Discovering the silk road: Nuclear and mitochondrial sequence data resolve the phylogenetic relationships among theraphosid spider subfamilies
Graphical abstract
Introduction
The family Theraphosidae, commonly referred to as “tarantulas”, are a group of mygalomorph spiders which includes some of the largest arachnids on earth. Tarantulas are among the most prominent spiders, widely known to the general public as horror movie performers and also commonly kept as pets (e.g. Klaas, 1989). Given this distinguished position among spiders, it is striking that their deep phylogeny and higher classification remain unstudied from a molecular perspective. The Theraphosidae contain 962 nominal species (World Spider Catalogue, 2017), and most authors accept 11 theraphosid subfamilies (Kambas, 2017) occurring on all continents except Antarctica (West et al., 2008) (Table 1). This diversity is also reflected in different ecological adaptations and natural histories. The group includes arboreal, terrestrial and cave dwelling taxa. Different theraphosid species display a variety of defensive mechanisms such as stridulation, venomous bite or urticating setae (Marshall et al., 1995, Bertani and Guadanucci, 2013). Furthermore, considering their wide range of distribution and their presumably poor dispersal capacity, theraphosids represent good models for biogeographic research.
As is commonplace in arachnological taxonomy (Coddington, 2005, Ramirez, 2014), the systematics of theraphosid spiders is largely based on analysis of morphological characters. Using those, Raven (1985) formulated a first evolutionary hypothesis on intrafamilial relationships, but later recognized the limits of this morphology-based approach by referring to Theraphosidae as a “taxonomic nightmare” (Raven, 1990). Subsequent revisions of theraphosid subfamilies and genera continued using mostly or only morphological characters (e.g. Smith, 1990, Pérez-Miles et al., 1996, Bertani, 2000, Gallon, 2003, Gallon, 2005a, West et al., 2008, West et al., 2012, West and Nunn, 2010, Guadanucci, 2011a, Guadanucci, 2011b, Guadanucci, 2014, Fukushima and Bertani, 2017), but other studies have shown that many traditional morphological characters of spiders can be affected by a high degree of homoplasy (Platnick and Gertsch, 1976, Goloboff, 1993, Bond and Opell, 2002, Hedin and Bond, 2006, Ayoub et al., 2007, Bond et al., 2012). Within Theraphosidae, the poorly defined subfamily Ischnocolinae comprises several taxa of uncertain taxonomical status and has been found to be paraphyletic (Raven, 1985). Several ischnocoline genera show an unresolved phylogenetic position, e.g. Nesiergus, which has been removed from Ischnocolinae (sensu strictu) by Guadanucci (2014) but not given an alternative (therefore herein referred to the Ischnocolinae sensu lato). Besides the Ischnocolinae conundrum, the monophyly and validity of several other subfamilies are also under discussion. Particularly, this regards the Selenocosmiinae (West and Nunn, 2010), Aviculariinae (West et al., 2008, Fukushima and Bertani, 2017) and Eumenophorinae (Guadanucci, 2011a, Guadanucci, 2011b, Mirza et al., 2014), as well as the validity of Poecilotheriinae (Schmidt, 1995, West et al., 2008, West et al., 2012) and Psalmopoeinae (Samm and Schmidt, 2010).
Comparative molecular studies on theraphosids thus far are limited to the species to genus level, for instance in Aphonopelma (Hamilton et al., 2011, Hamilton et al., 2016), Brachypelma (Longhorn et al., 2007, Mendoza and Francke, 2017), Grammostola (Montes De Oca et al., 2015) and Bonnetina (Ortiz and Francke, 2016), whereas phylogenetic relationships among genera and subfamilies remain significantly understudied from a molecular perspective, with few exceptions (e.g., Turner et al., 2017). Several recent studies comprehensively addressed the phylogeny of spiders in general (e.g. Fernandez et al., 2014, Wheeler et al., 2016, Garrison et al., 2016), including representatives of the Theraphosidae but not addressing the phylogeny within the family.
In this study, our goal is to provide a first molecular hypothesis on the broadest scale of theraphosid evolutionary history, i.e., on the relationships among major clades at the subfamily level, serving as a baseline to direct future in-depth studies. We discuss our phylogeny in terms of genus to subfamily level classification. Furthermore, we examine the geographic distribution of major clades identified.
Section snippets
Species sampling
Material was either obtained from wild caught specimens, museum collections or from the pet trade. For most samples we preserved voucher specimens, labeled with preliminary laboratory numbers (TP – Theraphosidae Project); specimens were deposited in the Zoologische Staatssammlung München (ZSM) under ZSM A20170052-A20170080, and in the Oxford University Museum of Natural History (OUMNH), plus pending deposition (Supplementary Table S1). Tissues were preserved in pure ethanol or RNAlater. DNA was
Results
Topologies of exploratory ML gene trees (see Supplementary Materials, Figs. S3–S7) did not present any major and strongly supported disagreements. The concatenated multi-gene alignment had a total length of 3498 bp for 52 taxa. Sequences of nuclear origin, consisting of the genes for 28S (522 bp), 18S (1024 bp) rRNA and Histone H3 (204 bp), represented 50% (1750 bp) of the total alignment. The remaining 50% (1748 bp) were of mitochondrial origin, including fragments of 12S/16S rRNA (1463 bp)
A preliminary molecular perspective on theraphosid phylogeny
Our molecular phylogenetic analysis reliably identified several major clades within Theraphosidae that largely agreed with the current subfamilial classification, but also provided numerous new insights. While we consider the well-supported major clades in our tree as reliable taxonomic groupings, our analysis also is affected by several restrictions.
For specimens obtained via the pet trade, although provenance was vague in several cases, all could be confidently assigned to countries of origin
Acknowledgements
We are grateful to Meike Kondermann and Marc Hoffmann for executing parts of the labwork. Jörg Bernhardt, Matthias Köhler, Richard Gallon, Volker von Wirth, Dirk Weinmann and Jean-Michel Verdez provided sample material to the study. We thank Michael Morra for assistance in dataset generation. Monty Oberländer, Aileen Schlag, Marcus Herrmann and Marlen Ziegler kindly provided and handled animals that were photographed and used to illustrate the phylogenetic tree. Susanne Hauswaldt provided COI
References (71)
- et al.
Utility of the nuclear protein-coding gene, elongation factor-1 gamma (EF-1y), for spider systematics, emphasizing family level relationships for tarantulas and their kin (Araneae: Mygalomorphae)
Mol. Phylogenet. Evol.
(2007) - et al.
Phylogenomics resolves a spider backbone phylogeny and rejects a prevailing paradigm for orb web evolution
Curr. Biol.
(2014) - et al.
Phylogenomic analysis of spiders reveals nonmonophyly of orb weavers
Curr. Biol.
(2014) - et al.
A combined molecular approach to phylogeny of the jumping spider subfamily Dendryphantinae (Araneae: Salticidae)
Mol. Phylogenet. Evol.
(2001) - et al.
Molecular phylogenetics of the spider infraorder Mygalomorphae using nuclear rRNA genes (18S and 28S): conflict and agreement with the current system of classification
Mol. Phylogenet. Evol.
(2006) - et al.
Two DNA barcodes and morphology for multi-method species delimitation in Bonnetina tarantulas (Araneae: Theraphosidae)
Mol. Phylogenet. Evol.
(2016) - et al.
Spiders (Araneae)
Male palpal bulbs and homologous features in Theraphosinae (Araneae, Theraphosidae)
J. Arachnol.
(2000)- et al.
Morphology, evolution and usage of urticating setae by tarantulas (Araneae: Theraphosidae)
Zoologica (Curitiba)
(2013) - et al.
New frog family from India reveals an ancient biogeographical link with the Seychelles
Nature
(2003)