Research Article |
Corresponding author: Marleen R. Greenberg ( marleen.greenberg@outlook.de ) Academic editor: Lorenzo Prendini
© 2021 Marleen R. Greenberg, Joel A. Huey, Volker W. Framenau, Danilo Harms.
This is an open access article distributed under the terms of the Creative Commons Attribution License (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Citation:
Greenberg MR, Huey JA, Framenau VW, Harms D (2021) Three new species of mouse spider (Araneae: Actinopodidae: Missulena Walckenaer, 1805) from Western Australia, including an assessment of intraspecific variability in a widespread species from the arid biome. Arthropod Systematics & Phylogeny 79: 509-533. https://doi.org/10.3897/asp.79.e62332
|
Abstract
Mouse spiders (genus Missulena Walckenaer, 1805) are a lineage of trapdoor spiders with males of many species having a brightly coloured red cephalic region, an abdomen that is tinged metallic blue, and the habit of wandering during the day in search of a mate. A total of 17 species of Missulena have been described in Australia to date but most descriptions are based exclusively on males and always small numbers of specimens. Here, we describe three new species of Missulena from the Pilbara and Goldfields regions of Western Australia based on morphology and genetic data: Missulena davidi sp. nov. (male and female), M. iugum sp. nov. (male) and M. manningensis sp. nov. (male). One of them is presently known only from its type locality and another one from a small range based on two specimens but M. davidi sp. nov. has a linear range of almost 300 km and is genetically highly structured. We use genetic data for 75 specimens as a foundation to evaluate morphological variability in this species and note substantial variation in several characters commonly used to identify species such as body size, colouration, rastellum shape and eye distances. This variation does not necessarily relate to phylogeographic structure as inferred from the genetic data, but rather seems to reflect natural variability both within and between localised populations. Overall, our results stress the need to evaluate a large series of specimens for mygalomorph taxonomy and provide an interesting example of intraspecific variability in hard-to-collect species that are usually underrepresented in museum collections.
Mygalomorphae, Pilbara, speciation, phylogeny, systematics, taxonomy
Mouse spiders (genus Missulena Walckenaer, 1805) are a lineage of trapdoor spiders from Australia and Chile. They are ambush hunters and build burrows that are lined with silk and usually sealed with one or more lids (
Species delineation in mygalomorph spiders can be challenging because many species are difficult to collect, morphologically conservative, and the taxonomic framework for species identification is often poor (e.g.,
In this paper, we describe three new species of Missulena from central and north-western Western Australia. Two of these new species follow the “typical” pattern of rarity in research collections and taxonomic descriptions that are based on a few males from single localities, although the morphological descriptions are backed up by a molecular phylogenetic framework. A third species is interesting insofar as it seems to be widespread across a wide area in north-western Australia, is morphologically and genetically variable across its range, and commonly collected. We use Missulena davidi sp. nov. not only to document female morphology based on a large sample size for the first time in this genus, but also to assess morphological variability of taxonomic key features in light of a sound molecular phylogenetic framework. A qualitative description of the morphological characters that are rather conserved versus those that show substantial variability within species will aid future species descriptions.
There were 89 Missulena cytochrome c oxidase I (COI) sequences available on GenBank, produced by previous studies (
The 167 Missulena sequences were aligned with two outgroup taxa, a specimen of Actinopus KY017543 (family Actinopodidae) and a specimen of Atrax KY017748 (Atracidae). Recent phylogenetic studies have shown that Atracidae is closely related to Actinopodidae (
All specimens chosen for morphological study belong to a monophyletic group comprising 70 specimens that was recovered by the phylogenetic analyses (Fig.
Maximum Likelihood phylogeny of all Missulena specimens. The ultimate outgroup taxon Atrax has been removed from the figure for convenience. All bootstrap values below 80 have been removed. Genetic clades within Missulena davidi sp. nov. are colour-coded: clade I brown, clade II red, clade III blue, clade IV pink.
We selected 26 adult females and 12 adult males of sequenced M. davidi sp. nov. from across the species range and assessed a total of 34 morphological characters, four of these qualitative (e.g., body colouration) and 30 quantitative (e.g., body measurements and spine count of the rastellum). All specimens are deposited at the Western Australian Museum in Perth, Australia (
A Leica DM4500 digital camera attached to a Leica M205A stereomicroscope and controlled by the Leica Application Suite X Version 3.0.1. was used for examination and morphological measurements in millimetres. Digital images were taken with a BK Plus Lab System by Dun, Inc. with integrated Canon EOS 5D Mark III and the program Capture One 9, or with a Hitachi TM4000Plus scanning electron micrograph (SEM). The images were stacked with Zerene stacking software (Zerene Systems LLC 2018). The drawings were created with the help of printed images that were traced on a Comicstar light table and scanned afterwards. Images were edited with Adobe Photoshop CS6. Maps were created using QGIS Version 3.0.
Box-Whisker-Plots for carapace and eye variation measurements as well as rastellum counts in M. davidi sp. nov. were compiled in RStudio Version 1.2.1335 (RStudio, Inc). For the comparison of leg spination and cheliceral dentition five specimens of each sex were examined in detail. The cuspules on the maxillae and labium were counted in the ventral position and represent minimum values because not all cuspules near the joints could be seen clearly. Carapace height was measured laterally from the highest point of the carapace vertically to the edge of the carapace.
Abbreviations used in the taxonomic sections are as follows: OQ ocular quadrangle, AME anterior median eyes, ALE anterior lateral eyes, PME posterior median eyes, PLE posterior lateral eyes, PLS posterior lateral spinnerets, PMS posterior median spinnerets, rl retrolateral, v ventral, pl prolateral, d dorsal.
Missulena Walckenaer, 1805: 8. Type species: Missulena occatoria Walckenaer, 1805, by monotypy.
Eriodon
Latreille, 1806: 85. Type species: Eriodon occatorius Latreille, 1806, by monotypy. Synonymised by
Holotype: AUSTRALIA – Western Australia • ♂; Juna Downs Station, 113 km NW of Newman; 22°41.23′S 118°53.55′E; 10 May 2011; C. Cole and P. Runham leg.; pit trap;
AUSTRALIA – Western Australia • 1♀; Carnarvon, 99 Gascoyne Road; 24°53′S 113°39′E; 23 July 2002; residents leg.; by hand;
Males of Missulena davidi sp. nov. share the red colouration of chelicerae and pars cephalica with M. langlandsi Harms and Harvey, 2013, M. occataria Walckenaer, 1805, M. insignis O. Pickard-Cambridge, 1877, M. iugum sp. nov. and M. manningensis sp. nov. that are morphologically most similar. They differ from M. langlandsi by having strong, conical spines of the rastellum (thin and not conical in the former) and a longer carapace (>3.00 mm; M. langlandsi up to 2.8 mm). They differ from M. occataria and M. insignis by having spines on patellae III and IV only and not on all four legs (on patellae I and II 1 spine, respectively). Missulena davidi sp. nov. males have more cuspules on maxillae and labium than those of M. manningensis sp. nov. (M. manningensis sp. nov.: 5 at labium, 30 at maxillae; M. davidi sp. nov.: 15–10 at labium, 35–100 at maxillae). Missulena davidi sp. nov. males differ from M. iugum sp. nov. by the ridge present in the cheliceral groove. Females of Missulena davidi sp. nov. have uniformly red chelicerae that they share with M. insignis; however, the fourth leg of M. davidi sp. nov. is the longest of all legs, whilst in M. insignis the longest leg is the first. Additionally, there are no cuspules recorded on the labium or the maxillae in M. insignis females.
MALE (based on holotype;
Missulena davidi sp. nov. Male holotype (WAMT119725): A habitus, dorsal view; B same, ventral view; C carapace, dorsal view; D abdomen, dorsal view; E same, ventral view; F maxillae, labium, and chelicerae, ventral view; G carapace, lateral view. Male paratype (WAMT119727), left pedipalp, H embolus with embolar tooth, prolateral view; I same, retrolateral view. Scale bars: A, B 4.0 mm; C–G 2.0 mm; H 100 µm; I 40 µm.
Missulena davidi sp. nov. Male holotype (WAMT119725): A chelicerae with cheliceral groove, ventral view; B patella III, dorsal view; C patella IV, dorsal view; D sternum, ventral view; E eye region, dorsal view; F rastellum, frontal view; G right pedipalp, retrolateral view; H same, ventral view; I same, prolateral view; J pattern of cheliceral teeth in cheliceral groove. Scale bars: A, D–H 2.0 mm; B, C 0.5 mm; J 1.0 mm.
FEMALE (based on allotype;
Missulena davidi sp. nov. Female allotype (WAMT107393): A carapace, dorsal view; B abdomen, dorsal view; C sternum, ventral view; D carapace, lateral view; E maxillae, labium and chelicerae, ventral view; F patella III, dorsal view; G patella IV, dorsal view; H eye region, dorsal view; I rastellum, frontal view. Scale bars: A, B 4.0 mm; C–I 2.0 mm.
Missulena davidi sp. nov. Variability of spermatheca in females: A allotype specimen WAMT107393, clade I; B specimen WAMT119995, clade IV; C specimen WAMT126272, clade II; D specimen WAMT122226, clade III. Systematic drawings based on allotype WAMT107393: E spermatheca; F pattern of cheliceral teeth in the cheliceral groove. Scale bars: A–E 0.5 mm; F 2.0 mm
The specific epithet is a patronym in honour of the senior author’s husband, David A. Greenberg.
Pilbara region of Western Australia, excluding the northern Pilbara subregion, extending into the Little Sandy Desert region. The known linear range of this species is 295 km (Fig.
This species had been labelled “MYG045” in previous barcoding studies (
Assessment of 19 characters in male M. davidi sp. nov. and 18 characters in females (Table
Summary of scored characters in M. davidi sp. nov., M. manningensis sp. nov. and M. iugum sp. nov.. M = Median, n = sample size.
Characters | M. davidi–males (n = 12) | M. davidi–females (n = 26) | M. manningensis (n = 1) | M. iugum (n = 2) | |||||||
Clades | All clades combined | Clades | All clades combined | ||||||||
I (n = 4) | II (n = 4) | III (n = 4) | I (n = 10) | II (n = 7) | III (n = 7) | IV (n = 5) | |||||
Carapace length [mm] | 3.86–4.2 | 3.95–4.74 | 3.58–4.11 | 3.58–4.74 M: 4.03 | 5.19–8.55 | 5.12–8.54 | 6.75–8.59 | 6.21–8.76 | 5.12–8.97 M: 7.91 | 3.6 | 3.68–3.87 |
Carapace width [mm] | 4.67–4.96 | 4.56–5.55 | 4.23–4.88 | 4.23–5.55 M: 4.77 | 6.28–10.7 | 6.71–10.44 | 8.39–12 | 8.14–10.85 | 6.28–12 M: 10.28 | 4.61 | 4.57–4.98 |
Carapace height [mm] | 1.73–2.54 | 1.92–2.24 | 1.75–2.31 | 1.73–2.54 M: 2.16 | Not measured | 1.96 | 1.67–1.69 | ||||
Pars cephalica ratio to carapace (in %) | 63.4–64.5 | 63.2–67.3 | 62.0–64.8 | 62–67% M: 64% | 62.6–70.3 | 65.6–74.6 | 61.2–71.9 | 62.5–73.4 | 61–75% M: 67% | 63% | 60-62% |
PLE ratio to ALE (in %) | 90.5–98.2 | 89.7–95.4 | 92.4–96.0 | 90–98% M: 94% | 84.3–93.0 | 84.6–90.4 | 86.1–94.3 | 85.5–94.7 | 84–95% M: 90% | 95% | 87–90% |
PME ratio to ALE (in %) | 60.7–64.5 | 59.6–62.9 | 59.0–64.1 | 59–64% M: 62% | 57.0–67.8 | 55.7–60.6 | 52.2–65.0 | 56.9–65.9 | 52–68% M: 60% | 64% | 65–69% |
Cuspules right maxillae | 50–80 | 55–60 | 50–90 | 50–90 M: 60 | 75–200 | 130–180 | 120–230 | 130 -200 | 75–230 M: 170 | 30 | 70–100 |
Cuspules left maxillae | 45–70 | 55–80 | 35–70 | 35–80 M: 60 | 80–190 | 110–160 | 120–220 | 90–150 | 80–220 M: 150 | 30 | 70–85 |
Cuspules labium | 25–30 | 15–25 | 20–30 | 15–40 M: 25 | 35–65 | 30–55 | 35–100 | 25–65 | 25–100 M: 48 | 5 | 30–35 |
Ridge in the Cheliceral groove | Not present | Not present | Not present | Present | |||||||
Right rastellum | 5–8 | 4–9 | 5–8 | 4–9 M: 6 | 8–15 | 11–17 | 7–18 | 5–12 | 5–18 M: 12 | 5 | 8 |
Left rastellum | 4–8 | 5–10 | 6–9 | 4–10 M: 8 | 7–19 | 11–17 | 4–16 | 7–15 | 4–19 M: 12 | 8 | 8–9 |
Spination 1st patella | Not individually measured | rl: 0; v: 0–1; pl: 1–7; d: 0 | Not individually measured | rl: 0; v: 0; pl: 0; d: 0 | rl: 0; v: 1; pl: 8; d: 0 | rl: 0; v: 3–9; pl: 9–12; d: 0 | |||||
Spination 2nd patella | rl: 0; v: 1; pl: 0; d: 0 | rl: 0; v: 0; pl: 0; d: 0 | rl: 0; v: 1; pl: 1; d: 0 | rl: 0; v: 2; pl: 0; d: 0 | |||||||
Spination 3rd patella | rl: 1–2; v: 0; pl/d: 23–25 | rl: 0; v: 0; pl/d: 26–28 | rl: 1; v: 0; pl/d: 26 | rl: 4; v: 2–3; pl/d: 27 | |||||||
Supination 4th patella | rl: 0–1; v: 0; pl: 16–18; d: 1 + ~12 (small) | rl: 0; v: 0; pl/d: 17–26 | rl: 0; v: 0; pl: 14 (small); d: 15 (small) | rl: 0; v: 3 (or just thickened setae); pl: 6 (small); d: 7 (small) |
In both sexes there is substantial variation in body colouration. In males, the chelicera and pars cephalica ranged from a bright red (specimen
Variability was also high in rastellum spination and cuspule counts. Some male specimens had about twice as many rastellum spines (Fig.
Eye ratio of the PLE and PME pair width in relation to the ALE pair width varied less than other characters. While still showing variation in males and females its range was less extreme (within 8% in males and 16% in females) than in other characters (Figs
There was no correlation between geographical distance of samples and morphological divergence. For example, two females of the clade III (
Holotype: AUSTRALIA – Western Australia • ♂; Mt Ida, 80 km NW of Menzies; 29°12.5′S 120°24.48′E; 29 Mar. 2012; V. Saffer leg.; pitfall trap;
Males of Missulena iugum sp. nov. share the red chelicerae and pars cephalica with M. davidi sp. nov., M. manningensis sp. nov., M. langlandsi, M. occataria and M. insignis that are morphologically most similar. They differ from M. langlandsi by a longer carapace (>3.00 mm; M. langlandsi up to 2.8 mm) and a rastellum with strong, conical spines (simple in M. langlandsi). Rastellum and cuspules on labium and maxillae stronger than in M. insignis (rastellum: 8–9 spines; M. insignis 2–5; cuspules: M. insignis none). Pars cephalica lower than in M. occatoria (up to 1.69; M. occataria approx. 3.0) and carapace shorter (3.87 long, 4.98 wide; M. occataria approx. 5.0 long, 7.0 wide). Differs from M. davidi sp. nov. and M. manningensis sp. nov. by the presence of a ridge in the cheliceral groove (Fig.
MALE (based on holotype;
Missulena iugum sp. nov. Male holotype (WAMT123110): A habitus, dorsal view; B ventral view; C carapace, dorsal view; D abdomen, dorsal view; E same, ventral view; F maxillae, labium, and chelicerae, ventral view; G eye region, dorsal view; H rastellum, anterior view. Scale bars: A, B 4.0 mm; C–H 2.0 mm.
Missulena iugum sp. nov. Male holotype (WAMT123110): A patella III, dorsal view; B patella IV, dorsal view; C sternum, ventral view; D right pedipalp, retrolateral view; E same, ventral view; F same, prolateral view; G cheliceral teeth in cheliceral groove, arrows pointing to the ridge on both sides of the groove; H carapace, lateral view; I pattern of cheliceral teeth in cheliceral groove. Male paratype (WAMT110243), right pedipalp (left pedipalp not complete): J embolus with embolar tooth, prolateral-dorsal view; K embolus with lamella and tooth, retrolateral view. Scale bars: A, B, G, I 1.0 mm; C–F, H 2.0 mm; J 40 µm; K 50 µm.
The specific epithet is a Latin noun (iugum = ridge) in apposition, referring to the strongly developed ridges along the cheliceral groove of the males.
Known only from the Mt Ida region approximately 16 km east of Ularring in the Goldfields region of Western Australia (Fig.
Holotype: AUSTRALIA – Western Australia • ♂; Mt Manning area, site CR2; 30°27.95′S 119°58.02′E; 21 June 2008; J. Francesconi leg.;
Males share with M. davidi sp. nov., M. iugum sp. nov., M. langlandsi, M. occatoria and M. insignis, the closest morphological matches, the red colouration of the chelicerae and pars cephalica. They differ from M. langlandsi by a longer carapace (>3.00 mm; M. langlandsi up to 2.8 mm) and the presence of strong, conical spines of the rastellum (simple in M. langlandsi). They differ from M. occataria and M. insignis by the lack of spines ventrally on patellae III and IV (at the most 1 thickened seta). Pars cephalica lower than in M. occataria (up to 1.96; M. occataria approx. 3.0) and carapace shorter (3.6 long, 4.61 wide; M. occataria approx. 5.0 long, 7.0 wide). More cuspules on the labium and maxillae than in M. insignis but less than in M. davidi sp. nov. (M. insignis: none; M. manningensis sp. nov.: 5 at labium, 30 at maxillae; M. davidi sp. nov.: 15–10 at labium, 35–100 at maxillae). Lacks a ridged cheliceral groove which is present in M. iugum sp. nov.
MALE (based on holotype; WAMT92071). Total length 8.95. Colour: pars cephalica and chelicerae orange (Fig.
Missulena manningensis sp. nov. Male holotype (WAMT92071): A habitus, dorsal view; B same, ventral view; C carapace, dorsal view; D abdomen, dorsal view; E abdomen, ventral view; F maxillae, labium, and chelicerae, ventral view; G sternum, ventral view; H eye region, dorsal view; I rastellum, anterior view. Scale bars: A, B 5.0 mm; C–I 2.0 mm.
Missulena manningensis sp. nov. Male holotype (WAMT92071): A right pedipalp, retrolateral view; B ventral view; C prolateral view; D carapace, lateral view; E patella III, dorsal view; F patella IV, dorsal view; G pattern of cheliceral teeth in cheliceral groove; H left pedipalp, embolus with embolar tooth, prolateral-dorsal view; I left pedipalp, embolus with lamella, retrolateral-ventral view. Scale bars: A–D 2.0 mm; E–G 1.0 mm; H, I 40 µm
The specific epithet refers to the type locality, Mt Manning, in the Goldfields region of Western Australia.
Known only from the Mt Manning area approximately 47 km northwest of Boorabbin in the Goldfields region of Western Australia (Fig.
This study highlights the difficulty in diagnosing Missulena species morphologically in the absence of a molecular framework (e.g., a barcoding framework in our study). There clearly is substantial intraspecific variation if large specimen series are available for taxonomic descriptions, but there also is morphological overlap in species that are clearly defined at the genetic level, as illustrated here by our species triplet that overlaps morphologically in several diagnostic characters (Figs
Variability in Missulena davidi sp. nov. males: A carapace, dorsal view (T119729, clade III); B carapace, dorsal view (T120081, clade II); C carapace, dorsal view (T119727, clade II); D carapace, dorsal view (T119733, clade III); E sternum, ventral view (T113596, clade I); F sternum, ventral view (T84005, clade I); G abdomen, dorsal view (T119727, clade II); H abdomen, dorsal view (T125176, clade III); I eye region, dorsal view (T120081, clade II); J eye region, dorsal view (T84005, clade I); K eye region, dorsal view (T119729, clade III); L rastellum, anterior view (T84005, clade I); M rastellum, anterior view (T119726, clade II). Scale bars: A–M 2.0 mm
Variability in Missulena davidi sp. nov. females: A carapace, dorsal view (T116874, clade IV); B carapace, dorsal view (T116776, clade I); C carapace, dorsal view (T113591, clade III); D carapace, dorsal view (T122226, clade III); E carapace, dorsal view (T126272, clade II); F sternum, ventral view (T116776, clade I); G sternum, ventral view (T119711, clade I); H left maxilla, ventral view (T119979, clade IV); I right maxillae, ventral view (T116868, clade I); J eye region, dorsal view (T116839, clade I); K eye region, dorsal view (T119979, clade IV); L eye region, dorsal view (T122865, clade I); M eye region, dorsal view (T126264, clade III); N rastellum, anterior view (T125316, clade III); O rastellum, anterior view (T102165, clade I). Scale bars: A–G 4.0 mm; H–O 2.0 mm.
Our analyses suggest that females are substantially more variable than males, and that most of the morphological characters that are used in Missulena taxonomy are also variable. Characters that appear to be most useful in distinguishing species, at least in males, are the the number of cuspules on maxillae and labium, the positioning of eyes, specifically the PME and PLE width in relation to the ALE width, and the overall shape and colour of the carapace. Whilst variable and subject to overlap between species, the number of cuspules is a good character for our species triplet but perhaps also other Missulena species for which original data are available (e.g.,
Characters that we found to be highly variable and of little use, both within and between species, are body colouration (often subject to variable storage conditions), sternum size and shape, position and shape of sigilla (specifically variable in females), and the number of spines on the rastellum. This result is irrespective of potential changes in the tree topology in the phylogeny if additional markers (e.g., nuclear genes) are used. A further character that should be treated with caution is the number and position of cheliceral teeth and a groove itself (e.g., ridge present in M. iugum sp. nov.; distal teeth blade in M. faulderi). For species delineation in females, we recommend evaluating patellae spination patterns further because there was little variability and the number of spines in M. davidi sp. nov. females (see Table
Although our phylogenetic analysis is based on a COI dataset rather than a multi-gene dataset that incudes additional nuclear markers, there may be several general implications. First of all, it is clear that Missulena is highly diverse at the genetic level (Fig.
Secondly, it is evident that females are substantially more variable than males and species identification based on morphology is harder, if not impossible. This may be trivial but considering that we provide the first detailed description of female Missulena based on a large sample size, this reinforces what is already known in other lineages of trapdoor spiders in Australia (e.g., Atracidae: Hadronyche, Gray 2019; Halonoproctidae: Conothele,
The third outcome is that variability in the characters examined is not simply an expression of isolation by distance at the genetic level but rather an expression of natural variability. Mitochondrial data alone have several limitations and the phylogenetic results given here may be affected by incomplete lineage sorting and/or even saturation at the deeper nodes of the phylogeny. However, we are assessing variability at shallow notes, in particular population structure, where the COI gene is clearly one of the most effective markers. Here it is interesting that there is substantial variation even within localised populations for several morphological characters we scored although COI sequences were (almost) identical. The examination of large specimen series will provide an account to such natural variability that may or may not have a genetic base in the nuclear genome.
Another outcome is that not necessarily all Missulena species have restricted ranges and that at least some species are quite widespread. This is perhaps not a surprising outcome given that several species, such as and M. bradleyi Rainbow, 1904, M. occatoria Walckenaer, 1905 and M. dipsaca Faulder, 1995, were always considered widespread (e.g., Wormersley 1943;
One may argue in the case of M. davidi sp. nov. that both the genetic and morphological variability may point to a radiation of closely related species that are cryptic, similar to Bertmainius tingle (Main 1991) in southwestern Australia for which a dataset comprising two genetic markers (COI and ITS-2; see
Research using genomic tools could help further resolve the species boundaries in this radiation, providing a more robust molecular framework to test species concepts (e.g.,
MG thanks Bennelongia Environmental Consultants (Jolimont, WA) and Stuart Halse in particular for supporting this research through a travel grant, the Western Australian Museum team for great company during her stay in 2019, Nadine Dupérré (LIB Hamburg) for technical support, and Tabea Schneider for company during the trip to Australia. All authors declare no conflict of interest. DNA sequencing of new specimens for this study was funded by the Gorgon Barrow Island Net Conservation Benefits Fund. The Fund is administered by the Department of Biodiversity, Conservation and Attractions and approved by the Minister for Environment after considering advice from the Gorgon Barrow Island Net Conservation Benefits Advisory Board. We finally thank Robert Raven and Michael Rix (Queensland Museum, Brisbane) for reviewing earlier drafts of this manuscript.
GenBank registration numbers
Data type: .xlsx
Explanation note: GenBank registration numbers and museum voucher numbers for all specimens that were sequenced in this study.