Molecular Phylogenetics and Evolution 66 (2013) 303–315
Contents lists available at SciVerse ScienceDirect
Molecular Phylogenetics and Evolution
journal homepage: www.elsevier.com/locate/ympev
Molecular phylogeny of the harvestmen genus Sabacon (Arachnida: Opiliones:
Dyspnoi) reveals multiple Eocene–Oligocene intercontinental dispersal events
in the Holarctic
Axel L. Schönhofer a,⇑, Maureen McCormack b, Nobuo Tsurusaki c, Jochen Martens d, Marshal Hedin a
a
Department of Biology, San Diego State University, San Diego, CA 92182-4614, USA
Wisconsin State Lab of Hygiene, University of Wisconsin-Madison, Madison, WI 53706, USA
Laboratory of Biology, Faculty of Regional Sciences, Tottori University, Tottori 680-8551, Japan
d
Institute of Zoology, Johannes Gutenberg University Mainz, D-55099 Mainz, Germany
b
c
a r t i c l e
i n f o
Article history:
Received 9 June 2012
Revised 6 September 2012
Accepted 3 October 2012
Available online 17 October 2012
Keywords:
Molecular systematics
Holarctic
Intercontinental dispersal
Divergence time estimation
Ancestral range reconstruction
Parametric biogeographic modelling
a b s t r a c t
We investigated the phylogeny and biogeographic history of the Holarctic harvestmen genus Sabacon,
which shows an intercontinental disjunct distribution and is presumed to be a relatively old taxon.
Molecular phylogenetic relationships of Sabacon were estimated using multiple gene regions and Bayesian inference for a comprehensive Sabacon sample. Molecular clock analyses, using relaxed clock models
implemented in BEAST, are applied to date divergence events. Biogeographic scenarios utilizing S-DIVA
and Lagrange C++ are reconstructed over sets of Bayesian trees, allowing for the incorporation of phylogenetic uncertainty and quantification of alternative reconstructions over time. Four primary well-supported subclades are recovered within Sabacon: (1) restricted to western North America; (2) eastern
North American S. mitchelli and sampled Japanese taxa; (3) a second western North American group
and taxa from Nepal and China; and (4) eastern North American S. cavicolens with sampled European Sabacon species. Three of four regional faunas (wNA, eNA, East Asia) are thereby non-monophyletic, and
three clades include intercontinental disjuncts. Molecular clock analyses and biogeographic reconstructions support nearly simultaneous intercontinental dispersal coincident with the Eocene–Oligocene transition. We hypothesize that biogeographic exchange in the mid-Tertiary is likely correlated with the
onset of global cooling, allowing cryophilic Sabacon taxa to disperse within and among continents. Morphological variation supports the divergent genetic clades observed in Sabacon, and suggests that a taxonomic revision (e.g., splitting Sabacon into multiple genera) may be warranted.
Ó 2012 Elsevier Inc. All rights reserved.
1. Introduction
1.1. Biogeographic studies of the Holarctic
Over the past 20 years the biogeographic history of the Northern Hemisphere has been the subject of several detailed studies
and meta-analyses (e.g. Donoghue and Smith, 2004; Enghoff,
1995; Sanmartín et al., 2001). The Holarctic is the largest planetary
ecozone, spanning a large proportion of the Northern Hemisphere
including parts of Europe, Asia, and North America, and many subdivisions exist according to ecology, geography, floristic and faunistic characteristics. For most biogeographic studies spanning
the Holarctic, four major infraregions have been applied (e.g.
Donoghue and Smith, 2004; Enghoff, 1995; Sanmartín et al.,
⇑ Corresponding author.
E-mail addresses: [email protected] (A.L. Schönhofer), mccmaureen@
gmail.com (M. McCormack), [email protected] (N. Tsurusaki), [email protected]
(J. Martens), [email protected] (M. Hedin).
1055-7903/$ - see front matter Ó 2012 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.ympev.2012.10.001
2001) corresponding to zones of temperate forest endemism in
the western Palearctic (Europe), eastern Palearctic (eastern Asia)
and the Nearctic (eastern and western North America; see Fig. 1
in Donoghue and Smith, 2004).
The temporal connectivity between the four major Holarctic
infraregions follows a complex reticulate pattern (Smith et al.,
1994; summarized in Sanmartín et al., 2001). Mid- to late Cretaceous (100–80 Ma) geographic settings were characterized by the
connection of Europe with eastern North America, and Asia with
western North America. At the end of the Cretaceous (70–65 Ma)
the American mid-Continental Seaway closed and the North American continent appeared as it does today. The opening of the Atlantic (90 Ma) isolated North America and Europe, but major land
bridges still connected these two continents, most importantly,
the Thulean Bridge around 55–50 Ma. Subsequent land bridges
were apparently less suitable for biotic interchange, especially for
warm-adapted species (McKenna, 1983; Sanmartín et al., 2001;
Tiffney, 1985). North America and Asia were repeatedly connected
via the Bering Bridge, with biotic interchange thought to be
304
A.L. Schönhofer et al. / Molecular Phylogenetics and Evolution 66 (2013) 303–315
Fig. 1. Sabacon habitus images. (A) S. simoni, female, Italy, Passo di Montevaca; (B) S. viscayanus ramblaianus, female, France, Pyrenees, Lac du Bethmale; (C) S. occidentalis,
juvenile, USA, CA, Humboldt Co., Berry Glenn; (D) S. paradoxus, female, France, Olargues. All pictures by A.L.S.; and (E) approximate distribution of Sabacon indicated by
outlined areas, with green circles indicating localities of sequenced specimens.
moderated mainly by climatic settings. A boreotropic forest connected eastern Asia and eastern North America during the warm
and humid Eocene, ultimately disrupted by global cooling and
orogeny of the Rocky Mountains at the end of this epoch
(35 Ma). In the Pleistocene (1.5–1 Ma) the colder climate restricted
Beringian interchange to mostly cold-adapted boreal species.
Development of novel parametric methods in biogeographic
studies allows for a more comprehensive reconstruction of complex Holarctic processes than previous methods (Donoghue and
Smith, 2004; Sanmartín et al., 2001 and summary herein). Most
Holarctic studies implementing parametric biogeographic methods
with molecular phylogenetic data focus on plant taxa, with animal
studies still scarce (e.g., Burbrink and Lawson, 2007; Vieites et al.,
2007). Comparisons of codistributed plant and animal taxa often
reveal contrasting distributional and biogeographic patterns at
comparable phylogenetic levels. For example, the degree and
direction of dispersal events between Holarctic infraregions varies
considerably, showing extensive interchange between floral elements of eastern North America and eastern Asia, whereas animal
disjuncts across these regions are rare (Donoghue and Smith, 2004;
Sanmartín et al., 2001). Holarctic plant genera are generally
widespread with few plant families endemic to the Holarctic.
Conversely, many Holarctic arthropod taxa show family-level
endemism at continental or smaller geographic scales. The
increased structuring seen in arthropods perhaps offers further
insight into the complex Holarctic biogeography, particularly when
conducted on groups considered as model taxa for biogeographic
analyses.
1.2. Harvestmen as model taxa for historical biogeography
Recent phylogenetic studies within the arachnid order Opiliones have shown that the diversification history of many groups
is tightly interconnected with Earth history processes (e.g., Boyer
et al., 2007; Giribet et al., 2010; Hedin et al., 2012; Schönhofer
and Martens, 2010). The integration of phylogeny and biogeography has therefore become necessary to delineate centers of Opiliones endemism and biodiversity, as illustrated by the wellstudied mite harvestmen (Boyer and Giribet, 2007; Boyer et al.,
2007; Giribet et al., 2012). Despite these many biogeographic contributions, few Opiliones groups with primarily Holarctic distributions have been the focus of phylogenetic or biogeographic
A.L. Schönhofer et al. / Molecular Phylogenetics and Evolution 66 (2013) 303–315
attention. Relationships of Holarctic Sironidae have been difficult
to resolve (Giribet and Shear, 2010), and this group seems mostly
absent from continental Asia. Species of Caddo (Eupnoi) provide insight into recent biogeographic exchange between eastern North
America and eastern Asia (Shultz and Regier, 2009). Regional Holarctic clades within the long-legged Phalangioidea are still being
defined (Hedin et al., 2012).
Members of the harvestmen suborder Dyspnoi are a promising
group for studies of Holarctic biogeography. This clade includes
about 350 species in approximately 40 genera, all with north temperate distributions. Most Dyspnoi genera are restricted to single
continents (Shear, 2010) while taxa above this rank are distributed
across continents. An exception is the genus Sabacon (Fig. 1), which
spans a larger distributional area than any other Dyspnoi genus,
with centers of endemism in eastern Asia (Martens, 1972, 1989;
Suzuki, 1974; Tsurusaki and Song, 1993a), eastern and western
North America (Shear, 1975), and south-western Europe (Martens,
1983; Fig. 1E). In these areas, Sabacon species show strong preferences for permanently cool and moist habitats, occurring in various
sheltered microhabitats in coniferous or mixed forests, in deep alpine stony gravel, and in caves. Although a phylogenetic analysis of
all Sabacon taxa has never been conducted, morphological similarities between European and eastern North American taxa (Martens,
1983; Shear, 1975, 1986), as well as between North American and
Asian species (Martens, 1983; Cokendolpher, 1984; Shear, 1975),
suggest possible intercontinental biogeographic connections.
Existing biogeographic patterns imply that diversification in
Sabacon must be relatively old. Fossil Sabacon dating to the Eocene
are known from European amber (Dunlop, 2006), indicating both
an old age and morphological stasis for this lineage. Lopez et al.
(1980) mentioned the cave-dwelling Sabacon paradoxus as an
example of a Tertiary relict, and Thaler (1976) discussed S. simoni
as a Holarctic relict taxon. Sabacon includes no transcontinental
species, and does not feature typical boreal species, as found for
example in the harvestmen suborder Eupnoi (e.g., Mitopus morio).
The pattern of disjunct distributions within species groups, but
narrow endemism of most Sabacon species, suggests generally poor
dispersal abilities for members of the genus. Dispersal across open
oceans is unlikely for Sabacon, and connection of landmasses can
be assumed to be the only means for intercontinental biogeographic exchange.
1.3. Applying novel biogeographic methods to Opiliones
The study of biogeography, in combining distributional patterns
of extant taxa and the reconstruction of their phylogenetic history,
has undergone considerable changes over the last decade. Utilizing
phylogenetic information derived from molecular data, novel modelling approaches using parsimony or likelihood reconstructions
over sets of plausible trees (Lagrange: Ree and Smith, 2008; Smith,
2009; S-DIVA: Yu et al., 2010) are replacing simple area cladograms and purely parsimony-based methods. These new methodologies are able to incorporate age estimates and phylogenetic
uncertainty, account for missing phylogenetic data, and add statistical confidence to reconstructions (Ree and Smith, 2008; Smith,
2009). While molecular phylogenetic analyses are becoming
increasingly common within Opiliones, biogeographic results remain largely descriptive, and statistical approaches have been restricted to estimating divergence times (Boyer and Giribet, 2007;
Boyer et al., 2007; Derkarabetian et al., 2010). With few exceptions
(e.g., Giribet et al., 2012; Sharma and Giribet, 2011), new biogeographic methodologies have yet to be incorporated into Opiliones
research.
The current study focuses on the systematics and biogeographic
history of Sabacon based on sampling of major lineages including
most described species diversity in North America and Europe,
305
and representative species sampling from Japan, Nepal, and China.
We use molecular phylogenetics to investigate relationships within Sabacon, and evaluate the systematic position of Sabacon within
the Dyspnoi. Phylogenetic information is used to infer ancestral
distribution areas of Sabacon clades, as well as the timing and
directionality of Holarctic intercontinental dispersal. We also summarize differences in Sabacon morphology as these relate to recovered molecular lineages.
2. Material and methods
2.1. Taxon sampling
Sabacon species are relatively rare and sometimes difficult to
collect. Many species are only known from high elevation habitats
or caves, and many original descriptions are based upon single or
few specimens. Nevertheless we have obtained a comprehensive
sample of North American and European taxa, and a representative
sample of Asian taxa. Overall, our sample covers much of the described morphological and geographical diversity in Sabacon (see
Table 1). Specimens were collected in the field via manual search
and preserved in 100% EtOH, or collected into 80% and later transferred to 100%. Material was then stored at minus 80 °C until DNA
extraction. Some older samples from Asia were retrieved from 70%
EtOH collections. Adult specimens were used whenever possible,
but due to the late-maturing seasonal phenology of Sabacon, inclusion of juveniles in this study was also necessary. The identity of
juveniles was corroborated by morphological comparison with
adults from the respective localities, partly loaned from the California Academy of Science (CAS), and from our own reference collections (San Diego State University Terrestrial Arthropod collection,
University of Mainz Opilionid collections of A.L.S and J.M.).
2.2. Gene data collection, sequence alignment and phylogenetic
analysis
Genomic DNA was extracted from leg tissue using the Qiagen
DNeasy kit. The polymerase chain reaction (PCR) was used to
amplify the following gene fragments: 28S rRNA (28S), using the
primers ZX1, ZR2 (Mallatt and Sullivan, 1998), and the newly
developed
ZX1elong
(50 -ACCCGCTGAATTTAAGCATATTAG-30 );
mitochondrial Cytochrome Oxidase 1 (CO1), using C1-J-1718SPIDERA and C1-N-2776SPIDER (Vink et al., 2005); nuclear Elongation
Factor 1-alpha (EF1a), using OP2BSAB and OPRC4 (Hedin et al.,
2010). PCR protocols followed Hedin et al. (2010) for EF1a, Thomas
and Hedin (2008) for CO1, and Hedin and Thomas (2010) for 28S,
the latter using an annealing temperature of 56 °C. PCR products
were purified on Millipore plates and amplicons were directly sequenced at Macrogen USA. Sequencher V4.5 was used to assemble
and edit sequence contigs, and all ambiguous sites were scored as
heterozygous using standard ambiguity codes.
CO1 sequences were aligned manually in MEGA 4.0 (Tamura
et al., 2007) using amino acid translation, while 28S and EF1a sequences were aligned with MAFFT (vers. 6; http://mafft.cbrc.jp/
alignment/software/). The G-INS-i strategy was used for the protein coding EF1a; to incorporate 28S structural information we
used Q-INS-i as recommended by Katoh and Toh (2008). To account for alignment uncertainty, 28S and EF1a alignments were
further trimmed using Gblocks (Castresana, 2000), applying a ‘‘less
stringent’’ criterion with standard settings. Models of DNA sequence evolution were evaluated using jModelTest 0.1.1 (Posada,
2008) under three substitution schemes (JC, HKY, GTR) on a fixed
BIONJ tree, allowing for unequal base frequencies and among-site
rate variation. Final model selection was based on the Akaike
Voucher no.
Country: locality, GPS
Ceratolasma tricantha Goodnight & Goodnight, 1942
Hesperonemastoma kepharti (Crosby and Bishop, 1924)
Ischyropsalis carli Lessert, 1905
Ischyropsalis hellwigii hellwigii (Panzer 1794)
Ischyropsalis kollari C.L. Koch, 1839
Ischyropsalis pyrenaea pyrenaea Simon, 1872
Sabacon akiyoshiensis Suzuki, 1963
Sabacon akiyoshiensis Suzuki, 1963
Sabacon akiyoshiensis Suzuki, 1963
Sabacon astoriensis Shear, 1975
Sabacon astoriensis Shear, 1975
Sabacon astoriensis Shear, 1975
Sabacon astoriensis Shear, 1975
Sabacon briggsi Shear, 1975
Sabacon briggsi Shear, 1975
Sabacon cavicolens (Packard, 1884)
Sabacon cavicolens (Packard, 1884)
Sabacon cavicolens (Packard, 1884)
Sabacon cavicolens (Packard, 1884)
Sabacon cavicolens (Packard, 1884)
Sabacon cavicolens (Packard, 1884)
Sabacon cavicolens (Packard, 1884)
Sabacon cavicolens (Packard, 1884)
Sabacon cavicolens (Packard, 1884)
Sabacon cavicolens (Packard, 1884)
Sabacon cavicolens (Packard, 1884)
Sabacon cavicolens (Packard, 1884)
Sabacon cavicolens (Packard, 1884)
Sabacon cavicolens (Packard, 1884)
Sabacon cavicolens (Packard, 1884)
Sabacon cavicolens (Packard, 1884)
Sabacon cavicolens (Packard, 1884)
Sabacon cavicolens (Packard, 1884)
Sabacon imamurai Suzuki, 1964
Sabacon jiriensis Martens, 1972
Sabacon mitchelli Crosby & Bishop, 1924
Sabacon mitchelli Crosby & Bishop, 1924
Sabacon occidentalis (Banks, 1894)
Sabacon occidentalis (Banks, 1894)
Sabacon occidentalis (Banks, 1894)
Sabacon occidentalis (Banks, 1894)
Sabacon occidentalis (Banks, 1894)
Sabacon occidentalis (Banks, 1894)
Sabacon occidentalis (Banks, 1894)
Sabacon occidentalis (Banks, 1894)
Sabacon paradoxus Simon, 1879
Sabacon pasonianus Luque, 1991
Sabacon sheari Cokendolpher, 1984
Sabacon simoni Dresco, 1952
Sabacon simoni Dresco, 1952
Sabacon simoni Dresco, 1952
Sabacon siskiyou Shear, 1975
Sabacon siskiyou Shear, 1975
OP989
OP812
AXLS145
AXLS240
AXLS296
AXLS620
OP2884
OP2886
OP2888
OP1718
OP1914
OP2027
OP2752
OP2441
OP2759
OP647
OP657
OP691
OP721
OP794
OP1283
OP1290
OP1315
OP1321
OP1326
OP1329
OP1348
OP1359
OP1497
OP2575
OP2579
OP2581
OP2584
OP1895
CJM6679
OP1819
OP1820
OP2589
OP2591
OP2695
OP2699
OP2700
OP2701
OP2702
OP2843
AXLS706
AXLS1009
OP2706
AXLS524
OP2521
OP2522
OP1034
OP229
USA: OR, Curry Co., E. Gold Beach, 42.5465, 124.1273
USA: GA, Dade Co., vic. Sitton’s Cave, 34.8597, 85.4847
Switzerland: Kanton Oberwalden, Lungern, Wengen Cave, 46.83, 8.24
Slovenia: Olševa-Mountains, Sadni Travniki, 46.45708, 14.68605
Austria: Steiermark, Koralpe, Wolfsberg, Großer Speikkogel, 46.79874, 14.95092
France: Midi-Pyrénées, Dép. Ariège, Grotte de l’Estelas, 46.91413, 7.55971
Japan: Shikoku, Ehime Pref., Kumakôgen-cho, Mt. Ohkawa-mine, 33.66, 132.90
Japan: Shikoku, Ehime Pref., Kumakôgen-cho, Mt. Ohkawa-mine, 33.66, 132.90
Japan: Shikoku, Ehime Pref., Kumakôgen-cho, Mt. Ohkawa-mine, 33.66, 132.90
USA: ID, Shoshone Co., Hobo Cedar Grove Bot. Area, 47.086, 116.1128
USA: WA, Cowlitz Co., Rose Valley Rd, 46.1676, 122.6374
USA: WA, Pacific Co., East Side of US 101, 46.3756, 123.9496
USA: ID, Benewah Co., Mannering Creek, 47.035, 116.6729
USA: CA, Marin Co., Tamalpais SP, 37.9043, 122.6038
USA: CA, Santa Cruz Co., Big Basin Redwoods SP, 37.174, 122.226
USA: TN, Franklin Co., Roarks Cove Rd, 35.13, 85.55
USA: VA: Lee Co., Cave Spring Rec. Area, 36.48, 82.55
USA: NC, Haywood Co., Hwy 209, below Hebo Mtn., 35.41, 82.54
USA: TN, Anderson Co., vic Norris Dam Cave, 36.2226, 84.0937
USA: NH, Pisgah SP, 42.868, 72.448
USA: NC, Macon Co., Bullpen Bridge, 35.0160, 83.1266
USA: NC, Transylvania Co., Looking Glass Creek, 35.29784, 82.7676
USA: GA, Towns Co., Brasstown Bald, 34.8600, 83.8020
USA: GA, Murray Co., Mulberry Gap Rd., 34.8012, 84.6162
USA: WV, Mercer Co., Camp Creek, 37.50, 81.13
USA: GA, Union Co., Lake Winfield, 34.7438, 83.9744
USA: TN, Graham Co., Joyce Kilmer, 35.3585, 83.9291
USA: NC, Cherokee Co., Dinkin Cove Rd., 35.1809, 83.9988
USA: VA, Smythe Co., Hurricane CG, 36.7240, 81.4904
USA: IL, Johnson Co., Fern Clyffe SP, 37.5385, 88.9801
USA: AR, Newton Co., Buffalo River, 36.0644, 93.1648
USA: AR, Newton Co., Cave Mtn Cave, 35.9569, 93.4069
USA: AR, Washington Co., Devil’s Den, 35.7802, 94.2461
Japan: Honshu, Tottori Pref., Kokufu-cho, Amedaki Falls, 35.5, 134.3
Nepal: Dolakha Distr., Hanumante Danda, 27.7, 86.2
USA: NC, Mitchell Co., Roan Mtn, 36.05, 82.05
USA: NC, Mitchell Co., Roan Mtn, 36.05, 82.05
USA: WA, Wahkiakum Co., Rock Creek, 46.2779, 123.2758
USA: WA, Cowitz Co., Germany Creek, 46.2544, 123.1341
USA: WA, Lewis Co, Kjesbu Rd, 46.5625, 122.4832
USA: WA, Clallam Co, Umbrella Creek, 48.135, 124.6234
USA: WA, Clallam Co, Hoko Falls, 48.194, 124.4499
USA: WA, Clallam Co, Hoko Falls, 48.194, 124.4499
USA: WA, Clallam Co, Trib Sitkum River, 47.9571, 124.2215
Canada: British Columbia, Haida Gwaii, Tow Hill, 54.08, 131.8
France: Languedoc-Roussillon, Dép. Hérault, Olargues, 43.5260, 2.9204
Spain: Cantabria Prov., Miera, Las Regadas Cave, 43.2712, 3.72357
USA: ID, Idaho Co, Burgdorf Rd 13.2 mi, 45.3831, 115.9995
Italy: Prov. Parma, Passo di Montevacà, 44.53546, 9.60968
France: Italy, Prov. Cuneo, Monesi di Triora, 44.07381, 7.75028
France: Italy, Prov. Cuneo, Monesi di Triora, 44.07381, 7.75028
USA: CA, Siskiyou Co., WNW McCloud, 41.2736, 122.1750
USA: OR, Polk Co., N Pedee, 44.77, 123.40
GenBank no.
28S rRNA
CO1
EF1a
JX573543
JX573548
JX573544
JX573545
JX573546
JX573547
JX573593
JX573594
JX573595
JX573549
JX573550
JX573551
JX573552
JX573554
–
JX573555
–
JX573556
JX573557
JX573558
JX573559
JX573560
JX573561
JX573562
JX573563
JX573564
JX573565
JX573566
JX573567
JX573568
JX573569
JX573570
JX573571
JX573572
JX573596
JX573573
JX573574
JX573575
JX573577
JX573578
JX573579
JX573580
JX573581
JX573582
JX573583
JX573584
JX573576
JX573553
JX573585
JX573586
JX573587
JX573588
JX573589
–
JX573642
JX573638
JX573639
–
JX573641
–
–
–
JX573643
JX573644
JX573645
JX573646
JX573647
JX573648
JX573649
GQ870659a
GQ870661a
GQ870658a
JX573650
GQ870660a
GQ870662a
JX573651
JX573652
JX573653
JX573654
JX573655
JX573656
JX573657
JX573658
JX573659
JX573660
JX573661
JX573662
–
JX573663
JX573664
JX573665
JX573667
JX573668
JX573669
JX573670
JX573671
JX573672
JX573673
JX573674
JX573666
–
–
JX573675
GQ870657
JX573676
JX573677
JX573601
JX573606
JX573602
JX573603
JX573604
JX573605
–
–
–
–
–
JX573607
–
JX573608
JX573609
JX573610
GQ872172a
GQ872174a
GQ872171a
JX573612
GQ872173a
GQ872175a
JX573613
JX573614
JX573615
JX573616
JX573617
JX573618
JX573619
JX573620
JX573621
JX573622
JX573623
JX573624
–
JX573625
JX573626
JX573627
–
–
–
–
–
JX573629
–
JX573630
JX573628
–
JX573631
JX573632
GQ872170
JX573633
JX573634
A.L. Schönhofer et al. / Molecular Phylogenetics and Evolution 66 (2013) 303–315
Species/taxon name
306
Table 1
List of species/taxa with voucher information, locality data, GenBank accession for sequences of species used in this study. Voucher number abbreviations are: OP (Opiliones Collection Marshal Hedin, San Diego State University), AXLS
(Collection Axel L. Schönhofer, University of Mainz, Germany), CJM (Collection Jochen Martens, University of Mainz, Germany).
A.L. Schönhofer et al. / Molecular Phylogenetics and Evolution 66 (2013) 303–315
France: Dép. Gard, Parc National des Cévennes, 44.08, 3.59
USA: CA, Mendocino Co., Hwy 128, 38.91956, 123.29644
Nepal: Manaslu Himal, upper Myagdi Khola Vall., 28.58, 84.66
China: Sichuan, Paoma, 29.9262, 101.9795
China: Gansu, Lianhua Shan, Shahetan, 33.43, 106.99
China: Sichuan, Muge Cuo, 30.1462, 101.8605
France: Midi-Pyrénées, Dép. Hautes-Pyrénées, Vallée de l’Ouzom, 43.02708, 0.25581
USA: CA, Shasta Co., Subway Cave, 40.41, 121.25
GQ912770b
JX573590
JX573597
JX573598
JX573599
JX573600
JX573591
JX573592
GQ912866b
JX573678
–
–
–
–
JX573679
JX573680
–
JX573635
–
–
–
–
JX573636
JX573637
Information Criterion (AIC) and individual models were applied to
respective partitions in all downstream analyses.
Bayesian inference using MrBayes v3.1.2 (Huelsenbeck and
Ronquist, 2001; Ronquist and Huelsenbeck, 2003) was applied to
single gene as well as concatenated datasets. Data were partitioned
by gene, and for CO1 and EF1a, by codon position. Bayesian analyses were run for 5,000,000 generations, where in all cases the standard deviation of split frequencies had dropped below 0.01
(Ronquist et al., 2005). Analyses were repeated to further check
for convergence. The first 40% of trees were discarded as burn-in,
with remaining trees used to reconstruct a 50% majority rule consensus tree. Split frequencies were interpreted as posterior probabilities (pp) of clades.
We also investigated phylogenetic relationships for a larger set
of Dyspnoi taxa to evaluate the relative phylogenetic position and
distinctiveness of Sabacon within the suborder. This analysis was
based on 28S rRNA, the only marker for which a comprehensive
sample of Dyspnoi taxa was available. These data were generated,
aligned and analyzed using the above-described methods. 28S sequences not generated for this study were downloaded from GenBank, based on studies of Giribet et al. (2010) and Schönhofer and
Martens (2010; see Fig. 2).
2.3. Divergence time estimation using BEAST
a
b
Hedin et al. (2010).
Giribet et al. (2010).
MCZDNA100711
OP790
CJM6608
CJM6697
CJM6710
CJM6695
AXLS663
OP027
Sabacon sp.
Sabacon sp.
Sabacon sp.
Sabacon sp.
Sabacon sp.
Sabacon sp.
Sabacon viscayanus ramblayanus Martens, 1983
Taracus sp.
307
We estimated taxon divergence times from the concatenated
dataset using BEAST 1.6.1 (Drummond and Rambaut, 2007). Because initial analyses did not reach convergence, we conducted
analyses on a reduced dataset of 37 specimens, excluding recently
diverged lineages within the S. occidentalis and S. cavicolens clades.
The Bayesian analysis was rerun and the reduced dataset topology
compared to the full dataset topology to check for congruence. Codon partitioning was applied to CO1 (1, 2, 3) and EF1a ((1 + 2), 3) to
accommodate heterogeneous rates of evolution (Brandley et al.,
2011). Models for CO1 and EF1a partitions were set to the second
best fitting HKY+I+G model; although other priors were high, effective sample size (ESS) values for rare substitutions within partitions for a GTR model remained low and analyses did not
converge. Strongly supported nodes (pp = 0.95–1.00) from the full
dataset were constrained to be monophyletic in the reduced dataset, to avoid entrapment in local optima (following Smith, 2009).
No treemodel operators were removed, allowing BEAST to re-estimate the remaining topology. Priors were set to gamma if required
to be modified. These settings were applied to all subsequent analyses. Analyses were run until ESS values exceeded 200 even for
poorly sampled priors, which was after 100,000,000 generations.
Analyses were replicated twice and checked for convergence using
Tracer 1.5.
We checked for rate heterogeneity in all datasets using the random local clock model implemented in BEAST. This clock model
estimates rate changes for each branch individually and identifies
lineage specific rate changes (Drummond and Suchard, 2010). An
uncalibrated log-normal analysis estimated different clock models
for the three gene partitions, and the standard deviation of the partition frequency (stdev) was investigated for deviation from clocklike behavior.
For final analyses in BEAST a Death-Birth model of speciation
and a relaxed log-normal clock were applied. Clock rates where unlinked and estimated for individual partitions. We simultaneously
applied both rate and fossil calibrations for a final BEAST reconstruction. For the clocklike CO1 (see Section 3.3.) we specified a
clock rate based on the standard arthropod clock from Brower
(1994; 2.3% per Ma) and a newly reported arthropod clock rate
(Papadopoulou et al., 2010; 2.69%/Ma). To accommodate differences of these rates a uniform prior was chosen for CO1. Meanrate
to include the range of 0.0115–0.0145/Ma. The fossil taxon Sabacon
308
A.L. Schönhofer et al. / Molecular Phylogenetics and Evolution 66 (2013) 303–315
claviger from European amber was used as a calibration point for
the minimum age of the cavicolens-Europe clade (see Section 3.2.).
This taxon dates to the Eocene (55–34 Ma; Dunlop, 2006) for
which a log-normal distributed prior was set with a mean of
2.53, a Stdev of 0.9, and an offset of 32 so the 2.5% percentile would
be 34 Ma and the median 44.18 Ma, constraining the minimum age
of the node but allowing it to be significantly older.
2.4. Ancestral range reconstructions using S-DIVA and Lagrange
Ancestral distributions were reconstructed using S-DIVA (Yu
et al., 2010), which evaluates the most parsimonious reconstructions of dispersal and vicariance events and integrates uncertainty
over a set of analyzed trees. Here we used 5000 post burn-in trees
resulting from Bayesian analysis of the full concatenated dataset.
Default settings were modified for ‘‘Allowed Reconstruction’’
(checked) and Hold set to 32.767 (to include all alternative reconstructions per node, see Ronquist (1997), DIVA 0.1.1 manual). We
specified five distributional areas for terminal taxa, including A:
continental Asia (including Nepal and China only), B: western
North America, C: eastern North America, E: Europe, and J: Japan.
We conducted analyses with maximum alternative scenarios at
each node set to 5, 4 and 3 to examine the effect of constraints
on reconstructed ancestral distributions (Bendiksby et al., 2010;
Kodandaramaiah, 2010).
Likelihood reconstruction of ancestral areas using Lagrange
(C++ version, Bendiksby et al., 2010; http://code.google.com/p/lagrange/downloads/detail?name=lagrange_cpp_0.1BETA2_%20for_MAC10.6.tar.gz&can=2&q) was performed on the dated Bayesian
consensus phylogeny from BEAST. Utilizing phylogenetic and temporal information, Lagrange reconstructs the biogeographic history
of a set of discrete areas based on the distribution of extant taxa.
Accounting for branch length information, Lagrange integrates
the probability of extinction and dispersal as stochastic events calculated over time. Nodes that showed ambiguous area reconstruction results in the Bayesian consensus tree were further
investigated by reconstructing over all 5000 retained BEAST trees,
following Smith (2009) and Bendiksby et al. (2010). Lagrange output was manipulated using TextWrangler for input into ‘‘R’’ 2.13.1
and Excel 2007, to calculate probabilities of all ancestral area scenarios at each node, and to segregate the five best scoring scenarios
by node age. Further confidence in reconstructions was attained by
including the percentage of individual ancestral area reconstructions per node (Smith, 2009). Lagrange C++ computation omits scenarios scoring less than 5% probability in individual
reconstructions, thereby ‘‘weeding out’’ weakly supported scenarios. No constraints on the maximal number of areas, or on the
directionality and timing of dispersal, were used in our analyses.
3. Results
3.1. Dyspnoi 28S phylogeny
Tree topologies based on 28S MAFFT versus MAFFT + Gblocks
alignments were similar (results not shown), so we used the
MAFFT-only alignment. The 28S phylogeny including multiple
Dyspnoi and Eupnoi outgroup taxa (1268 aligned base pairs for
82 terminals) strongly supports the superfamilies Ischyropsalidoidea and Troguloidea (1.00 pp; Fig. 2). The recently emended Sabaconidae (Giribet et al., 2010), including Hesperonemastoma,
Sabacon and Taracus, receives low support (0.78 pp), but Hesperonemastoma and Taracus are strongly supported as sister taxa (1.00
pp). Sabacon is strongly supported as monophyletic (1.00 pp;
Fig. 2) and shows high internal 28S sequence divergence, with separation into four well supported primary clades (Fig. 2). Based on
28S phylogram branch lengths, sequence divergence within Sabacon exceeds that of any other Dyspnoi genus included in the analysis, and is indeed greater than that of most Dyspnoi families
(Fig. 2).
3.2. Phylogenetic relationships within Sabacon
Fifty-four Sabacon specimens plus 7 outgroup taxa specimens
were used to generate sequence data for 28S, CO1 and EF1a (59,
50, and 42 sequences, respectively). Only 28S could be amplified
for some of the older samples and several EF1a PCR products
yielded no results in direct sequencing, possibly due to the close
proximity of the forward primer to additional introns reported
for Sabacon (Hedin et al., 2010). The final alignments of CO1
(968 bp) and EF1a (681 bp) were unambiguous. We excluded the
difficult-to-align (and sometimes partial) EF1a introns, as suggested by Gblocks. 28S was difficult to align and we compared tree
topologies from the MAFFT and MAFFT + Gblocks alignments. Results based on these alternative alignments were very similar, differing only for nodes that showed low support in analyses derived
from both alignments. An exception involved S. occidentalis sequence OP2591, which changed position within the occidentalisclade with high support from alternative alignments. The placement of this particular sequence does not greatly impact the biogeographic analyses presented below. We used the MAFFT-only
alignment (1236 aligned bp) in subsequent concatenated analyses.
For individual gene trees see Suppl. Figs. 1–3.
In agreement with results from the larger 28S dataset (Fig. 2),
analysis of the concatenated alignment (2885 total aligned bp)
recovers Sabacon as monophyletic with strong support (1.00 pp,
Fig. 3). Within Sabacon, four primary clades emerged that we define here for further discussion (Figs. 3–5): (1) western North
American occidentalis-clade, basal placement in Sabacon, unites S.
briggsi (1a), S. occidentalis and S. siskiyou (1b); (2) mitchelli-Japanclade, unites the eastern North American S. mitchelli and the Japanese S. akiyoshiensis and S. imamurai; (3) astoriensis-Asia-clade,
unites the western North American S. astoriensis and S. sheari
(3a) with S. jiriensis and Sabacon sp. from Nepal and sampled Chinese Sabacon (3b) – this clade is subdivided according to geographic origin; (4) cavicolens-Europe-clade, unites the eastern
North American S. cavicolens (4a) and sampled European Sabacon
species. These four clades are strongly supported in both individual
gene and concatenated analysis (Fig. 2). A conspicuous implication
of this phylogenetic structuring is that three of four regional faunas
(wNA, eNA, east Asia) are non-monophyletic, and that three clades
include intercontinental disjuncts.
3.3. Divergence time estimation using BEAST
Rate heterogeneity as estimated by the random local clock
model in BEAST (Drummond and Suchard, 2010) was generally
low throughout Sabacon and Sabaconidae, although we found significantly increased rate change in the astoriensis-Asia-clade with
additional increase within S. astoriensis. This result holds true for
the Dyspnoi, with minor rate changes between major groups (only
decreasing in Dicranolasmatidae) but marked changes in individual lower level taxa (Anelasmocephalus, Mitostoma). The uncalibrated log-normal analysis rejected clocklike evolution for EF1a
(stdev: 2.1) and 28S (2.2), but not for CO1 (stdev: 0.3), justifying final BEAST settings as previously specified.
Unconstrained BEAST trees differed noticeably from MrBayes
trees at deeper nodes, recovering a basal polytomy including Sabacon and outgroups, and implied different interrelationships of
the four major Sabacon clades. As these alternative BEAST reconstructions showed low support, and BEAST is known to converge
on local optima, we constrained basal nodes to force BEAST to
A.L. Schönhofer et al. / Molecular Phylogenetics and Evolution 66 (2013) 303–315
309
Fig. 2. Results of 28S Bayesian phylogenetic analysis of Dyspnoi including Eupnoi outgroups. GenBank accession numbers are given for taxa not listed in Table 1.
converge to the same topology as MrBayes, as suggested by Smith
(2009). This resulted in faster convergence and higher ESS values
for all parameters and allowed setting of more complex substitution models as estimated by jModelTest. The resulting BEAST chronogram is shown in Fig. 4 (see also Suppl. Fig. 4).
According to our reconstruction Sabacon originated about
85 Ma ago (53–132 highest posterior density (HPD)), with subsequent diversification into four primary clades occurring within a
relative short time frame of 20 Ma (Fig. 4). The estimated ages
of the most recent common ancestor (MRCA) of the primary clades
falls within a time window of less than 10 Ma (mean times of 30–
40 Ma), near the Eocene–Oligocene boundary. Intraspecific divergences within nominal taxa are estimated to be relatively old, often
predating the Pliocene. More recent divergences, partly dating to
the Pleistocene, were detected within the more densely sampled
Sabacon cavicolens (2.1–3.2 Ma) and S. occidentalis clades (1.5 Ma).
3.4. Ancestral area reconstruction using S-DIVA and Lagrange
S-DIVA analyses were conducted on a set of post-burnin trees
resulting from Bayesian analysis of the full concatenated dataset.
S-DIVA results vary depending upon whether the number of maximal areas per ancestral node was constrained to 3 or 4, or left
unconstrained at 5. Nodes for which S-DIVA calculated several
equally likely scenarios if left unconstrained often tended to favor
a single area scenario in more constrained analyses (Fig. 5). Also,
when optimal results could have yielded the same area scenario
(e.g., a scenario including three areas), the reconstructions did
not agree. Finally, nodes with alternative scenarios showed a higher number of possibilities that also differed considerably from the
Lagrange C++ results.
Lagrange C++ analyses were conducted on a set of trees resulting
from BEAST analysis of the reduced concatenated dataset. Fig. 4 pie
charts show the total proportion of each scenario at nodes with possible alternatives, whereas histograms show frequencies of the five
best scoring scenarios distributed over the 95% higher posterior density of node age reconstructions (see also Suppl. Fig. 4). Pie charts
thereby depict the total likelihood of each scenario at the nodes,
while the likelihood of timing of scenarios is shown by histograms.
Large differences in the timing of alternative scenarios were generally
not detectable in our dataset. While the two expanded histograms for
the astoriensis-Asia and mitchelli-Japan clades, and for the most recent common ancestor of Sabacon show a shift of about 10–15 Ma
at the most likely time of the individual area reconstruction, these
are small differences in comparison to the overall 95% higher posterior densities and the total histogram coverage (Suppl. Fig. 4).
310
A.L. Schönhofer et al. / Molecular Phylogenetics and Evolution 66 (2013) 303–315
Fig. 3. Bayesian phylogram of Sabacon and Ischyropsalidoid outgroups based on concatenated data. Members of primary Sabacon clades (1–4) discussed in the text are
indicated by different colors. Posterior probabilities given at each node with values in parentheses indicating support from single gene reconstructions of 28S, CO1 and EF1a
(see Suppl. Figs. 1–3) respectively for primary Sabacon clades.
Lagrange (Fig. 4) and S-DIVA (Fig. 5) reconstructions support
intercontinental dispersal in two of three intercontinental disjunct
clades. Within the cavicolens-Europe-clade, S-DIVA unambiguously
supports Europe as the center of origin. While Lagrange also supports a European center of origin with high likelihood, this reconstruction does not exclude a shared ancestral area of Europe and
eastern North America (i.e., ambiguous dispersal directionality).
The biogeographic connection between Europe and eastern North
America supports the assignment of the fossil Sabacon claviger to
a lineage of European (+eastern North American) origin, as emphasized here. Results for the mitchelli-Japan-clade differ between analyses. Lagrange shows higher support for an ancestral area
combining Japan with eastern North America, while S-DIVA favors
a Japanese center of origin. Neither reconstruction method provides
unambiguous results for dispersal directionality in the astoriensisAsia-clade. At deeper nodes, both methods weakly support western
North America as the ancestral area for Sabaconidae (Suppl. Fig. 4).
4. Discussion
4.1. Sabacon systematics
Sabacon is the nominative genus of the originally monotypic
family Sabaconidae (Dresco, 1970). Subsequent evaluations of the
A.L. Schönhofer et al. / Molecular Phylogenetics and Evolution 66 (2013) 303–315
311
Fig. 4. Consensus chronogram of Bayesian BEAST and Lagrange likelihood analysis based on 5000 trees. Scale at bottom shows time in millions of years. Letters at nodes, pie
charts and specimen identifier correspond to area-legend top left. Indicators (1–4) at nodes indicate clades discussed in the text. Numbers at nodes report mean node ages,
node bars the respective 95% posterior density based on the consensus tree. When more than one ancestral area was likely for the consensus tree percentage distributions of
scenarios were re-estimated over all BEAST trees and segregated by time shown in pie charts using Lagrange C++ and histogram bars at nodes. Dotted line, scaled on the left,
indicates change of average temperature over time according to Zachos et al. (2001). Outgroup branches (dashed) have been shortened and posterior density and histograms
omitted to focus on Tertiary events. For a fully expanded analysis see Suppl. Fig. 4.
family first rejected its status (Gruber, 1978; Martens, 1972; Shear,
1975) but then accepted an altered family diagnosis (Martens,
1976, 1980, 1983; Shear, 1986). Martens et al., (1980), Martens
(1983) based evidence on the peculiar and derived genital morphology, unique within the Dyspnoi. Later, Shear (1986) downgraded genital morphology as genus specific for Sabacon and
included Taracus in the Sabaconidae on the basis of a cladistic analysis. The latest emendation was by Giribet et al. (2010), adding
Hesperonemastoma to Sabaconidae based on molecular phylogenetic affinities to Taracus. Our reassessment of Sabaconidae based
upon a more comprehensive taxon sample finds low support for
the relationship of Sabacon with Taracus plus Hesperonemastoma
(Fig. 2; comparable to Giribet et al., 2010).
Compared to 28S rRNA divergence of other families within the
Dyspnoi (Fig. 2), Sabacon emerges as an isolated and highly divergent clade, and perhaps should be re-classified into multiple genera. While we do not attempt to taxonomically subdivide
Sabacon in this paper, we provide evidence for deep divergence
encouraging such endeavors. Support for a taxonomic split also
stems from morphology, where comparisons of Sabacon molecular
lineages reveal clade-specific differences in male genital morphology. This situation is similar to that seen within the Dyspnoi family
312
A.L. Schönhofer et al. / Molecular Phylogenetics and Evolution 66 (2013) 303–315
Fig. 5. S-DIVA reconstructions. Histograms on the left show differences of analyses constraint for 3, 4, and 5 areas per node. Colored circles and letters at nodes show identical
results for all three constraints. Indicators (1–4) at nodes indicate clades discussed in the text.
A.L. Schönhofer et al. / Molecular Phylogenetics and Evolution 66 (2013) 303–315
Nemastomatidae, members of which exhibit relative stasis of body
form but display greater variation in male genital and/or secondary
sexual characteristics (Schönhofer and Martens, 2012).
Sabacon is separated into four well-defined molecular clades
(Figs. 3–5); here we discuss morphological support for these
clades.
(1) Occidentalis-clade: Sabacon briggsi is sister to a group comprising S. occidentalis and S. siskiyou. The latter are very similar but unique within Sabacon in exhibiting inflated penis
glandes with a firmly attached, three-dimensional and partly
divided stylus (Shear, 1975). From the single male S. briggsi
available for our analysis we can confirm stylus modifications, including a regular row of backwards pointing teeth.
All three species share large conical cheliceral apophyses
very similar in shape and size. Within the occidentalis and
siskiyou group OP2591 is genetically unique and needs to
be further investigated. Specimens collected from this location are smaller than other members of this clade, but only
females are available. We expect at least stylus-specific
differences for males from this potentially undescribed
species.
(2) Mitchelli-Japan-clade: The eastern North American S. mitchelli is phylogenetically related to Japanese taxa, a result suggested by Cokendolpher (1984) based on comparative
morphology. All three species in the recovered mitchelliJapan-clade lack male cheliceral glands, but differ in genital
morphology. The relatively simple penis of S. mitchelli
(Shear, 1975) differs from the three-dimensional glandes
penis with distinct spination, found in the two Japanese species (Suzuki, 1974). For the nine described Japanese Sabacon
species, Suzuki (1974) recognized the following three species groups based on external morphology (penis, ovipositor,
armaments of male chelicera and palpal patella, scutum, and
body size): pygmaeus-group, dentipalpis-group, and akiyoshiensis-group. It is interesting that both S. akiyoshiensis of the
akiyoshiensis-group, which is the smallest Japanese species
(ca. 1.5 mm in male body length) and S. imamurai of the
dentipalpis-group, which is the largest Japanese species (ca.
4 mm), are placed in the same clade despite their large difference in external morphology and body size. The phylogenetic position of the pygmaeus-group, not sampled in this
study, remains to be determined. Also, there are Sabacon
species from the Chinese Sichuan that appear morphologically-allied to Japanese taxa (Martens, 1972; Tsurusaki and
Song, 1993a, J.M. unpublished results); the phylogenetic
position of these taxa remains to be determined.
(3) Astoriensis-Asia-clade: This group is subdivided according to
geographic origin. The western North American clade
includes S. sheari and S. astoriensis. Cokendolpher (1984)
considered S. sheari to be closely related to S. mitchelli based
upon external morphology, but this relationship is not supported by our data. There is considerable sequence divergence between the Oregon/Washington versus Idaho
populations of S. astoriensis. Investigation of specimens from
these disjunct areas reveals differences in male chelicerae
and penis morphology (pers. obs.), perhaps consistent with
species-level differentiation. There remains uncertainty in
the correct phylogenetic placement of taxa in this clade. Different analyses swapped the position of S. sheari (Figs. 2 and
3), and basal relationships in the astoriensis-sheari-clade of
the full dataset (Fig. 3) are partly unresolved. We also found
considerable rate change within this clade suggesting alignment or systematic errors, which might be overcome by
incorporating additional Asian taxa. Our current sparse sampling prevents strong conclusions regarding the phyloge-
313
netic structuring of mainland Asian taxa. According to our
results and morphological evidence we argue that most of
the Chinese and Nepalese divergence is relatively recent,
but we acknowledge that this fauna is morphologically disparate, and that our current sample does not capture most
of this morphological divergence.
(4) Cavicolens-Europe-clade: Based on comparative morphology,
Shear (1975) and Martens (1983) suggested a close relationship between S. cavicolens and European Sabacon, corroborated in this study. Also, speciation in this clade is clearly
older than the Pleistocene, consistent with the hypothesis
of Martens (1983). Basal nodes within the group receive only
low support, placing S. simoni closest to S. cavicolens.
Strongly supported European relationships include S. paradoxus with S. viscayanus (as expected by Martens, 1983),
and S. pasonianus with the Sabacon juvenile from the Cevennes in southern France (MCZDNA100711). These latter
taxa are not in geographic proximity and are genetically
divergent. Sabacon cavicolens from eastern North America
probably comprises several cryptic species (estimated divergence mean between 2 and 12 Ma). Shear (1975) mentioned
morphological divergence within S. cavicolens, a matter that
requires further investigation.
4.2. Historical biogeography
While the northern latitudes climate of the Eocene was more
tropical, a sharp decrease in average global mean temperatures
at the Eocene–Oligocene boundary (33 Ma; Zachos et al., 2001)
marks the almost simultaneous intercontinental biogeographic exchange observed in all Sabacon clades (Fig. 4). Most extant Sabacon
taxa show preferences to permanently cool and moist habitats,
even to include subtropical cloud forest taxa occurring in Nepal
(some included in our sample; Martens, 1972, 1983). We hypothesize that the onset of global cooling enabled cryophilic Sabacon
to disperse into more suitable habitats, and to diversify into the
many lineages we see today, a causal connection also found for
cold adapted snakes (Lynch, 2009) and salamanders (Vieites
et al., 2007; Zhang et al., 2008). Biotic conditions during connection
of the trans-Atlantic de Geer land bridge provide an illustrative
example. At the Eocene–Oligocene boundary the climate is reconstructed as becoming more seasonal with cooler winters but moderate annual temperatures (Eldrett et al., 2009; Ivany et al., 2000).
In Greenland, as part of the de Geer Bridge, these conditions included high annual precipitation promoting a humid coniferous
forest (Eldrett et al., 2009). Modern humid coniferous forests house
rich Sabacon faunas, as for example, in western North America.
Fennoscandian Sabacon fossils from Baltic amber (Dunlop, 2006),
also from the de Geer Bridge, most likely originated from Sciadopityaceae (Wolfe et al., 2009), which together with other dominant
conifers, went almost globally extinct in the late Oligocene. This
climatic and fossil evidence suggests high latitudinal dispersal conditions likely suitable for Sabacon. We hypothesize that global
cooling created large and maybe continuous suitable habitats, enabling a general expansion of cryophilic Sabacon and connecting
habitat patches from the warmer Eocene.
Phylogenetic relationships within the cavicolens-Europe-clade
clearly favor Europe as the center of origin for this clade. A possible
scenario is that the complex and repeated fragmentation of Europe
(Smith et al., 1994) caused early and deep divergence of European
Sabacon, while the trans-Atlantic de Geer land bridge facilitated
dispersal to eastern North America. Further speciation within S.
cavicolens is considerably younger and must have happened long
after the land bridge had been severed.
For the mitchelli-Japan-clade, ancestral areas are either Japan or
a combination of Japan and eastern North America; the former fa-
314
A.L. Schönhofer et al. / Molecular Phylogenetics and Evolution 66 (2013) 303–315
vored by S-DIVA and the latter by Lagrange. Major events disrupting a connection between Asia and eastern North America include
the uplift of the Rocky Mountains during the Eocene (45–36 Ma)
followed by a sharp decrease in average temperatures during the
Oligocene (33 Ma) to the present (Zachos et al., 2001). While
decreasing temperatures might have facilitated dispersal of cold
adapted Sabacon, the increasing aridification of mid-continental
North America could have prohibited subsequent faunistic
interchange.
According to the MrBayes reconstruction of the full dataset
(Fig. 3), the astoriensis-Asia-clade comprises two well supported
geographic lineages, with western North American S. astoriensis
and S. sheari as sister to the Asian Sabacon. The separating split,
dating to 32 Ma (17–53 HPD) falls in the time frame for increased
faunistic interchange between western Nearctic and eastern Palearctic (Sanmartín et al., 2001), while no recent Pleistocene exchange is apparent. While our data does not allow an inference
of dispersal directionality in this clade, high morphological diversity of recent Asian Sabacon indicates continental Asia as the potential source area.
Although some biogeographic patterns seen in Sabacon are also
found in other taxa, Sabacon is peculiar in several respects. For
example, one connection for which Donoghue and Smith (2004)
recovered the highest proportion of disjunct distributions, Asia
and Europe, is completely absent from our biogeographic scenarios. Also, there is no apparent phylogenetic affinity between species in eastern versus western North America. Donoghue and
Smith (2004) re-analyzed the animal data of Sanmartín et al.
(2001) and found few animal disjuncts between eastern North
America and Europe (7%), with a low proportion (4%) of these
reconstructed dispersal events occurring from Europe to eastern
North America. The eastern North America to Japan connection
is rare for animal disjuncts and dispersal corridors (Brewer
et al., 2012; Donoghue and Smith, 2004; Sanmartín et al., 2001).
Another peculiarity concerning the mitchelli-Japan-clade is the
estimated divergence time for the geographical groups
(mean = 32 Ma, 16–53 HPD). Xiang et al. (2000) dated splits between eleven species pairs of plants, and found most estimates
in the last 10 Ma (max. 12.6 Ma) in the late Miocene and Pliocene.
These estimates probably correspond to other Opiliones taxa, e.g.,
Crosbycus dasycnemus (Shear, 1986), Acropsopilio boopis (Suzuki,
1976), and two Caddo species (Suzuki, 1976; Shultz and Regier,
2009) with disjunct distributions in Japan (and China for Crosbycus; Tsurusaki and Song, 1993b) and eastern North America.
Divergence in these harvestmen taxa is likely relatively young,
as evidenced by morphological similarity and low genetic divergence (when estimated).
Our molecular clock-derived divergence date estimates differ
considerably from similar molecular clock estimates reported by
Giribet et al. (2010). They estimated the split between S. cavicolens and European Sabacon at 75 Ma, which greatly exceeds
our estimates (mean of 33 Ma, 24–46 HPD). Similarly, the Giribet
et al. (2010) estimate for the MRCA of Ceratolasma and Ischyropsalis (160 Ma) again exceeds our estimate (65 Ma, 31–109
HPD), as does the overall age of the Ischyropsalidoidea (243 versus 137 Ma). These differences may stem from the predominantly
ancient calibration points used by Giribet et al. (2010), which may
not account for fluctuations along the calibrated regression and
influence younger nodes (Conroy and van Tuinen, 2003; Rutschmann et al., 2007). Applying their root age of 243 Ma to our analysis caused considerable increase in overall HPD ranges (results
not shown), and we therefore refrained from using this possible
calibration. Ultimately, incorporation of additional reliable calibration points and more DNA sequence data will be needed to resolve this discrepancy.
To summarize, the onset of diversification within the four major
Sabacon clades, as well as intercontinental exchange in three
clades, is nearly simultaneous and seems connected to global cooling, allowing dispersal across northern land bridges. Our data supports the perspective that the study of arthropods provides more
fine-scaled views on Holarctic biogeography. For example, while
cryophilic Sabacon were able to survive in suitable microhabitats,
the aforementioned coniferous megaflora went mostly extinct
and is no longer available for molecular based reconstructions.
Acknowledgements
We thank Jared Grummer, Dean Leavitt, Casey Richart and Jordan Satler for help with the phylogenetic methods and for general
discussion. Richard Ree and Stephen Smith provided help with the
Lagrange analysis and methods used in Smith (2009). Fred Coyle,
Shahan Derkarabetian, Lucia Labrada, Dean Leavitt, Steven Thomas,
Robin Keith, Wolfgang Schawaller, Joachim Schmidt, Jim Starrett,
Jordan Satler, Adrienne Richart, Dale Richart, Dan Richart, Casey Richart and Peter Scott helped to collect specimens. Research efforts
of M. McCormack were funded by the American Arachnological
Society, and a U.S. Fish and Wildlife Service contract to M.H.
A.L.S. was funded by a postdoc grant of the Deutsche Forschungsgemeinschaft (DFG) to work on systematics and biogeography of
Holarctic Opiliones, and collected Sabacon in the southwestern
Alps with the help of an EDIT/ATBI+M travel Grant (2008). Field research in Asia was funded by Feldbausch Stiftung and Wagner Stiftung, both at Fachbereich Biologie, Mainz University, as annual
grants to of J.M.
Appendix A. Supplementary material
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.ympev.2012.10.001.
References
Bendiksby, M., Schumacher, T., Gussarova, G., Nais, J., Mat-Salleh, K., Sofiyanti, N.,
Madulid, D., Smith, S.A., Barkman, T., 2010. Elucidating the evolutionary history
of the Southeast Asian, holoparasitic, giant-flowered Rafflesiaceae: Pliocene
vicariance, morphological convergence and character displacement. Mol.
Phylogenet. Evol. 57, 620–633.
Boyer, S.L., Giribet, G., 2007. A new model Gondwanan taxon: systematics and
biogeography of the Gondwanan harvestman family Pettalidae (Arachnida,
Opiliones, Cyphophthalmi), with a taxonomic revision of genera from Australia
and New Zealand. Cladistics 23, 337–361.
Boyer, S.L., Clouse, R., Benavides, L.R., Sharma, P., Schwendinger, P.J., Kuranarathna,
I., Giribet, G., 2007. Biogeography of the world: a case study from
cyphophthalmid Opiliones, a globally distributed group of arachnids. J.
Biogeogr. 34, 2070–2085.
Brandley, M.C., Wang, Y., Guo, X., de Oca, A.N., Fería-Ortíz, M., Hikida, T., Ota, H.,
2011. Accommodating heterogeneous rates of evolution in molecular
divergence dating methods: an example using intercontinental dispersal of
Plestiodon (Eumeces) lizards. Syst. Biol. 60, 3–15.
Brewer, M.S., Spruill, C.L., Rao, N.S., Bond, J.E., 2012. Phylogenetics of the millipede
genus Brachycybe Wood, 1864 (Diplopoda: Platydesmida: Andrognathidae):
patterns of deep evolutionary history and recent speciation. Mol. Phyl. Evol. 64,
232–242.
Brower, A.V.Z., 1994. Rapid morphological radiation and convergence among races
of the butterfly Heliconuius erato inferred from patterns of mitochondrial DNA
evolution. Proc. Nat. Acad. Sci. USA 91, 6491–6495.
Burbrink, F.T., Lawson, R., 2007. How and when did the Old World rat snakes cross
into the New World? Mol. Phylogenet. Evol. 43, 173–189.
Castresana, J., 2000. Selection of conserved blocks from multiple alignments for
their use in phylogenetic analysis. Mol. Biol. Evol. 17, 540–552.
Cokendolpher, J.C., 1984. A new species of Sabacon Simon from Oregon (Arachnida:
Opiliones: Sabaconidae). Can. J. Zool. 62, 989–991.
Conroy, C.J., van Tuinen, M., 2003. Extracting time from phylogenies: positive
interplay between fossil and genetic data. J. Mammal. 84, 444–455.
Derkarabetian, S., Steinmann, D.B., Hedin, M., 2010. Repeated and time-correlated
morphological convergence in cave-dwelling harvestmen (Opiliones,
Laniatores) from montane western North America. PLoS ONE 5, 1–13.
Donoghue, M.J., Smith, S.A., 2004. Patterns in the assembly of temperate forests
around the Northern Hemisphere. Philos. Trans. Royal Soc. B 359, 1633–1644.
A.L. Schönhofer et al. / Molecular Phylogenetics and Evolution 66 (2013) 303–315
Dresco, E., 1970. Recherches sur la variabilité et la phylogénie chez les Opiliones du
genre Ischyropsalis C. L. Koch (Fam. Ischyropsalidae), avec la création de la
famille nouvelle des Sabaconidae. Bull. Mus. Natl. Hist. Nat. 2e (41), 1200–1213.
Drummond, A.J., Rambaut, A., 2007. BEAST: Bayesian evolutionary analysis by
sampling trees. BMC Evol. Biol. 7, 214.
Drummond, A.J., Suchard, M.A., 2010. Bayesian random local clocks, or one rate to
rule them all. BMC Biol. 8, 1–12.
Dunlop, J.A., 2006. Baltic amber harvestman types (Arachnida: Opiliones: Eupnoi
and Dyspnoi). Foss. Rec. 9, 167–182.
Eldrett, J.S., Greenwood, D.R., Harding, C.I., Huber, M., 2009. Increased seasonality
through the Eocene to Oligocene transition in northern high latitudes. Nature
459, 969–973.
Enghoff, H., 1995. Historical biogeography of the Holarctic: area relationships,
ancestral areas, and dispersal of nonmarine animals. Cladistics 11, 223–263.
Giribet, G., Shear, W.A., 2010. The genus Siro Latreille, 1796 (Opiliones,
Cyphophthalmi, Sironidae), in North America with a phylogenetic analysis
based on molecular data and the description of four new species. Bull. Mus.
Comparat. Zool. 160, 1–33.
Giribet, G., Vogt, L., Pérez González, A., Sharma, P., Kury, A.B., 2010. A multilocus
approach to harvestman (Arachnida: Opiliones) phylogeny with emphasis on
biogeography and the systematics of Laniatores. Cladistics 26, 408–437.
Giribet, G., Sharma, P., Benavides, L.R., Boyer, S.L., Clouse, R.M., de Bivort, B.L.,
Dimitrov, D., Kawauchi, G.Y., Murienne, J.Y., Schwendinger, P.J., 2012.
Evolutionary and biogeographic history of an ancient and global group of
arachnids (Arachnida, Opiliones, Cyphophthalmi) with a new taxonomic
arrangement. Biol. J. Linn. Soc. 105, 92–130.
Gruber, J., 1978. Redescription of Ceratolasma tricantha Goodnight and Goodnight,
with notes on the family Ischyropsalidae (Opiliones, Palpatores). J. Arachnol. 6,
105–124.
Hedin, M., Thomas, S.M., 2010. Molecular systematics of eastern North American
Phalangodidae (Arachnida: Opiliones: Laniatores), demonstrating convergent
morphological evolution in caves. Mol. Phylogenet. Evol. 54, 107–121.
Hedin, M., Derkarabetian, S., McCormack, M., Richart, C., Shultz, J.W., 2010. The
phylogenetic utility of the nuclear protein-coding gene EF-1a for resolving recent
divergences in Opiliones, emphasizing intron evolution. J. Arachnol. 38, 9–20.
Hedin, M., Tsurusaki, N., Macías-Ordóñez, R., Shultz, J.W., 2012. Molecular systematics
of sclerosomatid harvestmen (Opiliones, Phalangioidea, Sclerosomatidae):
geography is better than taxonomy in predicting phylogeny. Mol. Phylogenet.
Evol. 62 (1), 224–236.
Huelsenbeck, J.P., Ronquist, F.R., 2001. MrBayes: Bayesian inference of phylogenetic
trees. Bioinformatics 17, 754–755.
Ivany, L.C., Patterson, W.P., Kyger, C.L., 2000. Cooler winters as a possible cause of
mass extinctions at the Eocene/Oligocene boundary. Nature 407, 887–890.
Katoh, K., Toh, H., 2008. Improved accuracy of multiple ncRNA alignment by
incorporating structural information into a MAFFT-based framework. BMC
Bioinformatics 9, 1–13.
Kodandaramaiah, U., 2010. Use of dispersal–vicariance analysis in biogeography – a
critique. J. Biogeogr. 37, 3–11.
Lopez, A., Emerit, M., Rambla, M., 1980. Contribution à l’etude de Sabacon
paradoxum Simon 1879 (Opiliones, Palpatores, Ischyropsalidae). Stations
nouvelles, particularités électro-microscopiques du prosoma et de ses
appendices. Comptes Rendus de la Vième Colloque d’Arachnologie
d’Expression Française, 1979, Barcelona, IX, pp. 147–158.
Lynch, V.J., 2009. Live-birth in vipers (Viperidae) is a key innovation and adaptation
to global cooling during the Cenozoic. Evolution 63, 2457–2465.
Mallatt, J., Sullivan, J., 1998. 28S and 18S rDNA sequences support the monophyly of
lampreys and hagfishes. Mol. Biol. Evol. 15, 1706–1718.
Martens, J., 1972. Opiliones aus dem Nepal-Himalaya. I. Das Genus Sabacon Simon
(Arachnida: Ischyropsalididae). Senck. Biol. 53, 307–323.
Martens, J., 1976. Genitalmorphologie, System und Phylogenie der Weberknechte
(Arachnida: Opiliones). Entomol. Germ. 3 (1/2), 51–68.
Martens, J., 1983. Europäische Arten der Gattung Sabacon Simon 1879 (Arachnida:
Opiliones: Sabaconidae). Senck. Biol. 63, 265–296.
Martens, J., 1989. Sibirische Arten der Gattung Sabacon Simon (Arachnida:
Opiliones: Sabaconidae). Senck. Biol. 69, 369–377.
Martens, J., 1980. Versuch eines Phylogenetischen Systems der Opiliones. In:
Gruber, J. (Ed.), 8 Internationaler Arachnologen-Kongress. Verlag H. Egerman,
Wien, pp. 355–360.
McKenna, M.C., 1983. Cenozoic paleogeography of North Atlantic land bridges. In:
Bott, M.H.P., Saxov, S., Talwani, M., Thiede, J. (Eds.), Structure and Development
of the Greenland-Scotland Bridge: New Concepts and Methods. Plenum, New
York, pp. 351–395.
Papadopoulou, A., Anastasiou, I., Vogler, A.P., 2010. Revisiting the insect
mitochondrial molecular clock: the mid-Aegean trench calibration. Mol. Biol.
Evol. 27, 1659–1672.
Posada, D., 2008. JModelTest: phylogenetic model averaging. Mol. Biol. Evol. 25,
1253–1256.
315
Ree, R.H., Smith, S.A., 2008. Maximum likelihood inference of geographic range
evolution by dispersal, local extinction, and cladogenesis. Syst. Biol. 57, 4–14.
Ronquist, F., 1997. Dispersal–vicariance analysis: a new biogeographic approach to
the quantification of historical biogeography. Syst. Biol. 46, 195–203.
Ronquist, F., Huelsenbeck, J.P., 2003. MRBAYES 3: Bayesian phylogenetic inference
under mixed models. Bioinformatics 19, 1572–1574.
Ronquist, F., Huelsenbeck, J.P., van der Mark, P., 2005. MrBayes 3.1 Manual, Draft 5/
26/2005. <http://mrbayes.csit.fsu.edu/manual.php> (last accessed 20.12.11.).
Rutschmann, F., Eriksson, T., Salim, K.A., Conti, E., 2007. Assessing calibration
uncertainty in molecular dating: the assignment of fossils to alternative
calibration points. Syst. Biol. 56, 591–608.
Sanmartín, I., Enghof, H., Ronquist, F., 2001. Patterns of animal dispersal, vicariance
and diversification in the Holarctic. Biol J. Linn. Soc. 73, 345–390.
Schönhofer, A.L., Martens, J., 2010. Hidden Mediterranean diversity: assessing
species taxa by molecular phylogeny within the opilionid family Trogulidae
(Arachnida, Opiliones). Mol. Phylogenet. Evol. 54, 59–75.
Schönhofer, A.L., Martens, J., 2012. The enigmatic Alpine opilionid Saccarella schilleri
gen. n., sp. n. (Arachnida: Nemastomatidae) – isolated systematic placement
inferred from comparative genital morphology. Org. Divers. Evol., http://
dx.doi.org/10.1007/s13127-012-0073-7.
Sharma, P.S., Giribet, G., 2011. The evolutionary and biogeographic history of the
armoured harvestmen – Laniatores phylogeny based on ten molecular markers,
with the description of two new families of Opiliones (Arachnida). Invertebr.
Syst. 25, 106–142.
Shear, W.A., 1975. The opilionid genera Sabacon and Tomicomerus in America
(Opiliones, Troguloidea, Ischyropsalidae). J. Arachnol. 3, 5–29.
Shear, W.A., 1986. A cladistic analysis of the opilionid superfamily
Ischyropsalidoidea, with description of the new family Ceratolasmatidae, the
new genus Acuclavella and four new species. Am. Mus. Novit. 2844, 1–29.
Shear, W.A., 2010. New species and records of ortholasmatine harvestmen from
México, Honduras, and the western United States (Opiliones, Nemastomatidae,
Ortholasmatinae). ZooKeys 52, 9–45.
Shultz, J.W., Regier, J.C., 2009. Pattern and timing of phylogenetic events in Caddo
and implications for the biogeography of Caddidae (Arachnida: Opiliones). J.
Arachnol. 37, 238–241.
Smith, S., 2009. Taking into account phylogenetic and divergence-time uncertainty
in a parametric biogeographical analysis of the Northern Hemisphere plant
clade Caprifolieae. J. Biogeogr. 36, 2324–2337.
Smith, A.G., Smith, D.G., Funnel, B.M., 1994. Atlas of Mesozoic and Cenozoic
Coastlines, first ed. Cambridge University Press, Cambridge.
Suzuki, S., 1974. The Japanese species of the genus Sabacon (Arachnida, Opiliones,
Ischyropsalididae). J. Sci. Hiroshima Univ. Ser. B, Div. 1 25, 83108.
Suzuki, S., 1976. The harvestmen of family Caddidae in Japan (Opiliones, Palpatores,
Caddoidea). J. Sci. Hiroshima Univ. Ser. B, Div. 1 26, 261–273.
Tamura, K., Dudley, J., Nei, M., Kumar, S., 2007. MEGA4: molecular evolutionary
genetics analysis (MEGA) software version 4.0. Mol. Biol. Evol. 24, 1596–1599.
Thaler, K., 1976. Two remarkable relict arachnids from northern Italy: Sabacon
simoni Dresco (Opiliones, Ischyropsalididae), Louisfagea rupicola (Simon)
(Araneae, Tetragnathidae). Bull. Brit. Arachnol. Soc. 3, 205–210.
Thomas, S.M., Hedin, M., 2008. Multigenic phylogeographic divergence in the
paleoendemic southern Appalachian opilionid Fumontana deprehendor Shear
(Opiliones, Laniatores, Triaenonychidae). Mol. Phylogenet. Evol. 46, 645–658.
Tiffney, B.H., 1985. The Eocene North Atlantic land bridge: its importance in Tertiary
and modern phytogeography of the Northern Hemisphere. J. Arnold Arboretum
66, 243–273.
Tsurusaki, N., Song, D., 1993a. Two new species of Sabacon from Sichuan Province,
China (Arachnida: Opiliones: Sabaconidae). Zool. Sci. 10, 155–159.
Tsurusaki, N., Song, D., 1993b. Occurrence of Crosbycus dasycnemus (Crosby)
(Opiliones, Palpatores, Ceratolasmatidae) in China. Jpn. J. Entomol. 61, 175–176.
Vieites, D.R., Min, M.S., Wake, D.B., 2007. Rapid diversification and dispersal during
periods of global warming by plethodontid salamanders. Proc. Nat. Acad. Sci.
USA 104, 19903–19907.
Vink, C.J., Thomas, S.M., Paquin, P., Hayashi, C.Y., Hedin, M., 2005. The effects of
preservatives and temperatures on arachnid DNA. Invertebr. Syst. 19, 99–104.
Wolfe, A.P., Tappert, R., Muehlenbachs, K., Boudreau, M., McKellar, R.C., Basinger,
J.F., Garrett, A., 2009. A new proposal concerning the botanical origin of Baltic
amber. Proc. Royal Soc. B 276, 3403–3412.
Xiang, Q.Y., Soltis, D.E., Soltis, P.S., Manchester, S.R., Crawford, D.J., 2000. Timing the
eastern Asian-eastern North American floristic disjunction: molecular clock
corroborates paleontological estimates. Mol. Phylogenet. Evol. 15, 462–472.
Yu, Y., Harris, A.J., He, X., 2010. S-DIVA (statistical dispersal-vicariance analysis): a
tool for inferring biogeographic histories. Mol. Phylogenet. Evol. 56, 848–850.
Zachos, J., Pagani, M., Sloan, L., Thomas, E., Billups, K., 2001. Trends, rhythms, and
aberrations in global climate 65 Ma to present. Science 29, 686–693.
Zhang, P., Papenfuss, T.J., Wake, M.H., Qu, L., Wake, D.B., 2008. Phylogeny and
biogeography of the family Salamandridae (Amphibia: Caudata) inferred from
complete mitochondrial genomes. Mol. Phylogenet. Evol 49, 586–597.