processes structure biodiver-

sity Trends in Ecology and Evolution 28 78ndash85Weisrock DW amp Larson A (2006) Testing hypotheses of

speciation in the Plethodon jordani species complex with

allozymes and mitochondrial DNA sequence Biological

Journal of the Linnaean Society 89 20ndash51Xie WG Lewis PO Fan Y Kuo L amp Chen MH

(2011) Improving marginal likelihood estimation for Baye-

sian phylogenetic model selection Systematic Biology 60

150ndash160Zhang J Kapli P Pavlidis P amp Stamatakis A (2013) A

general species delimitation method with applications to

phylogenetic placements Bioinformatics 29 2869ndash2876

SUPPORTING INFORMATION

Additional Supporting Information may be found in the

online version of this article

Appendix S1 Sample information including GenBank

numbers

Appendix S2 PCR conditions and marker development

Appendix S3 COI Bayesian skyline plots

DATA ACCESSIBILITY

Aligned matrices have been deposited at Dryad (doi105061

dryadmk32s)

BIOSKETCHES

The Hedin lab studies the evolution and diversity of arach-

nids including spiders and harvestmen with a biogeographi-

cal emphasis in the southern Appalachians

Author contributions MH implemented the study design

directed data collection conducted analyses and wrote the

manuscript MM implemented the study design collected

specimens and DNA sequence data

Editor Brent Emerson


1678


INTRODUCTION

Short-range endemic (SRE) taxa defined as species or

higher taxa with naturally small geographical distributions

(Harvey 2002 Harvey et al 2011 Keith amp Hedin 2012)

present a biogeographical paradox SRE taxa are often dis-

tributed as arrays of parapatric or allopatric taxa consis-

tent with a biogeographical history dominated by low

dispersal and vicariance However the larger encompassing

geographical distribution of such allo- or parapatric arrays

implies some level of dispersal during the history of the

lineage (eg Garcıa-Vazquez amp Ribera 2016) An illustra-

tive example is the North American spider genus Hypochi-

lus with five species in the southern Appalachians two

species in the southern Rockies and three species in mon-

tane California All Hypochilus species are habitat-specia-

lized short-range endemics where allopatry prevails even

within montane regions (Hedin 2001) Moreover available

genetic data for individual species suggests extreme popula-

tion genetic fragmentation (Hedin amp Wood 2002 Keith amp

Hedin 2012) Overall vicariance clearly dominates the his-

tory of this SRE taxon But how do we explain the overall

larger geographical distribution in California or in Appala-

chia or in North America Biogeographers are left to infer

historical dispersal based on indirect evidence sometimes

despite a complete lack of evidence for modern-day dis-

persal abilities in such taxa This is a problem with

ancient and rare dispersal ndash dispersal must have happened

likely under special environmental conditions but details

have been lost to history

Biogeographical patterns observed in the harvestman

genus Sabacon seem consistent with a history dominated by

vicariance with occasional rare dispersal Sabacon includes

over 50 described species known from disjunct geographical

centres in the Northern Hemisphere (North America west-

ern Europe eastern Siberia Nepal Korea and Japan see

Scheuroonhofer et al 2013 Martens 2015) Sabacon prefer

moist cool microhabitats in habitats such as caves upland

forests under rocks or woody debris and under rocks near

streams (summarized in Scheuroonhofer et al 2013) It appears

that physiological constraints to cryophilic microhabitats

restrict gene flow and higher level phylogenetic results are

consistent with this assertion Larger phylogenetic clades

correspond to geographical centres of endemism on differ-

ent continents each of these centres with arrays of closely

related SRE taxa occurring in allopatry or parapatry How-

ever interwoven in this vicariance-dominated history is evi-

dence for rare long-distance dispersal Because harvestmen

are not known to be phoretic (Giribet amp Sharma 2015)

this dispersal must have occurred via intrinsic short-range

movements over hundreds or thousands of generations In

Sabacon an apparent pulse of trans-continental dispersal

events is temporally coincident with special environmental

conditions of the lsquoicehousersquo Eocene-Oligocene transition

(Ivany et al 2000 Scheuroonhofer et al 2013) These ancient

dispersal events occurred in multiple lineages and are

inferred based on the combination of geographical distribu-

tion time-calibrated phylogenies and algorithmic biogeo-

graphical analyses that indirectly imply dispersal

(Scheuroonhofer et al 2013)

This paper focuses on a single wide-ranging Sabacon

species (S cavicolens Packard 1884) from the eastern Uni-

ted States This taxon is related to species from southern

Europe with the lineage that ultimately led to S cavicolens

hypothesized to have dispersed to eastern North America

near the Eocene-Oligocene transition (Scheuroonhofer et al

2013) Sabacon cavicolens occurs most commonly in moist

rocky habitats in the richly forested mountains of the

southern Appalachians but also includes more northern

populations in previously glaciated regions (eg Wisconsin

Michigan Maine etc) and disjunct populations in the

Ozark highlands (Fig 1 Shear 1975) We hypothesize that

S cavicolens displays a lsquocommon vicariance rare dispersalrsquo

biogeographical history and as such represents a micro-

cosm of biogeographical patterns observed more broadly

within Sabacon However because the timeframe is more

recent inferred patterns provide more direct and thus

more powerful evidence for biogeographical processes For

example intraspecific phylogeographic data can be used to

more directly infer demographic expansion events such

demographic information is lost in comparisons at deeper

phylogenetic levels

Geographical fragmentation has promoted SRE specia-

tion in many terrestrial lineages that occupy the mountains

of the southern Appalachians Vertebrate examples are

numerous in salamanders (eg Weisrock amp Larson 2006

Kozak amp Wiens 2010 Herman amp Bouzat 2016) whereas

arthropod examples include other harvestmen (Thomas amp

Hedin 2008 Hedin amp Thomas 2010) spiders (Hedin

1997 2001 Hendrixson amp Bond 2005 Keith amp Hedin

2012 Hedin et al 2015) and others Also intraspecific

phylogeographical fragmentation is evident within more

wide-ranging species with many taxa including divergent

populations on distinct massifs separated by lowland river-

ine barriers For example a conspicuous lowland riverine

barrier is the Asheville Basin which includes the French

Broad River This relatively dry lowland is unsuitable for

habitat-specialized upland species and many phylogenetic

and phylogeographic studies have implicated this barrier

(eg Thomas amp Hedin 2008 Kozak amp Wiens 2010 Hedin

et al 2015) More generally Thomas amp Hedin (2008)

hypothesized that multiple riverine barriers in the southern

Blue Ridge physiographical province may act to promote

divergence a pattern also seen in salamanders (Herman amp

Bouzat 2016)

Climatic stability through time and in particular a lack of

glaciation in the southernmost mountains during the Pleis-

tocene has also promoted high diversity in the southern

Blue Ridge However many lineages have ultimately

expanded out of these refugial mountains (reviewed in Soltis

et al 2006 Herman amp Bouzat 2016) with genetic studies

revealing patterns consistent with post-glacial range


1666
























time-scales





















1667



analyses















































































Primer combo

name

Aligned lengthno

sequences

Substitution


COI 954168 266 0081

28S 114917 HKY+I+G 17 0006

EF1-a 66621 K80+I 18 0010




Zinc finger protein





1668

















































significance






















Clinch Mountain (8)














1669






(99000 replicates)



















tent results

RESULTS























clades (Fig 1)




























Riverine barriers









Bullpen 0134 0069 0076 0125 0018 54 91 00177






1670









(Table 3)


















































1671







range expansion

DISCUSSION

Cryptic species
























1672





















1673



(Hedin et al 2015)
























































































Joyce Kilmer 0215 0557 11304 0002 0007 0271 0006 0901





1674





Sabacon
















tive release
























hypotheses

























ACKNOWLEDGEMENTS



















REFERENCES













844ndash853


1675












































1929ndash1945



































phylo











1676












































bio ed ac ukTracer















































1677

















numbers



DATA ACCESSIBILITY


dryadmk32s)

BIOSKETCHES










1678
























time-scales





















1667



analyses















































































Primer combo

name

Aligned lengthno

sequences

Substitution


COI 954168 266 0081

28S 114917 HKY+I+G 17 0006

EF1-a 66621 K80+I 18 0010




Zinc finger protein





1668

















































significance






















Clinch Mountain (8)














1669






(99000 replicates)



















tent results

RESULTS























clades (Fig 1)




























Riverine barriers









Bullpen 0134 0069 0076 0125 0018 54 91 00177






1670









(Table 3)


















































1671







range expansion

DISCUSSION

Cryptic species
























1672





















1673



(Hedin et al 2015)
























































































Joyce Kilmer 0215 0557 11304 0002 0007 0271 0006 0901





1674





Sabacon
















tive release
























hypotheses

























ACKNOWLEDGEMENTS



















REFERENCES













844ndash853


1675












































1929ndash1945



































phylo











1676












































bio ed ac ukTracer















































1677

















numbers



DATA ACCESSIBILITY


dryadmk32s)

BIOSKETCHES










1678



analyses















































































Primer combo

name

Aligned lengthno

sequences

Substitution


COI 954168 266 0081

28S 114917 HKY+I+G 17 0006

EF1-a 66621 K80+I 18 0010




Zinc finger protein





1668

















































significance






















Clinch Mountain (8)














1669






(99000 replicates)



















tent results

RESULTS























clades (Fig 1)




























Riverine barriers









Bullpen 0134 0069 0076 0125 0018 54 91 00177






1670









(Table 3)


















































1671







range expansion

DISCUSSION

Cryptic species
























1672





















1673



(Hedin et al 2015)
























































































Joyce Kilmer 0215 0557 11304 0002 0007 0271 0006 0901





1674





Sabacon
















tive release
























hypotheses

























ACKNOWLEDGEMENTS



















REFERENCES













844ndash853


1675












































1929ndash1945



































phylo











1676












































bio ed ac ukTracer















































1677

















numbers



DATA ACCESSIBILITY


dryadmk32s)

BIOSKETCHES










1678

















































significance






















Clinch Mountain (8)














1669






(99000 replicates)



















tent results

RESULTS























clades (Fig 1)




























Riverine barriers









Bullpen 0134 0069 0076 0125 0018 54 91 00177






1670









(Table 3)


















































1671







range expansion

DISCUSSION

Cryptic species
























1672





















1673



(Hedin et al 2015)
























































































Joyce Kilmer 0215 0557 11304 0002 0007 0271 0006 0901





1674





Sabacon
















tive release
























hypotheses

























ACKNOWLEDGEMENTS



















REFERENCES













844ndash853


1675












































1929ndash1945



































phylo











1676












































bio ed ac ukTracer















































1677

















numbers



DATA ACCESSIBILITY


dryadmk32s)

BIOSKETCHES










1678






(99000 replicates)



















tent results

RESULTS























clades (Fig 1)




























Riverine barriers









Bullpen 0134 0069 0076 0125 0018 54 91 00177






1670









(Table 3)


















































1671







range expansion

DISCUSSION

Cryptic species
























1672





















1673



(Hedin et al 2015)
























































































Joyce Kilmer 0215 0557 11304 0002 0007 0271 0006 0901





1674





Sabacon
















tive release
























hypotheses

























ACKNOWLEDGEMENTS



















REFERENCES













844ndash853


1675












































1929ndash1945



































phylo











1676












































bio ed ac ukTracer















































1677

















numbers



DATA ACCESSIBILITY


dryadmk32s)

BIOSKETCHES










1678









(Table 3)


















































1671







range expansion

DISCUSSION

Cryptic species
























1672





















1673



(Hedin et al 2015)
























































































Joyce Kilmer 0215 0557 11304 0002 0007 0271 0006 0901





1674





Sabacon
















tive release
























hypotheses

























ACKNOWLEDGEMENTS



















REFERENCES













844ndash853


1675












































1929ndash1945



































phylo











1676












































bio ed ac ukTracer















































1677

















numbers



DATA ACCESSIBILITY


dryadmk32s)

BIOSKETCHES










1678







range expansion

DISCUSSION

Cryptic species
























1672





















1673



(Hedin et al 2015)
























































































Joyce Kilmer 0215 0557 11304 0002 0007 0271 0006 0901





1674





Sabacon
















tive release
























hypotheses

























ACKNOWLEDGEMENTS



















REFERENCES













844ndash853


1675












































1929ndash1945



































phylo











1676












































bio ed ac ukTracer















































1677

















numbers



DATA ACCESSIBILITY


dryadmk32s)

BIOSKETCHES










1678





















1673



(Hedin et al 2015)
























































































Joyce Kilmer 0215 0557 11304 0002 0007 0271 0006 0901





1674





Sabacon
















tive release
























hypotheses

























ACKNOWLEDGEMENTS



















REFERENCES













844ndash853


1675












































1929ndash1945



































phylo











1676












































bio ed ac ukTracer















































1677

















numbers



DATA ACCESSIBILITY


dryadmk32s)

BIOSKETCHES










1678



(Hedin et al 2015)
























































































Joyce Kilmer 0215 0557 11304 0002 0007 0271 0006 0901





1674





Sabacon
















tive release
























hypotheses

























ACKNOWLEDGEMENTS



















REFERENCES













844ndash853


1675












































1929ndash1945



































phylo











1676












































bio ed ac ukTracer















































1677

















numbers



DATA ACCESSIBILITY


dryadmk32s)

BIOSKETCHES










1678





Sabacon
















tive release
























hypotheses

























ACKNOWLEDGEMENTS



















REFERENCES













844ndash853


1675












































1929ndash1945



































phylo











1676












































bio ed ac ukTracer















































1677

















numbers



DATA ACCESSIBILITY


dryadmk32s)

BIOSKETCHES










1678












































1929ndash1945



































phylo











1676












































bio ed ac ukTracer















































1677

















numbers



DATA ACCESSIBILITY


dryadmk32s)

BIOSKETCHES










1678












































bio ed ac ukTracer















































1677

















numbers



DATA ACCESSIBILITY


dryadmk32s)

BIOSKETCHES










1678

















numbers



DATA ACCESSIBILITY


dryadmk32s)

BIOSKETCHES










1678


Date post:	09-Jun-2020
Category:	Documents
Upload:	others
View:	3 times
Download:	0 times

Biogeographical evidence for common vicariance and rare ...€¦ · hypothesized to have dispersed...

Documents