Research Article |
Corresponding author: Somsak Panha ( somsak.pan@chula.ac.th ) Academic editor: Didier Vanden Spiegel
© 2020 Natdanai Likhitrakarn, Sergei I. Golovatch, Ekgachai Jeratthitikul, Ruttapon Srisonchai, Chirasak Sutcharit, Somsak Panha.
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:
Likhitrakarn N, Golovatch SI, Jeratthitikul E, Srisonchai R, Sutcharit C, Panha S (2020) A remarkable new species of the millipede genus Trachyjulus Peters, 1864 (Diplopoda, Spirostreptida, Cambalopsidae) from Thailand, based both on morphological and molecular evidence. ZooKeys 925: 55-72. https://doi.org/10.3897/zookeys.925.49953
|
A new, giant species of Trachyjulus from a cave in southern Thailand is described, illustrated, and compared to morphologically closely related taxa. This new species, T. magnus sp. nov., is much larger than all other congeners and looks especially similar to the grossly sympatric T. unciger Golovatch, Geoffroy, Mauriès & VandenSpiegel, 2012, which is widespread in southern Thailand. Phylogenetic trees, both rooted and unrooted, based on a concatenated dataset of the COI and 28S genes of nine species of Cambalopsidae (Trachyjulus, Glyphiulus, and Plusioglyphiulus), strongly support the monophyly of Trachyjulus and a clear-cut divergence between T. magnus sp. nov. and T. unciger in revealing very high average p-distances of the COI gene (20.80–23.62%).
cave, diplopod, molecular-based phylogeny, morphological character, taxonomy
In South and Southeast Asia, as well as China, the juliform millipede family Cambalopsidae Cook, 1895 is among the most diverse, common, and often highly abundant groups that clearly dominate cave millipede faunas (
By far the largest genus is Glyphiulus Gervais, 1847 with its 60+ species ranging across China and Southeast Asia to Borneo in the east (
The genus Hypocambala Silvestri, 1895 is the smallest, but particularly widespread, presently containing 14 species in Southeast Asia, as well as scattered across several islands of the Pacific and Indian oceans (
The more diverse genus Trachyjulus Peters, 1864 is currently known to comprise 32 described species (
During recent field surveys in southern Thailand, a new, unusually large Trachyjulus species was taken from a cave. From the first glance, it seemed to be particularly similar to the grossly sympatric T. unciger, but both are readily distinguished by body size and several other characters, including gonopodal structures. To better understand the species delimitations and their variations, we compare this new species to topotypes of T. unciger (Pung-Chang (= Tham Nam) Cave, Phang-Nga Province, Thailand) not only based on their morphological characters, but also on molecular evidence. In addition, molecular-based phylogenetic relationships within the genus Trachyjulus are revealed and discussed for the first time using mitochondrial cytochrome c oxidase subunit I (COI) and nuclear gene 28S rRNA sequences. These were obtained from para- or topotypes of nine species of Cambalopsidae, including not only five Trachyjulus, but also two Glyphiulus and two Plusioglyphiulus as outgroups. Two members of the family Harpagophoridae from the same order Spirostreptida, as well as two of the family Julidae, order Julida, are also included as more distant outgroups for tree rooting.
Specimens were collected from southern Myanmar and southern Thailand under the Animal Care and Use Protocol Review No. 1723018. The collecting sites were located by GPS by using a Garmin GPSMAP 60 CSx, and all coordinates and elevations were rechecked with Google Earth. Photographs of live animals were taken using a Nikon 700D digital camera with a Nikon AF-S VR 105 mm macro lens. The specimens collected were euthanized by a two-step method following AVMA Guidelines for the Euthanasia of Animals (
The holotype, as well as most of the paratypes are housed in the Museum of Zoology, Chulalongkorn University (CUMZ), Bangkok, Thailand; a few paratypes have also been donated to the collections of the Zoological Museum, State University of Moscow, Russia (ZMUM) and the Natural History Museum of Denmark, University of Copenhagen, Denmark (NHMD), as indicated in the text.
The specimens were examined, measured, and photographed under a Nikon SMZ 745T trinocular stereo microscope equipped with a Canon EOS 5DS R digital SLR camera. Scanning electron micrographs (SEM) were taken with a JEOL, JSM-5410 LV microscope using gold-coated samples, and the material returned to alcohol upon examination. Digital images obtained were processed and edited with Adobe Photoshop CS5. Line drawings were executed based on photographs and specimens examined under a Nikon SMZ 745T trinocular stereo microscope, equipped with a Canon EOS 5DS R digital SLR camera. The terminology used and the carinotaxic formulae in the descriptions follow those in
Total genomic DNA was extracted from the dissected midbody ring tissues using the DNA extraction kit for animal tissue (NucleoSpin Tissue extraction kit, Macherey-Nagel, Germany), following the standard procedure of the manual. Fragments of the mitochondrial cytochrome c oxidase subunit I (COI, 690 bp) gene were amplified using LCO1490 (5'-GGTCAACAAATCATAAAGATATTGG-3';
The PCR amplification was performed using a T100™ thermal cycler (BIO-RAD) with a final reaction volume of 20 μL (15 μL of EmeraldAmp GT PCR Master Mix, 1.5 μL of each primer, 10 ng of template DNA and distilled water up to 20 μL total volume). Thermal cycling was performed at 94 °C for 3 min, followed by 35 cycles of 94 °C for 30 s, annealing at 42–56 °C (depending on samples and the primer paired) for 60 s, extension at 72 °C for 90 s, and a final extension at 72 °C for 5 min. Amplification of PCR products were confirmed through 1.5% (w/v) agarose gel electrophoresis before purification by PEG precipitation. Purified PCR products were sequenced in both directions (forward and reverse) using an automated sequencer (ABI prism 3730XL). All nucleotide sequences in this study were deposited in the GenBank Nucleotide sequences database under submission numbers MN893771–MN893781 for COI, and MN897820–MN897826 for 28S. The collecting localities and submission codes of each nominal species are listed in Table
Our phylogenetic analyses included a specimen (paratype) of T. magnus sp. nov. and six individuals of four previously described species, namely T. singularis (Attems, 1938), T. phylloides Golovatch, Geoffroy, Mauriès & VandenSpiegel, 2011, T. bifidus Likhitrakarn, Golovatch, Srisonchai, Brehier, Lin, Sutcharit & Panha, 2018, and T. unciger Golovatch, Geoffroy, Mauriès & VandenSpiegel, 2011. Specimens from other genera, i.e. Glyphiulus sattaa Golovatch, Geoffroy, Mauriès & VandenSpiegel, 2011, G. duangdee Golovatch, Geoffroy, Mauriès & VandenSpiegel, 2011, Plusioglyphiulus erawan Golovatch, Geoffroy, Mauriès & VandenSpiegel, 2011, and P. saksit Golovatch, Geoffroy, Mauriès & VandenSpiegel, 2011, were utilized as outgroups (Table
List of the species used for molecular phylogenetic analyses and their relevant information. * = paratype, ** = topotype.
Voucher number | Species | Locality | Geographical coordinates | GenBank accession numbers | |
---|---|---|---|---|---|
COI | 28S | ||||
CAM059* |
Trachyjulus bifidus |
Yae Gu Cave (River Cave), Tanintharyi, Myanmar | 11°13'4.50"N, 99°10'32.33"E | MN893771 | MN897820 |
CAM061* |
Trachyjulus bifidus |
Thin Bow Gu Cave (Linno Gu #2), Tanintharyi Region, Myanmar | 11°11'23.0"N, 99°10'18.3"E | MN893772 | MN897821 |
CAM027** | Trachyjulus phylloides Golovatch et al., 2012 | Phra Kayang Cave, Ranong, Thailand | 10°19'35.62"N, 98°45'53.54"E | MN893773 | N/A |
CAM079** | Trachyjulus unciger Golovatch et al., 2012 | Pung-Chang Cave, Phang-Nga, Thailand | 8°26'35.67"N, 98°30'57.32"E | MN893774 | MN897822 |
CAM070* | Trachyjulus magnus sp. nov. | Wat Tham Khrom Wanaram, Surat Thani, Thailand | 8°46'12.07"N, 99°22'6.36"E | MN893775 | MN897823 |
CAM044* | Trachyjulus singularis (Attems, 1938) | Tham Kao Havot Cave, Chon Buri, Thailand | 13°09'46.95"N, 101°35'51.97"E | MN893776 | MN897824 |
CAM107** | Trachyjulus singularis (Attems, 1938) | Khao Loi Cave (Wat Ma Duea), Rayong, Thailand | 13°03'27.00"N, 101°36'27.00"E | MN893777 | MN897825 |
Outgroup Cambalopsidae | |||||
CAM030* | Glyphiulus sattaa Golovatch et al., 2011 | Tham Pum-Tham Pla Cave, Chiang Rai, Thailand | 20°19'42.54"N, 99°51'50.12"E | MN893778 | N/A |
CAM022* | Glyphiulus duangdee Golovatch et al., 2011 | Chan Cave, Uttaradit, Thailand | 17°35'39.00"N, 100°25'18.30"E | MN893779 | MN897826 |
CAM031* | Plusioglyphiulus erawan Golovatch et al., 2011 | Erawan Cave, Lamphun, Thailand | 18°19'37.79"N, 98°52'22.41"E | MN893780 | N/A |
CAM021* | Plusioglyphiulus saksit Golovatch et al., 2011 | Tham Nennoi Cave, KhonKaen, Thailand |
16°43'4.77"N, 101°53'39.08"E | MN893781 | MN897826 |
Outgroup Harpagophoridae (Sporostreptida) | |||||
CUMZ-D00057 | Thyropygus bearti Pimvichai et al., 2009 | Si-Chon, Nakhon Si Thammarat, Thailand | 9°14'48.1"N 99°45'51.1"E | KC519519 | N/A |
CUMZ-D00021 | Thyropygus allevatus (Karsch, 1881) | Siam-Nakorn-Thani village, Nakhon Si Thammarat, Thailand | 8°25'23.9"N 99°58'07.0"E | KC519487 | N/A |
Outgroup Julidae (Julida) | |||||
BIOUG22537 | Julus scandinavius (Latzel, 1884) | Provincial Park, Ontario, Six Mile Lake, Canada: | 44°53'52.8"N 79°45'25.2"W | MG320199 | N/A |
09BBMYR_083 | Brachyiulus pusillus (Leach, 1815) | Gros Morne NP, Newfoundland and Labrador, Canada | 49°25'37.2"N 57°44'20.4"W | KM611731 | N/A |
The sequences were edited and aligned using Clustal W, implemented in MEGA7 (
Genus Trachyjulus Peters, 1864
Holotype ♂ (CUMZ), Thailand, Surat Thani Province, Ban Na San District, Wat Tham Khrom Wanaram, 8°46'12.07"N, 99°22'6.36"E, 16.06.2018, leg. W. Siriwut, E. Jeratthitikul and N. Likhitrakarn.
Paratypes. 15 ♂, 20 ♀ (CUMZ), 1 ♂, 1 ♀ (ZMUM), 1 ♂, 1 ♀ (NHMD), same locality, together with holotype.
To emphasize the largest body size of this species compared to all other species known in the genus.
This new species differs from all other Trachyjulus spp. by the largest body size (43.5–64.2 mm long, 2.1–2.8 mm wide), and also from the particularly similar and grossly sympatric T. unciger (23–42 mm long, 1.2–2.0 mm wide) in having the tegument of rings 2 and 3 nearly smooth (vs evidently carinate), carinotaxic formulae of typical rings (11–8/11–8+I/i+2/2+m/m vs 8–6/8–6+I/i+2/2+m/m), combined with the number of ommatidia (5–6+5–6 vs 4+4), and the posterior gonopods showing medial coxosternal processes (mcp) subtrapezoid (vs shorter and lobe-shaped).
Length of holotype ca 62.5 mm (Fig.
Coloration of live animals red-brown to yellow-brownish (Fig.
Body with 80p+2a+T rings (holotype); paratypes with 68–86p+1–3a+T (♂) or 69–93p+1–4a+T (♀) rings. Eyes large, flat, ovoid, with 6(5)+6(5) ommatidia arranged in a single vertical row (Fig.
In width, head = ring 2 < ring 4 = 5 < 3 < 6 < 7 < 8 < 9 < 10 < collum = midbody ring (close to 12th to 14th); body abruptly tapering towards telson on a few posteriormost rings (Fig.
Collum (Fig.
Epiproct (Fig.
Ventral flaps behind gonopod aperture on ♂ ring 7 barely distinguishable as low swellings, forming no marked transverse ridge.
Legs short, on midbody rings about 2/3 (♂, ♀) as long as body height (Figs
Trachyjulus magnus sp. nov., A–C, I–P ♀ paratype, D–H ♂ paratype. A, B anterior part of body, lateral and dorsal views, respectively C collum, dorsal view D cephalic capsule, dorsal view E gnathochilarium, ventral view F antenna, lateral view G tip of antenna H bacilliform sensilla on antennomere 5, lateral view I cross-section of midbody ring J midbody rings, ventral view K claw of midbody leg L enlarged ozopore region, lateral view M midbody prozona, dorsal view N–P posterior part of body, lateral, dorsal and ventral views, respectively.
♂ legs 1 highly characteristic (Figs
♂ legs 2 (Figs
♂ legs 3 (Figs
Anterior gonopods rather simple (Figs
Posterior gonopods (Figs
The often striking colour difference between head+collum+ring 2 and the remaining rings observed in SEM micrographs (Fig.
Trachyjulus magnus sp. nov., ♂ paratype. A, B Legs 1, frontal and caudal views, respectively C legs 2, caudal view D penes, caudal view E legs 3, frontal view F, G anterior gonopods, caudal and frontal views, respectively H telopodite tips of anterior gonopods I, J posterior gonopods, caudal and frontal views, respectively K, L telopodite tips of anterior gonopods, caudal and frontal views, respectively.
Trachyjulus magnus sp. nov., ♂ holotype. A Antenna, lateral view B gnathochilarium, ventral view C, D legs 1, caudal and frontal view, respectively E, F legs 2, caudal and frontal view, respectively G midbody leg, frontal view H legs 3, frontal view I, J anterior gonopods, frontal and caudal views, respectively K, L posterior gonopods, frontal and caudal views, respectively. Scale bars: 0.2 mm.
Our concatenated dataset contained 15 individuals, including seven Trachyjulus ingroup and eight outgroup species, and an alignment of approximately 1,501 base pairs (bp). We were unable to obtain sequences of the 28S gene from T. phylloides, G. sattaa, and P. erawan. The final alignment of the COI gene fragment yielded 690 bp (298 variable sites, 270 parsimony informative), while the 28S gene fragment comprised 811 bp (100 variable sites, 45 parsimony informative). The phylogenetic tree estimated by both ML and BI revealed equivalent topologies. As only one position within the outgroup taxa was controversial, solely a ML tree is shown in Figure
The interspecific divergence of the COI uncorrected p-distance among these nine cambalopsid species was found to be generally high (13.48–24.49%; Table
Phylogenetic analyses of Trachyjulus species and some related taxa. A Maximum likelihood tree based on a 1,501 bp alignment dataset of the nuclear 28S rRNA and mitochondrial COI genes. Numbers on nodes indicate bpp from Bayesian inference analysis (BI) and bootstrap values from maximum likelihood (ML), respectively B neighbour-joining tree (NJ) based on 230 amino acid alignments of peptide sequences corresponding to the mitochondrial COI dataset. Numbers on nodes indicate bootstrap values.
Matrix of the average interspecific genetic divergence (uncorrected p-distance: % ± SE) for the 690 bp barcoding region of the COI gene between Trachyjulus species and some related Cambalopsidae taxa.
Taxa | 1. | 2. | 3. | 4. | 5. | 6. | 7. | 8. |
---|---|---|---|---|---|---|---|---|
1. Trachyjulus bifidus | ||||||||
2. Trachyjulus phylloides | 15.07 ± 1.31 | |||||||
3. Trachyjulus unciger | 20.00 ± 1.51 | 19.13 ± 1.45 | ||||||
4. Trachyjulus magnus sp. nov. | 20.14 ± 1.52 | 20.00 ± 1.46 | 20.43 ± 1.52 | |||||
5. Tachyjulus singularis | 21.16 ± 1.53 | 20.80 ± 1.50 | 23.62 ± 1.64 | 21.52 ± 1.54 | ||||
6. Glyphiulus sattaa | 18.84 ± 1.50 | 17.68 ± 1.40 | 21.45 ± 1.58 | 18.12 ± 1.46 | 20.51 ± 1.53 | |||
7. Glyphiulus duangdee | 21.16 ± 1.51 | 21.16 ± 1.59 | 22.61 ± 1.56 | 23.62 ± 1.61 | 23.48 ± 1.59 | 17.97 ± 1.48 | ||
8. Plusioglyphiulus erawan | 21.74 ± 1.53 | 20.43 ± 1.46 | 24.49 ± 1.59 | 20.29 ± 1.48 | 21.45 ± 1.61 | 17.39 ± 1.40 | 19.42 ± 1.50 | |
9. Plusioglyphiulus saksit | 20.72 ± 1.48 | 21.30 ± 1.44 | 24.06 ± 1.53 | 20.58 ± 1.51 | 20.00 ± 1.41 | 19.13 ± 1.43 | 21.16 ± 1.47 | 13.48 ± 1.26 |
Trachyjulus magnus sp. nov. clearly represents a taxonomically valid species based on both morphological and molecular evidence. In the latest taxonomic review of Trachyjulus,
Morphologically, the new species looks especially similar to T. unciger, but both are clearly distinguishable (see Diagnosis above). Molecular evidence likewise reveals a sufficiently strong genetic divergence between T. magnus sp. nov. and T. unciger (p-distance = 20.43±1.52) (Table
The phylogenetic trees, both rooted and inrooted, and based on the concatenated dataset, provide strong support to the monophyly of the genus Trachyjulus in both ML and BI analyses (bpp = 1.0 for BI and bootstrap value = 100% for ML) (Fig.
Trachyjulus singularis was recovered as the basal clade of the tree. It also showed the highest genetic divergence from the other Trachyjulus species (21.16–23.62%). These results are in accordance with their geographic distributions, as T. singularis occurs only in eastern Thailand, i.e. far away from the congeners in southern Thailand. In addition, T. singularus has retained the ancestral character of a divided promentum of the gnathochilarium, a trait absent from the other members of Trachyjulus, but present in two other related genera, Glyphiulus and Plusioglyphiulus.
In conclusion, we put on record the first results of a molecular phylogenetic study on Trachyjulus, a largely cavernicolous genus, using a combination of the nuclear 28S rRNA and mitochondrial CO1 genes for a total of 1,501 bp. Our results reveal high rates of interspecific divergence among Trachyjulus species and other closely related genera. Given that Thailand and the neighbouring countries are extremely rich in karst and karst caves, there can hardly be any doubt that additional new species of Cambalopsidae generally and Trachyjulus in particular still await discovery. A combination of morphological and molecular studies in Cambalopsidae seems the best to provide further insights into the taxonomy and phylogenetic relationships in this large and widespread group.
This project was partly funded through grants received from the Office of the Royal Development Projects Board (RDPB), while most of the financial support was obtained from TRF Strategic Basic Research BDG 6080011 (2017–2019) to CS and NL, and Centre of Excellence on Biodiversity (BDC-PG2-161002) to SP. We thank the members of the Animal Systematics Research Unit for their invaluable assistance in the field. One of us (SIG) was partly supported by the Presidium of the Russian Academy of Sciences, Program No. 41 “Biodiversity of natural systems and biological resources of Russia”. Special thanks go to Henrik Enghoff (NHMD) for his critical review of an advanced draft that has allowed us to considerably improve the paper.