Biodiversity Data Journal :
Taxonomic paper
|
Corresponding author:
Academic editor: Pavel Stoev
Received: 23 Apr 2015 | Accepted: 05 Jun 2015 | Published: 09 Jun 2015
© 2015 Jeanne Wilbrandt, Paul Lee, Helen Read, Thomas Wesener
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:
Wilbrandt J, Lee P, Read H, Wesener T (2015) A first integrative study of the identity and origins of the British Dwarf Pill Millipede populations, Trachysphaeracf.lobata (Diplopoda, Glomerida, Glomeridae). Biodiversity Data Journal 3: e5176. https://doi.org/10.3897/BDJ.3.e5176
|
|
Three populations of the pill millipede genus Trachysphaera Heller 1858 are present in Great Britain, one on the Isle of Wight, one in South Wales and one in mid-Wales. To identify and characterize the British Trachysphaera populations, the intraspecific and interspecific variation of the populations in South Wales and on the Isle of Wight were studied and evaluated in a first integrative study of members of Trachysphaera, utilizing barcoding and SEM. DNA was extracted from 28 British Trachysphaera and 10 French T. lobata (Ribaut 1954) specimens, one each of French T. cf. drescoi (Conde and Demange 1961) and T. pyrenaica (Ribaut 1908), and one of Spanish T. cf. rousseti (Demange 1959); the barcoding fragment of the COI gene was amplified and their genetic intra- and interpopulation distances compared with one another using two Italian T. spp. and one Croatian T. schmidti Heller 1858 specimens as near outgroups. To compare the genetic distances with the morphological characters, 15 characters of a total of 13 British Trachysphaera, together with two specimens of T. pyrenaica, two T. cf. drescoi and one of T. cf. rousseti were imaged, using the same individuals utilized for DNA extraction. Albeit both British populations are genetically distant, they are closely related (1.9–2.5% p-distance) to French T. lobata, corroborating results of earlier studies. Between different Trachysphaera species, genetic distance was high (16.7–18.8%). The morphological study showed the non-reliability of key taxonomic characters in Trachysphaera, with genetically identical individuals exhibiting morphological variation, especially on the telopods. The only observed morphological characters constant within and different between species were the number of rows of sclerotized bacilli on the tergites, as well as the shape of the male and female anal shield. Both, barcoding and the morphological study identify the British Trachysphaera populations as T. lobata.
Integrative study, barcoding, taxonomic characters, SEM, Glomerida
Some of the most ornamented members of the pill millipedes (order Glomerida) are the dwarf species of the genus Trachysphaera
The position of Trachysphaera vis-à-vis other Glomerida is uncertain, with several authors putting the genus into a family of its own, the Trachysphaeridae. This placement is based on the fusion of the last tergite with the anal shield, reduction of defense glands, an unusual technique of rolling up, circular grooves on the thoracic shield, and other external peculiarities (
Even more problematic than the position of Trachysphaera within Glomerida is the alpha-taxonomy of the genus. Species were first described based on their peculiar external morphology alone (
Finding and collecting Trachysphaera specimens constitute further difficulties, since species are small, cryptic, and rare. Their dwarf habitus makes them difficult to find for a generalist researcher, as specimens are <5 mm long when walking and form a sphere with a diameter of 2–3 mm when rolled up. Discovery is also hampered by their crypsis: specimens do not move for up to 1h when disturbed (T. Wesener, personal observation), and resemble a tiny pebble of calcareous stone. Additionally, while some Austrian species are now known from a relatively wide area (
Adding to these morphological peculiarities, sequencing of Trachysphaera specimens seems to be difficult as well. Of the 10 specimens analyzed during the 'Fauna Bavarica' project and sequenced at the Barcode of Life facility in Guelph, none yielded sequence data (
In 1984, an isolated population of Trachysphaera was discovered at Bembridge on the Isle of Wight, England. The species was tentatively determined as T. lobata (
For a comprehensive analysis and evaluation of the British Trachysphaera populations we chose an integrative (molecular and morphological) approach, applied for the first time in pill millipedes. The analysis of morphology and sequence data from the same specimen allows a direct comparison of morphological and genetic variation. With this approach we aimed to answer the question of the origin of the two British Trachysphaera populations, and to discover whether the two disjunct populations resembled relics or more recent (potentially anthropogenic) introductions. We also evaluated the intrapopulational, intraspecific and interspecific differences of 15 sexual and non-sexual morphological characters apart from the telopods in order to advance future taxonomic classifications of the more than 30 Trachysphaera species.
To investigate the morphological and molecular diversity of the British Trachysphaera, 28 specimens from Wales and the Isle of Wight were selected. From the mainland population of T. lobata, 10 specimens from a French population (Génis) were used. Additional specimens were included from the northern French-Spanish border region from three sites: (1) one male and female of T. pyrenaica from Grotte de l'Estellas, France, (2) one female of T. cf. rousseti from Leitza, Spain, (3) one male and female of T. cf. drescoi from Sare, France.
As near outgroups, two female specimens of undetermined Italian Trachysphaera, as well as a specimen of the eastern central European species and type of the genus, T. schmidti
All animals were collected by hand. Two slightly different preservation methods were utilized: (1) Immediately after capture, specimens were put into 1.8 ml screw top vials with 98% ethanol; British Trachysphaera were stored in individual tubes, while all French T. lobata and all Italian Trachysphaera specimens were each put together into a single vial. (2) The specimens of T. pyrenaica, T. cf. rousseti and T. cf. drescoi were captured and stored in 80% ethanol, before being transferred to 98% ethanol >12 months later.
Before the disintegration of the specimens for the removal of tissue and SEM preparations, colored multi-layer photographs of the enrolled specimens were taken under a Leica Z6 Imaging-System. A total of 11 specimens from the Isle of Wight population, and 8 from the Welsh population were photographed. For optimal depth of field, the 10-15 single photographs taken from each specimen were put together into one multi-layer photograph using the software Auto-Montage.
The tiny size of the specimens made DNA extraction difficult. In at least one case, the animal was filled with a long and massive nematomorph >4x the length of the host specimen. Due to the possible presence of parasites and the fact that no sequences of the genus (or family) were present on NCBI GenBank for comparison, no whole body extraction was utilized, but every specimen was carefully dissected. Intersegmental muscle tissue was chosen as the extraction target. Due to the small size of the specimens, the specimen was pulled apart along the tergite margins and divided into several parts, where the muscle tissue binding the tergites could be removed with fine forceps. Dissected muscle tissue was washed in a dish of ethanol to remove attached particles of the intestine, which are usually filled with cephaline gregarines.
The muscle tissue was processed with a DNAeasy Blood & Tissue kit from Qiagen following the manufacturer’s extraction protocol, except that two times 50 µl elution buffer were used to heighten the DNA yield of the extraction. DNA was extracted from a total of 44 specimens: 13 specimens of T. cf. lobata from the Isle of Wight and 15 from Wales; 10 specimens of T. lobata from Génis, France; one specimen of T. pyrenaica from the Grotte de L'Estellas; one specimen of T. cf. rousseti from Leitza, Spain; one specimen of T. cf. drescoi from Sare, France; one specimen of T. schmidti from Croatia and two specimens of the Italian Trachysphaera. The dissected specimens were also used for the SEM study (see below), while the remaining parts are conserved as voucher specimens at the ZFMK. Genomic DNA is archived in Qiagen extraction buffer and stored at -20°C at the ZFMK.
To gain insight into the genetic diversity of the British Trachysphaera populations as well as their distance to the French and Spanish taxa, the standard barcoding fragment of the cytochrome c oxidase subunit I (COI), a mitochondrial gene, was chosen as a marker. The COI gene was amplified using polymerase chain reaction (PCR) (
Purified PCR products from 34 specimens were outsourced for double-strand sequencing to a contract sequencing facility (Macrogen, Seoul, Korea) on an ABI3730 XL automatic DNA sequencer, using the same primer sets as for PCR. Sequences of a total of 27 specimens (13x Isle of Wight, 12x Wales, 2x Italy) could be obtained, while the sequences for the French T. lobata only contained various contaminations. Overall, PCR and sequencing success was limited (Table
Localities and method application. Locality ID [LocID] as given in Checklist Materials, Extracted specimens [# Extracted], PCR and sequencing success [# PCR success, # Sequencing success] and number of Trachysphaera specimens studied under SEM [# SEM]. Numbers in parentheses refer to specimens where PCR and/or sequencing were unsuccessful at the ZFMK but successful at the BGI. More detailed locality info in Checklist section.
LocID |
Locality |
# Extracted |
# PCR success |
# Sequencing success |
# SEM |
1 |
Isle of Wight, GB |
13 |
13 |
13 |
5 |
2 |
South Wales, GB |
15 |
15 |
12 |
8 |
3 |
Génis, FR |
10 |
4 (2) |
- (2) |
- |
4 |
Grotte de l'Estellas, FR |
1 |
- |
- |
2 |
5 |
Leitza, ESP |
1 |
- |
- |
1 |
6 |
Sare, FR |
1 |
- |
- |
2 |
7 |
Oropa, IT |
2 |
2 |
2 |
- |
8 |
Velika-Kapela, CRO |
1 |
- (1) |
- (1) |
- |
Sequencing reads were assembled with Bioedit 7.1.3. (
The analysis involved 32 nucleotide sequences (COI from 32 specimens), with a total of 660 positions in the final dataset. The number of base substitutions per site from between sequences were determined using MEGA (v. 5.2,
The tree with the highest log likelihood (-2177.4216) is shown below (Analysis section). The percentage of trees in which the associated taxa clustered together is shown next to the branches (bootstrap). The tree is drawn to scale, with branch lengths reflecting the number of substitutions per site.
To evaluate the intra- and interspecific variation of closely related Trachysphaera species, a total of 19 specimens (3 males, 2 females from Isle of Wight; 1 male, 7 females from Wales; 1 male, 3 females of T. pyrenaica; 1 male, 1 female T. cf. drescoi) of Trachysphaera, all of which were also subject to DNA extraction, were studied using scanning electron microscopy. Objects prepared for SEM were: (1) anal shield; (2) anterior body part including head, collum, thoracic shield and tergite 3; (3) midbody tergite (for body part and tergite nomenclature see Fig.
Body part and tergite nomenclature. SEM micrographs of a female of Trachysphaera pyrenaica (GEU157, MBiIDs 852870 and 852871). Abbreviations: Ant = antenna; AS = anal shield; Co = collum (tergite 1); h = head; th-sh = thoracic shield (tergite 2); T# = tergite number.
A total of 15 (11, 4 of which separately given for male and female) morphological characters commonly employed in the taxonomy of Trachysphaera were investigated for their taxonomic value (see Table
Characters and states. Character numbers [C #], states are exemplified with SEM images partly in figures, partly with links to MorphBank, then given as MBiIDs.
C # |
Character |
States |
1 |
Color of freshly preserved specimen. |
(0) brownish (Fig. |
2 |
Collum (reduced tergite 1), number of toothed ridges (Fig. |
|
3 |
Thoracic shield (enlarged tergite 2) anterior margin, number of rows of sclerotized nodules (Fig. |
Real number |
4 |
Thoracic shield, number of rows of large sclerotized protuberances (Fig. |
Real number |
5 |
Endotergum (underside of posterior margin of tergites), structure. |
(0) simple margin with one row of sclerotized nodules, and single row of short setae (Fig. |
6 |
Endotergum, number of rows of setae. |
(0) 1 (Fig. |
7 |
Endotergum, number of rows of sclerotized nodules. |
(0) 1 (Fig. |
8 |
Male tergite 10, posterior margin, number of rows of large bacilli. |
|
9 |
Female tergite 10, posterior margin, number of rows of large bacilli (Fig. |
|
10 |
Male anal shield, shape. |
(0) well-rounded (852928), (1) with special protuberance (852829) |
11 |
Female anal shield, shape (Fig. |
0) well-rounded (852960), (1) with special protuberance (852856) |
12 |
Male anal shield, setae at posterior margin. |
(0) isolated, 1-2 rows, (852978), (1) many, >2 rows (852826) |
13 |
Female anal shield, setae at posterior margin (Fig. |
(0) isolated, 1-2 rows, (853011), (1) many, >2 rows (853027) |
14 |
Male anal shield, large circular grooves. |
|
15 |
Female anal shield, large circular grooves (Fig. |
Character matrix. Abbreviations: [N] refers to number of individuals investigated by SEM; IoW = Isle of Wight population; U = character conflict within one population, i.e., unresolved.
Species/Population |
1 |
2 |
3 |
4 |
5 |
6 |
7 |
8 |
9 |
10 |
11 |
12 |
13 |
14 |
15 |
T. pyrenaica [N=2] |
U |
0 |
0 |
1 |
0 |
0 |
0 |
0 |
0 |
1 |
0 |
0 |
0 |
0 |
0 |
T. cf. rousseti [N=1] |
1 |
0 |
0 |
1 |
0 |
0 |
0 |
n/a |
0 |
n/a |
1 |
n/a |
0 |
n/a |
1 |
T. cf. drescoi [N=2] |
U |
U |
0 |
U |
0 |
0 |
0 |
0 |
0 |
1 |
1 |
1 |
1 |
1 |
1 |
T. lobata IoW [N=5] |
U |
0 |
0 |
U |
0 |
0 |
0 |
1 |
1 |
0 |
0 |
0 |
U |
U |
U |
T. lobata Wales [N=8] |
U |
U |
0 |
U |
0 |
0 |
0 |
1 |
1 |
0 |
0 |
U |
U |
U |
U |
Addtional individual information in Table
Additional sample information for Trachysphaera lobata specimens. Material specimen number as used above, Extraction voucher number, MorphBank specimen ID (with the number of SEM images available in parentheses), NCBI GenBank accession number.
Material specimen # |
Extraction voucher # |
MBspecimenID |
GenBank accession # |
a |
TW01 |
852812 (15) |
|
b |
TW02 |
– |
|
c |
TW03 |
– |
|
d |
TW04 |
– |
|
e |
TW05 |
– |
|
f |
TW11 |
– |
|
g |
TW12 |
852813 (17) |
|
h |
TW13 |
852814 (11) |
|
i |
TW14 |
852815 (23) |
|
j |
TW15 |
– |
|
k |
TW16 |
– |
|
l |
TW17 |
852816 (17) |
|
m |
TW18 |
– |
|
n |
TW19 |
– |
|
o |
TW20 |
– |
|
p |
TW21 |
852817 (16) |
– |
q |
TW22 |
852818 (16) |
|
r |
TW23 |
– |
|
s |
TW24 |
852819 (21) |
|
t |
TW25 |
852820 (18) |
|
u |
TW26 |
852821 (12) |
|
v |
TW27 |
852822 (13) |
– |
w |
TW28 |
– |
|
x |
TW29 |
852823 (28) |
|
y |
TW30 |
852824 (17) |
|
z |
TW31 |
– |
|
aa |
TW32 |
– |
|
ab |
TW33 |
– |
– |
ac |
TW47 |
– |
– |
ad |
TW48 |
– |
– |
ae |
TW49 |
– |
– |
af |
TW50 |
– |
– |
ag |
TW51 |
– |
|
ah |
TW52 |
– |
|
ai |
TW53 |
– |
– |
aj |
TW54 |
– |
– |
ak |
TW55 |
– |
– |
al |
TW57 |
– |
– |
Additional individual information in Table
Additional sample information for Trachysphaera pyrenaica specimens. Material specimen number as used above, Extraction voucher number, MorphBank specimen ID (with the number of SEM images available in parentheses), NCBI GenBank accession number.
Additional individual information in Table
Additional sample information for Trachysphaera cf. rousseti specimen. Material specimen number as used above, Extraction voucher number, MorphBank specimen ID (with the number of SEM images available in parentheses), NCBI GenBank accession number.
Material specimen # |
Extraction voucher # |
MBspecimenID |
GenBank accession # |
a |
GL017 |
852825 (16) |
– |
Additional individual information in Table
Additional sample information for Trachysphaera cf. drescoi specimens. Material specimen number as used above, Extraction voucher number, MorphBank specimen ID (with the number of SEM images available in parentheses), NCBI GenBank accession number.
Additional individual information in Table
Additional sample information for Trachysphaera schmidti specimen. Material specimen number as used above, Extraction voucher number, MorphBank specimen ID (with the number of SEM images available in parentheses), NCBI GenBank accession number.
Material specimen # |
Extraction voucher # |
MBspecimenID |
GenBank accession # |
a |
BGI-MYR-16 |
– |
Additional individual information in Table
Additional sample information for Trachysphaera sp. specimens. Material specimen number as used above, Extraction voucher number, MorphBank specimen ID (with the number of SEM images available in parentheses), NCBI GenBank accession number.
Material specimen # |
Extraction voucher # |
MBspecimenID |
GenBank accession # |
a |
TW06 |
– |
– |
b |
TW07 |
– |
Despite differences in the morphology and coloration (see below), no variation in the mitochondrial haplotypes were found within Trachysphaera populations. Genetic distances (uncorr. p-dist.) in the COI sequence between the UK Trachysphaera populations and between T. lobata from France were low (1.9–2.5%). The Isle of Wight Trachysphaera show a 2.5% divergence to the French population and a 1.9% genetic distance to the population from Wales. The Wales population shows an equal genetic distance of 1.9% to both the French and the Isle of Wight Trachysphaera.
All T. lobata populations show high genetic distances of 17.3–18.8% to the other Trachysphaera species from Italy and Croatia. The latter two, T. schmidti from Croatia and T. sp. from Italy, differ by 16.7%.
The close relationship of the UK and French Trachysphaera populations is also reflected in the phylogenetic tree (Fig.
Maximum Likelihood Tree of 32 COI sequences of Trachysphaera and outgroup specimens. Rate model: discrete Gamma distribution, 5 categories (+G, parameter = 0.7759); nucleotide frequencies: A / T = 0.3273, C / G = 0.1727; tree is drawn to scale, branch lengths reflect number of substitutions per site; 1st, 2nd, 3rd codon positions and non-coding positions were included, gaps and missing data excluded; node values = 1000 replicates ML bootstrap support.
Most of the studied somatic characters show considerable variation (see Table
Habitus. Preserved Trachysphaera specimens from two locations, multi-layer photographs, showing high variation in colour and/or encrusted dirt. Not to scale; for further specimen information see Checklists.
Trachysphaera SEM characters. A+B: MBsID 852812, C-E: MBsID 852823. Abbreviations: AS = anal shield; Co = collum; Gr = groove, 'Ohrgrube'; h = head; T# = refers to number of tergite; th-sh = thoracic shield; c2 = collum, number of toothed ridges; c3 = thoracic shield anterior margin, number of rows of sclerotized nodules; c4 = thoracic shield, number of rows of large sclerotized protuberances; c5 = endotergum structure; c6 = endotergum, number of rows of setae; c7 = endotergum, number of rows of sclerotized nodules; c9 = female tergite 10, posterior margin, number of rows of large bacilli (with row counts 1, 2); c11 = female anal shield, shape; c13 = female anal shield, setae at posterior margin; c15 = female anal shield, large circular grooves (for more information on characters see Table
The only unambiguous morphological characters that neither show great variation within a species, nor are constant between closely related species, are the characters 8–12 (Table
The telopods of the studied Trachysphaera species show extensive variation between species, within species, within populations and even within the same individual (Figs
Telopods. SEM micrographs of Trachysphaera lobata males from Isle of Wight. For TW# refer to Checklists. Abbreviations: fem = femur; pre = prefemur; syn = syncoxite; ta = tarsus; ti = tibia.
Telopods. SEM micrographs of Trachysphaera lobata male from South Wales. For TW# refer to Checklists. Abbreviations: fem = femur; pre = prefemur; syn = syncoxite; ta = tarsus; ti = tibia.
Telopods. SEM micrographs of Trachysphaera cf. drescoi and T. pyrenaica males from France. For TW# refer to Checklists. Abbreviations: fem = femur; pre = prefemur; syn = syncoxite; ta = tarsus; ti = tibia; 18 = leg (pair) 18.
The UK Trachysphaera populations belong, based on their mitochondrial DNA (COI sequence), clearly to T. lobata, an observation also corroborated by the morphological data (see below). Both populations analyzed (from the Isle of Wight and South Wales) show unique haplotypes.
Success of DNA extraction and PCR was generally low, indicating the necessity of special treatments to ensure success. Some specimens rolled-up so tight that ethanol was not able to penetrate the specimen, and muscle tissue was already partly decayed. We thus recommend to conserve specimens individually and open them shortly after conservation (i.e., a few minutes after death at best) so that ethanole can enter the organism. Furthermore, it is reasonable to not extract sequences from whole specimens but only from muscular and other tissues free of contamination from gut content or parasites. PCR success was improved (i.e., obtaining positive bands at all) with another deviation from standard protocols, namely using 5µl of DNA instead of the usual 1-2 µl. We were not able to study the reason for this necessity and can thus only speculate. Possibilities are insufficient primer matching, decay, or other, inherent peculiarities.
The differing results of sequencing attempts at the ZFMK and BGI despite of presumably standardized methods highlight the persistent difficulties of obtaining glomerid DNA sequences. Here, we deem a further investigation of best practice methods for Glomerida extremely important.
For studies of the intraspecific variation of Trachysphaera species, faster evolving markers that show a greater amount of variation should provide a better resolution than the COI gene fragment studied here. Unfortunately, such a marker is currently not available in Diplopoda.
Between Trachysphaera species, the genetic distance is quite high, on par with those observed between different genera of the order (see
Telopods need to be studied and illustrated with great care. Slight variations of the angle of view can provide very different observations (compare Figs
Since trachysphaeran surface structures are tiny, complex, and manifold, their evaluation as well as drawing conclusions is obfuscated. Many characters are too fine-scaled or complex to be evaluated in a discrete fashion and it is hard to define categories due to variability in quasi-countable characters. The possible inner variation of extended structures (e.g., endotergum) confounds decisions of species identification based on small details of the structure. However, even small details vary in their peculiarities: large sclerotized protuberances (LSPs; Fig.
Taxonomically informative character examples. Abbreviations: AS = anal shield; T10 = tergite 10; c8 = male tergite 10, posterior margin, number of rows of large bacilli; c12 = male anal shield, shape. For more information on characters and states see Table
Problems did not only arise from the structural diversity itself but also from the method of documentation. Too few pictures were taken and not all angles were available due to fixation. Comparing all specimens, not all structures were shot from the same angle. A better chance to find true species-specific characters and evaluate them properly would be given by zoomable 3D images, as can be obtained by micro-CT studies. Additionally, physical influences may have distorted the evaluation. LSPs and similar extended structures (e.g., setae) especially may have been evaluated inconsistently due to prior abrasion (during struggle in life, dissection, mounting and sputtering). Grooves may have been evaluated incorrectly if clogged by dirt. Here, testing whether ultra-sonic cleaning is feasible would be beneficial (we did not try due to the risk of loosing of one specimen). Given these complications and our relatively small sample size, further corroboration of our results is desirable. Meanwhile, our conclusions have to be considered as preliminary, but nevertheless relevant.
Both, barcoding and the morphological study identify the British Trachysphaera populations as T. lobata. We were not able to trace the origin of these British populations because only one French mainland population could be sampled due to external circumstances. However, an independent origin of the South Wales and the Isle of Wight populations could be ascertained. Whether this independent origin is the result of two different recent anthropogenic introduction events or if they represent two relic populations of a once widespread occurrence currently cannot be determined. Once the COI sequence of more French populations of T. lobata becomes known, a clearer picture of the origins of the British Trachysphaera can be drawn.
We thank Desmond Kime for his advice and the laborious collecting efforts of French Trachysphaera specimens without which this study would have been impossible, Ian Morgan for his collection of Welsh specimens, and Steve Gregory for assistance with collecting on the Isle of Wight. Courtesy for the images used in Figure 1 go to Paul Richards. Molecular lab manager Claudia Etzbauer (ZFMK) provided help and advice during the tricky gathering of molecular information, while Karin Ulmen (ZFMK) advised at the SEM. Dr. Hans Reip (Jena) thankfully supplied specimens of T. schmidti from Croatia and supported us on numerous occasions. We also thank the support team of MorphBank for their patience and help, as well as our reviewers Peter Decker, Slobodan Marakov, and Jan Philip Oeyen for constructive suggestions. Collection of specimens from the Isle of Wight was undertaken as part of biodiversity conservation work co-ordinated by Hymettus and funded by the UK Government Department for the Environment, Food and Rural Affairs (Project code WC0786). Much of the data was assembled during a class of JW undertaken in the international study program M.Sc. OEP-Biology at the University of Bonn. Funding for chemicals and sequencing was provided by the ZFMK.