﻿Redescription of the giant Southeast Asian millipede Spirobolusmacrurus Pocock, 1893 and its assignment to the new genus Macrurobolus gen. nov. (Diplopoda, Spirobolida, Pachybolidae)

﻿Abstract A new genus of the millipede family Pachybolidae from Southeast Asia is described: Macrurobolusgen. nov., with Spirobolusmacrurus Pocock, 1893 as type species. This latter species is DNA barcoded (COI) and redescribed based on male morphological characters, which hitherto were unknown. The new genus differs from other pachybolid genera by having (1) the preanal ring process long and protruding beyond the anal valves and (2) the anterior gonopod telopodite distally abruptly narrowed, forming an extremely long, slender, elevated process curved caudad. Given that Macrurobolusgen. nov. is a monotypic genus, it is aphyletic and thus requires further taxonomic revision.


Introduction
Spirobolus macrurus Pocock, 1893 is, with its length of up to 110 mm and diameter of up to 10 mm, the largest pachybolid millipede in SE Asia, but despite its large size, the species is still poorly known. Its original description was based on a single female specimen from Kawkareet, Tenasserim, Myanmar, and did not include the genital parts. Yet, Pocock (1893) separated S. macrurus from other Spirobolus species by its much longer and thinner preanal ring process. Much later, Hoffman (1962: 773) transferred the species to the genus Tonkinbolus Verhoeff, 1938 and remarked "said to be closely related to moulmeinensis, differing only in the longer and more slender epiproct". However, based on gonopod characters and strongly supported by DNA sequence data, Pimvichai et al. (2018) assigned Tonkinbolus scaber Verhoeff, 1938 (type species of Tonkinbolus) to the genus Litostrophus Chamberlin, 1921. Thus, Tonkinbolus became a subjective junior synonym of Litostrophus. At the same time, Pimvichai et al. (2018) moved all other Tonkinbolus species, including T. macrurus, to the genus Atopochetus Attems, 1953 because they share the unique anterior gonopod telopodite of this genus. Yet, since T. macrurus was until then only characterised on the basis of a single female specimen, its transfer to Atopochetus was qualified as "incertae sedis" (Pimvichai et al. 2018).
In the present paper we redescribe and barcode Spirobolus macrurus based on an old male specimen discovered in the collections of the Natural History Museum of Denmark, Copenhagen, and new live material, including an adult male specimen, collected during recent fieldwork in Thailand. As a result we also create the new genus Macrurobolous gen. nov. to accommodate Spirobolus macrurus, so that this species will be referred to as Macrurobolus macrurus comb. nov.

Material and methods
Live specimens were hand collected and preserved in 70% ethanol for morphological study or placed in a freezer at -20 °C for DNA analysis. Specimens were also examined from the following collections: CUMZ Museum of Zoology, Chulalongkorn University, Bangkok, Thailand; NHMD Natural History Museum of Denmark, University of Copenhagen, Denmark.
This research was conducted under the approval of the Animal Care and Use regulations (numbers U1-07304-2560 and IACUC-MSU-037/2019) of the Thai government.

Morphology
Gonopods were photographed with a digital camera manipulated via the program Helicon Remote (v. 3.1.1.w). The Zerene Stacker Pro software was used for image-stacking. Drawings were made using a stereomicroscope. Samples for scanning electron microscopy (SEM) were air-dried directly from alcohol and sputter-coated for 250 s with gold. SEM micrographs were taken with an environmental scanning electron microscope (ESEM)-FEI Quanta 200. Voucher specimens were deposited in the collections of CUMZ and NHMD.

DNA extraction, amplification, and sequencing
Total genomic DNA was extracted from legs of a male specimen of Macrurobolus macrurus, comb. nov. from Wat Tham Inthanin, Mae Sot District, Tak Province, Thailand (CUMZ-D00147) using the NucleoSpin Tissue kit (Macherey-Nagel, Düren, Germany) following the manufacturer's instructions. PCR amplifications and sequencing of the standard mitochondrial COI DNA barcoding fragment (Hebert et al. 2003) were done as described by Pimvichai et al. (2020). The COI fragment was amplified with the primers LCO-1490 and HCO-2198 (Folmer et al. 1994). The new COI nucleotide sequence has been deposited in GenBank under accession number MZ905519. Sample data and voucher codes are provided in Table 1.
CodonCode Aligner (v. 4.0.4, CodonCode Corporation) was used to assemble the forward and reverse sequences and to check for errors and ambiguities. Sequences were checked with the Basic Local Alignment Search Tool (BLAST) provided by NCBI and compared with reference sequences in GenBank. Next, sequences were aligned using MUSCLE (v. 3.6, see http://www.drive5.com/muscle; Edgar 2004 Phylogenetic trees were constructed using maximum likelihood (ML), Bayesian inference (BI), and neighbor-joining (NJ). The shape parameter of the gamma distribution, based on 16 rate categories, was estimated using maximum-likelihood analysis. ML trees were inferred with RAxML (v. 8.2.12, see http://www.phylo.org/index.php/ tools/raxmlhpc2_tgb.html; Stamatakis 2014) through the CIPRES Science Gateway (Miller et al. 2010) using a GTR+G substitution model and 1000 bootstrap replicates to assess branch support. BI trees were constructed with MrBayes (v. 3.2.7a, see http://  (Miller et al. 2010). BI trees were run for 2 million generations (heating parameter was 0.05), sampling every 1000 generations. Convergences were confirmed by verifying that the standard deviations of split frequencies were below 0.01. Then the first 1000 trees were discarded as burn-in, so that the final consensus tree was built from the last 3002 trees. Support for nodes was assessed by posterior probabilities. NJ trees were constructed with MEGA v. X using the Kimura 2-parameter model and 1000 bootstrap replicates.
For ML and NJ trees we consider branches with bootstrap values (BV) of ≥ 70% to be well supported (Hillis and Bull 1993) and < 70% as poorly supported. For BI trees, we consider branches with posterior probabilities (PP) of ≥ 0.95 to be well supported (San Mauro and Agorreta 2010) and below as poorly supported.

Results
The uncorrected p-distance between the sequences ranged from 0.03 to 0.25 (Tables  2, 3). The mean interspecific sequence divergence within Atopochetus was 0.13 (range: 0.08-0.16). The mean sequence divergence between Atopochetus and M. macrurus comb. nov. was 0.15 (range: 0.14-0.17). The mean interspecific sequence divergence within Litostrophus was 0.10 (range: 0.09-0.11). The mean sequence divergence between Litostrophus and M. macrurus comb. nov. was 0.13 (range: 0.11-0.14). PartitionFinder indicated that the best substitution model for BI analysis was GTR+ G. The ML, BI, and NJ trees were congruent with respect to some of the well-supported branches (by visual inspection of the branching pattern). Yet, in several instances BI provided good support for branches that were not well-supported by both ML and NJ (e.g., the Litostrophus + Benoitolus clade or the Coxobolellus + Pseudospirobolellus clade).
In the phylogenetic trees ( Fig. 1) the clade of Pachybolidae + Benoitolus is poorly supported by ML (BV = 63) and NJ (BV = 27), but well supported by BI (PP = 0.97), while Trigoniulinae is well supported by the three methods (BV = 96 and 92; PP = 1.00). Although the monophyly of Pachybolidae is clearly challenged by the inclusion of Benoitolus, which involves a long branch, removing Benoitolus from the analysis yields a Pachybolidae clade with the same pattern of support as the Pachybolidae + Benoitolus clade (Suppl. material 1).
Irrespective of the in-or exclusion of Benoitolus, Macrurobolus macrurus comb. nov. is nested within a clade comprising Litostrophus and Atopochetus. Yet, this clade is poorly supported by ML, well supported (but just so) by NJ, and convincingly well supported by BI. The position of M. macrurus comb. nov. within this clade, however, is poorly supported by the three methods.

Diagnosis.
A genus of Pachybolidae characterised by the following combination of characters: preanal ring with long process protruding beyond anal valves; the anterior gonopod telopodite distally abruptly narrowed, forming an extremely long, slender, elevated process curved caudad.
Collum smooth, with a marginal furrow along lateral part of anterior margin; lateral lobes narrowly rounded, extending as far ventrad as the ventral margin of body ring 2.
Colour. Living animal reddish brown except for grey pro-and mesozona (Fig. 4). Male sexual characters. Tarsus from third to before the last 4 body rings with large ventral soft pad occupying entire ventral surface. Body ring 7 entirely fused ventrally, no trace of a suture. Tip of anterior gonopods visible when the animal is stretched out (not when it is rolled up). Anterior gonopods (Fig. 3A, B, D, E) with triangular mesal sternal process, not reaching so far as the tip of coxae, apical margin bilobed, with basal longitudinal triangular ridge in posterior view. Coxa oval, apically gradually narrowed, rounded, projecting slightly beyond sternal process. Telopodite apically far overreaching coxa, distally abruptly narrowed, forming an extremely long, slender, elevated process curved caudad.
Posterior gonopods (Fig. 3C, F, H-I) strongly curved mesad, laterally with a massive ridge; with efferent canal (Enghoff 2011) running along mesal margin terminating in slender, pointed meso-distad process, covered with fine hairlike spinules      Female vulvae (Fig. 3G). Valves prominent, of equal size; basally with open space between free margins. DNA barcode. The GenBank accession number of the COI barcode of the Thai specimen is MZ905519 (voucher code CUMZ-D00147).
Ecology. Found under leaf litter. Notes on the male from Meetan, Myanmar. This specimen is labelled as "ex typ" in the NHMD collection and was, like the female type specimen, collected by Fea. It agrees with the Thai male in all characters, including all details of gonopod shape, with the following exceptions: Colour after > 100 years in alcohol is faded, but there is still a clear contrast between greyish pro-and mesozone and reddish-brown metazona. Size: length ca 8 cm, diameter 6.7 mm, 50 podous rings, no apodous rings in front of telson. Head capsule smooth. 11 vertical rows of ommatidia, of which 3 are very incomplete, 7 horizontal rows, 47 ommatidia per eye. Antennomeres 2-4 with some ventral setae, 5 and 6 densely setose. Gnathochilarium not dissected.

Discussion
The male specimen of Spirobolus macrurus from Meetan in NHMD, although labelled "ex typ.", should not a priori be regarded as a type (ICZN Art. 72.4.7.) because Pocock (1893: 396) explicitly mentioned that the species description was based on "A single ♀ from Kawkareet (Tenasserim)". However, its non-sexual characters agree with Pocock's (1893) description. Hence, we do not hesitate to refer it to Macrurobolus macrurus comb. nov.
The new male specimen from Thailand and the old specimen from Myanmar share the long preanal ring process with the female type specimen, which is a remarkable character for a pachybolid, since most pachybolid genera (except Aulacobolus Pocock, 1903 and Trachelomegalus Silvestri, 1896) have a short preanal ring process. So, in this respect, Macrurobolus gen. nov. is clearly differentiated from most other pachybolid genera, including Atopochetus and Litostrophus, the two genera with which Macrurobolus gen. nov. appears the be most closely related in our phylogenetic tree (Fig. 1). Similarly, the anterior gonopod telopodites of Macrurobolus (telopodite distally abruptly narrowed, forming an extremely long, slender, elevated process curved caudad) clearly differ from those of Litostrophus (telopodite simple, without process, narrowly rounded) or Atopochetus (telopodite with a triangular process directed laterad originating on posterior surface at ~1/2 or 2/3-4/5 of its height). Hence, given that Macrurobolus shares neither the defining morphological synapomorphies of Atopochetus, nor those of Litostrophus, we think that the creation as a separate monotypic genus is warranted.
The interpretation of Macrurobolus as a separate genus is somehow in line with the COI tree ( Fig. 1), which places the new genus in a clade comprising Atopochetus and Litostrophus, but which supports neither joining M. macrurus comb. nov. with Atopochetus (which itself forms a consistently well-supported clade), nor joining it with Litostrophus (which itself forms also a well-supported clade) (Fig. 1). Moreover, the mean interspecific COI sequence divergence between M. macrurus and other pachybolid and pseudospirobolellid species is 18% (range: 11-23%) (Tables 2, 3), a value that rather points to an intergeneric divergence (Table 3).
In conclusion, this study suggests that Pimvichai et al. (2018) appropriately labelled the transfer of Tonkinbolus macrurus to the genus Atopochetus as "incertae sedis". Indeed, the species can be accommodated in neither Atopochetus nor Litostrophus, i.e., the two genera with which it appears to be most closely associated. Hence, it would be ill-advised to maintain Macrurobolus macrurus comb. nov. in the genus Atopochetus, for this would undermine both the definition and the support of the monophyly of this taxon. Therefore, the creation of the monotypic genus Macrurobolus gen. nov. seems the best solution to provide a generic name for Spirobolus macrurus Pocock, 1893. Still, the monotypy of Macrurobolus gen. nov. renders it aphyletic sensu Ebach and Williams (2010), and hence in need of further study (Williams and Ebach 2020: 134).
al Belgian Institute of Natural Sciences (RBINS). We thank, Pongpun Prasankok for assistance in collecting material. We are indebted to Julien Cillis (RBINS) for help with SEM photographs, to Yves Barette (RBINS) for help with gonopod photographs and to Thita Krutchuen for the excellent drawings. Last but not least, we thank Thomas Wesener (Bonn) and Sergei Golovatch (Moscow) for their helpful reviews of this paper.