Organisms Diversity & Evolution

, Volume 11, Issue 1, pp 43–59

Sympatry in Mantophasmatodea, with the description of a new species and phylogenetic considerations

Authors

    • Behavioural Physiology, Department of BiologyHumboldt-Universität zu Berlin
  • Mike D. Picker
    • Zoology DepartmentUniversity of Cape Town
  • Klaus-Dieter Klass
    • Museum für TierkundeSenckenberg Naturhistorische Sammlungen Dresden
Original Article

DOI: 10.1007/s13127-010-0037-8

Cite this article as:
Eberhard, M.J.B., Picker, M.D. & Klass, K. Org Divers Evol (2011) 11: 43. doi:10.1007/s13127-010-0037-8

Abstract

We describe a new genus of Mantophasmatodea, Viridiphasma gen. n. (Austrophasmatidae), represented by one new species, V. clanwilliamense sp. n. The new species differs from previously described species in features of the male and female postabdomen including the genitalia, in morphometrics and details of colouration. The new species occurs syntopically with another austrophasmatid, Karoophasma biedouwense Klass et al., 2003; this is the first well-documented case of sympatry of two mantophasmatodean species. We therefore survey the morphological differences between these two species, document the absence of any morphological evidence of hybridisation, and also report on differences in life history. While a previous molecular phylogeny using COI and 16S genes ambiguously placed V. clanwilliamense sp. n. near the base of Austrophasmatidae (but not as sister to all other Austrophasmatidae), morphological characters strongly support V. clanwilliamense sp. n. to be the sister taxon of a clade comprising all remaining Austrophasmatidae. This phylogenetic placement challenges the current hypothesis of a linear north-to-south diversification of Austrophasmatidae.

Keywords

AustrophasmatidaeFynbosMorphologySympatrySyntopyPhylogeny

Abbreviations

Measurements

antl

antenna length

bodl

body length

emad

distance between eye and mandibular articulation

eyeh

eye height

eyel

eye length

genh

gena height

heaw

head width

iocw

interocular width

msnl

mesonotum length

mtfl

metafemur length

mtnl

metanotum length

mttl

metatibia length

prfl

profemur length

prnl

pronotum length

prnw

pronotum width

prtl

protibia length

t10l

minimum length of tergum X

t10l+

maximum length of tergum X

t10w

width of tergum X

thol

thorax length

Parts of postabdomen

at8

anterior apodeme/tendon of abdominal laterocoxosternum VIII

dpc

dorsal projection of cercus

gp8

gonapophysis VIII

gp9

gonapophysis IX

GP9

sclerotization of gonapophysis IX

IP

sclerite of posterior intervalvula

LS8

abdominal laterocoxosternum VIII (subgenital plate)

PH1–8

phallic sclerites 1–8

phl

phallic hook, long

phs

phallic hook, short

pli

phallic lobe, internal

spb

spermathecal bulb

spt

spermathecal tube

va

vagina

Natural history museums

MTD

Museum für Tierkunde, Dresden

SAMC

South African Museum, Cape Town

Introduction

The initial description of Mantophasmatodea (heelwalkers) (Klass et al. 2002) has stimulated extensive research on the new insect order, including taxonomy and intra-ordinal phylogeny (e.g. Damgaard et al. 2008; Klass 2004; Klass et al. 2003a, Klass et al. 2003b; Zompro and Adis 2006; Zompro et al. 2002, 2003), morphology (Baum et al. 2007; Beutel and Gorb 2006; Dallai et al. 2003; Drilling and Klass 2010; Eberhard et al. 2009; Hockman et al. 2009; Klass and Eulitz 2007; Klug and Klass 2007), embryonic development (Machida et al. 2004), reproductive biology (Tojo et al. 2004), life cycle (Hockman et al. 2009), egg structure (Tsutsumi et al. 2004), fossil record (Arillo and Engel 2006; Engel and Grimaldi 2004; Huang et al. 2008; Zompro 2001, 2005), vibrational communication (Eberhard and Picker 2008; Eberhard et al. 2010), and inter-ordinal phylogenetic position (Beutel and Gorb 2006; Cameron et al. 2006; Klass 2004, 2009; Klass et al. 2003b; Terry and Whiting 2005). Extensive collecting has increased the number of known extant species from two (Klass et al. 2002) to more than 20 in the most recent and comprehensive phylogenetic study (Damgaard et al. 2008). In addition, five fossil species have been assigned to Mantophasmatodea, the oldest ones from the mid-Jurassic (Huang et al. 2008). On the other hand, current knowledge on the ecology, distribution, behaviour, and overall biology of mantophasmatodeans has remained quite incomplete.

Morphological characters by which extant species can be distinguished are focally located in the postabdomen (Klass et al. 2003b). In the male, diagnostic characters include the presence and shape of particular genitalic sclerites, structural details of the vomeroid (a sclerotized process above the genitalia) and cerci (used as clasping organs during copulation), and the shape of abdominal tergum 10. In the female, the shape and sclerotization of the subgenital plate (abdominal ‘sternite’ 8), the shape of the spermathecal bulb, and the condition of vaginal sclerites are most important. While the overall body colouration shows much intra-specific variation and is reflected in polymorphism, some aspects of colouration do provide useful characters. This concerns, for instance, the pattern of dark epidermal patches (close-correlation versus loose-correlation type, referring to the spatial correlation between setal bases and pigmented areas), and the distribution of compound eye pigments (striped versus mosaic type) (Klass et al. 2003b).

The morphology-based results on the delimitation of mantophasmatodean species match the species boundaries suggested by molecular studies. In Damgaard et al.’s (2008) analysis of mitochondrial 16S and COI, in which most species were represented by several specimens usually from different localities, all morphologically defined species were supported. Some of these taxonomic units have been confirmed also by (so far sporadic) comparative analysis of selected neuropeptides (Predel et al. 2004). The combination of all this information has provided a rigid and consistent taxonomic framework for Mantophasmatodea (Damgaard et al. 2008 provide an up-to-date classification; see also Fig. 7). At least in South Africa mantophasmatodean species appear to be narrow-range endemics. Typically the pattern is strictly allo- or parapatric, with the majority of species having a limited distribution range; only two additional cases of sympatry have been recorded so far, but these have not been investigated in any detail and might even be the result of confusion of samples.

Some of the species-level taxa of Mantophasmatodea established as monophyletic in Damgaard et al. (2008) were not described formally and therefore given preliminary names only. Here we describe one of these species (treated as “Austrophasmatidae sp.n. 3” in Damgaard et al. 2008), give arguments for its basal phylogenetic position within Austrophasmatidae, and consequently erect a new genus for it. This species was also the subject of a study of vibrational communication (part of the mate recognition system in Mantophasmatodea) by Eberhard and Picker (2008), who found that characteristics of the male and female calls in this species differed significantly from those in a sympatric austrophasmatid, Karoophasma biedouwense Klass et al. 2003, at Clanwilliam Dam, Western Cape Province, South Africa.

Previously, K. biedouwense had been known only from a small area in the Western Cape region that included the settlements of Biedouw, Driefontein (both in the Succulent Karoo biome), and Wolfdrif (in the fynbos biome) (Damgaard et al. 2008: fig. 1; Klass et al. 2003b). The new finding at Clanwilliam added an outlying record for this species, only some 50 km southwest of the other localities but separated from these by the Cederberg Mountains. Since the various populations occur across two different biomes, divergence between populations of these wingless insects might be expected. As no genetic evidence is available yet, the Clanwilliam population of K. biedouwense was re-examined morphologically. Additionally, this allows an assessment whether any introgression has taken place between the two sympatric species.

Material and methods

Specimens

The specimens used for this study were collected in May–September 2006 and August–September 2007 near the town of Clanwilliam (South Africa, Western Cape Province, 32.21°S 18.88°E) using the bush beating technique. The site falls within the winter rainfall fynbos biome, and supports Leipoldtville Sand Fynbos vegetation (FFd 2 of Rebelo et al. 2006). This open, semi-arid shrubland community includes a variety of often spiny shrubs, succulents and small trees. While K. biedouwense occurred on a wide range of plant species, it was largely restricted to shrubs of various sizes on which its colouration matched the mass of bare stems and branches (most of the fynbos plants have microphyllous leaves). In contrast, by far the majority of specimens of the new species described below were collected from small trees at the site on which their bright green colouration afforded some degree of camouflage from avian and reptilian predators. Males and females of both species were collected as nymphs or adults, and taken back to the laboratory for rearing. Individual nymphs were kept in separate plastic pots (height 50–90 mm, diameter 55–95 mm) and fed on a ‘vestigial’ Drosophila species. Water was supplied by saturated wads made from paper towels. Five adult males and females of each of the two sympatric taxa were anaesthetized by chilling and fixed in 96% ethanol. All individuals were given a unique ID comprising the collection site (Cw), running number (3–12; separate for each sex; numbers 1 and 2 preoccupied by specimens studied in Damgaard et al. 2008), and sex (♂ or ♀); e.g. Cw3♀.

Morphology and measurements

Colouration was examined, and most measurements were taken under a Leica M125 stereo microscope using the unmacerated preserved specimens, often based on photographs taken from different angles. The same distances were measured and ratios calculated as described in Klass et al. (2003b: p. 17–18, fig. 3). To visualize interspecific differences in morphometric data, a principal component analysis was performed using PASW Statistics 18 (version 18.0.0); the significance of the first two principal components was tested (Mann-Whitney U test).

For the study of postabdominal exoskeletal structures, segments VII and following (females) or IX and following (males) were cut off and macerated in 10% KOH for about 1–2 hours. Tracheae were cut near the spiracles and removed. The exoskeleton was then examined in 70% ethanol under the Leica M125. The terminology for sclerites and other structural elements, including abbreviations, follows Klass et al. (2003b).

Photographs were taken through the Leica M125 with a Leica DFC420 digital camera using the software Leica Application Suite (version 3.4.1; Leica Microsystems CMS GmbH, Switzerland); they were subsequently edited with Adobe Photoshop CS2 (version 9.0.2). Drawings were produced using CorelDraw and Corel Photo-Paint (version 12.0.0.536; Corel Corporation).

Taxonomic results and descriptions

Viridiphasma gen. n.

Etymology

The name is composed from the Latin adjective ‘viridis’ (= green), reflecting the bright green live colouration in the single known species of this genus, and the Classical Greek noun ‘to phasma’ (= apparition, phantom), a frequent component of names for stick insect and mantophasmatodean genera, e.g. Mantophasma Zompro, Klass, Kristensen & Adis, 2002. As in other such composites ending in ‘-phasma’, the nomenclatural gender of the genus name is neuter (see also Damgaard et al. 2008; Zompro 2005).

Type species

Viridiphasma clanwilliamense sp. n.

Other included species

None thus far.

Description and diagnosis

Viridiphasma gen. n. is assigned to Austrophasmatidae based on its morphological characters and the molecular-based phylogeny of Damgaard et al. (2008). It conforms with the diagnosis of Austrophasmatidae in Klass et al. (2003b: p. 27), but there are four exceptions: (i) in males, phallic sclerite PH2 bears, besides a ridge, also a process; (ii) a sclerite extension at the ventral base of the cercus is absent; (iii) in females, there are a few setae on the membrane between the gonapophyses of segment 9 (gp9); (iv) the height of the eye and the height of the gena are subequal. All these features are so far unique within Austrophasmatidae.

In both sexes Viridiphasma gen. n. differs from Praedatophasma and Tyrannophasma by the lack of spination in most body regions (e.g. the thoracic nota); from Karoophasma, Hemilobophasma, Namaquaphasma, and Austrophasma rawsonvillense by the striped-type colouration of the compound eyes and uniform type of its body colouration (Fig. 1a, c).
https://static-content.springer.com/image/art%3A10.1007%2Fs13127-010-0037-8/MediaObjects/13127_2010_37_Fig1_HTML.gif
Fig. 1

Habitus and important colour characteristics of Viridiphasma clanwilliamense sp. n. (a, c, e) and Karoophasma biedouwense (b, d, f). (a, b) Respective males and females in copula; male on top, bending its abdomen around right side of female. (c) Head of adult male V. clanwilliamense sp. n., body colouration faded due to alcohol fixation; note eye of striped type, and absence of dark patches from maxillary palpomeres (arrows), scapus (arrowhead) and head. (d) Head of adult male K. biedouwense; note eye of mosaic type, and dark patches on 3rd and 4th maxillary palpomeres (arrows) and on scape of antenna (arrowhead). (e) Fore femur of female V. clanwilliamense sp. n., showing setal spots (arrows), one beneath each seta. (f) Fore femur of female K. biedouwense; note circumsetal rings (arrows) and dark femoral stripes. (a, b) Photos by Stefan H. Eberhard. Scale bars: 1 mm (c, d); 0.5 mm (e, f)

Male Viridiphasma gen. n. differ from those in genera of Mantophasmatidae by, e.g., the presence of sclerotized phallic hooks (phl, phs), the absence of phallic sclerites PH5 (opposite to PH2) and PH8, and by the oblique rather than straight transverse course of the ridge on the vomeroid. They differ from males in all other genera of Austrophasmatidae by the presence of a process on phallic sclerite PH2 (Fig. 2k) and absence of a basal sclerite extension of the cercus (Fig. 2d).
https://static-content.springer.com/image/art%3A10.1007%2Fs13127-010-0037-8/MediaObjects/13127_2010_37_Fig2_HTML.gif
Fig. 2

Elements of male postabdomen in Viridiphasma clanwilliamense sp. n. (a, c, d, g, i, k, l, o, p) and Karoophasma biedouwense (b, e, f, h, j, m, n, q, r). (a, b) Cerci in dorsal and slightly mesal view, posterior ends at top; respective bottom contour represents border of sclerotization. (c, e) Distal third of left cercus in posterolateral view that is perpendicular on widened part; dorsal side at left, posteromesal at top; transverse baseline indicates section cutoff. (d, f) Base of left cercus in posteroventral view, lateral side at left, anterior (proximal) end at bottom; narrow oval area at top indicates section cutoff. (g, h) Abdominal tergum X in dorsal view and undistorted condition, with midline portions of anterior and posterior margins on same level; bases of cerci indicated at top. (i, j) Vomeroid in dorsal view, with tips of left and right arms on same level; anterolateral corners of tergum X, articulated upon the arms, indicated schematically at bottom. (k, m) Phallic sclerite PH2 in dorsal (internal) view (if genitalia invaginated), anterior end at top. (l, n) Phallic hook phl and its sclerite PH3 in posterodorsal (external) view (if genitalia invaginated), anterodorsal side at top; arrow indicates position of tip of second phallic hook phs. (o, q) Phallic sclerite PH1 in dorsal (internal) view (if genitalia invaginated); anterior side at top. (p, r) Phallic lobe pli in anterodorsal and slightly left view (if genitalia invaginated) that is perpendicular on flattened lobe. Scale bars: 0.5 mm (a, b, g, h); 0.2 mm (cf, ir)

Female Viridiphasma gen. n. differ from those in genera of Mantophasmatidae by the short length of apodemes at8 of the subgenital plate and the lack of any vaginal sclerotization. They differ from female Karoophasma or Hemilobophasma by the near-circular shape of the spermathecal bulb, from Lobatophasma by the entirely unlobed hind margin of the subgenital plate, from Austrophasma by the much shorter subgenital plate, and from Namaquaphasma by the sclerotization of the subgenital plate continuing across the midline. In contrast to all other Austrophasmatidae, female Viridiphasma gen. n. possess some setae between the gonapophyses gp9 (Fig. 4).

Viridiphasma clanwilliamense sp. n.

Etymology

The specific epithet reflects the name of the type locality, Clanwilliam settlement. It is to be treated as adjectival for the purposes of nomenclature.

Material examined

Holotype: ♂ (deposited at SAMC, specimen Cw3♂); South Africa, Western Cape Province, near Clanwilliam at Clanwilliam Dam, 32.21°S 18.88°E, August 2007, leg. M. Eberhard & S.H. Eberhard.

Paratypes: 2 ♂♂, 2 ♀♀ (SAMC; specimens Cw1♂, Cw4♂, Cw1♀, Cw3♀, Cw4♀); 4 ♂♂, 4 ♀♀ (MTD; specimens Cw2♂, Cw5♂, Cw6♂, Cw7♂, Cw2♀, Cw5♀, Cw6♀, Cw7♀); other data as for holotype. DNA samples and vouchers of specimens Cw1♂, Cw2♂, Cw1♀, Cw2♀ in Damgaard et al. (2008; under the name “Austrophasmatidae sp.n. 3”) belong to V. clanwilliamense sp. n., as their morphological features (Damgaard et al. 2008: table 2) match those of the specimens studied here; their collection site (Damgaard et al. 2008: “32.22°S 18.85°E”) is approx. 3 km away from the GPS-determined type locality.

Diagnosis

See the corresponding genus section above.

Description

Measurements [in mm]. Males Cw3♂/Cw4♂/Cw5♂/Cw6♂/Cw7♂: bodl = 11.5/11.2/11.4/11.6/11.2; heaw = 2.30/2.23/2.36/2.46/2.44; iocw = 1.52/1.44/1.55/1.56/1.67; antl = 10/?/11/12/12; thol = 4.81/4.41/5.13/4.55/5.09; prnl = 2.17/2.03/2.05/2.27/2.17; msnl = 1.59/1.57/1.74/1.61/1.61; mtnl = 1.35/1.03/1.38/1.14/1.52; prnw = 1.90/1.87/2.03/2.15/2.00; prfl = 3.33/3.24/3.28/3.51/3.36; prtl = 3.56/3.36/3.56/3.76/3.35; mtfl = 3.86/3.69/3.94/4.21/3.74; mttl = 4.99/4.62/4.79/5.25/4.97; eyel = 0.85/0.80/0.81/0.90/0.88; eyeh = 0.50/0.47/0.48/0.51/0.51; genh = 0.49/0.50/0.57/0.56/0.56; emad = 0.77/0.74/0.76/0.80/0.80; t10w = 1.68/1.67/1.63/1.61/1.66; t10l+ = 1.01/0.90/0.83/0.88/0.88; t10l= 0.90/0.79/0.72/0.73/0.74. – Females Cw3♀/Cw4♀/Cw5♀/Cw6♀/Cw7♀: bodl = 15.8/16.0/14.8/15.0/16.0; heaw = 2.75/2.86/2.66/2.73/2.72; iocw = 1.95/1.81/1.71/1.94/1.97; antl = 11.00/11.30/11.20/11.00/11.40; thol = 5.68/6.10/5.70/5.56/5.61; prnl = 2.62/2.63/2.46/2.54/2.53; msnl = 1.79/2.12/1.74/1.60/1.84; mtnl = 1.38/1.53/1.53/1.36/1.40; prnw = 2.17/2.75/2.38/2.51/2.44; prfl = 3.35/3.76/3.45/3.51/3.69; prtl = 3.68/3.70/3.54/3.65/3.74; mtfl = 4.19/4.39/4.06/4.15/4.06; mttl = 5.19/5.40/4.95/5.06/5.25; eyel = 1.01/1.03/1.01/0.95/0.98; eyeh = 0.65/0.67/0.55/0.57/0.61; genh = 0.69/0.75/0.66/0.67/0.64; emad = 1.05/1.10/0.96/1.01/0.98. Selected ratios are shown in Tables 1, 2, 3 and 4.
Table 1

Antenna lengths, antennomere and tibial spine numbers of the Mantophasmatodea specimens studied here

  

Antenna

Antennomere number

Tibial spine number

  

length [mm]

basifl. div.

all antms.

fore a

fore p

mid a

mid p

Species

Specimen

L

R

L

R

L

R

L

R

L

R

L

R

L

R

V. clanwilliamense

Cw3♀

11.0

11.2

11

11

27

27

9

9

11

9

7

6

6

7

V. clanwilliamense

Cw4♀

11.3

11.4

10

11

28

27

8

8

10

12

6?

6

7?

6

V. clanwilliamense

Cw5♀

10.6

11.2

11

12

27

26

7

8

10

10

7

7

?

6

V. clanwilliamense

Cw6♀

11.0

11.0

10

10

28

28

11

9

9

9

6

6

7

7

V. clanwilliamense

Cw7♀

11.4

?

8

?

30

?

8

10

10

10

6

6

6

6

V. clanwilliamense

Cw3♂

10.1

10.4

10

11

28

27

8

6

7

9

6

5

8

8

V. clanwilliamense

Cw4♂

?

?

8

10

?

?

7

8

8

9

7

6

7

6

V. clanwilliamense

Cw5♂

11.4

11.2

8

8

30

30

6

6

8

7

7

7

6

6

V. clanwilliamense

Cw6♂

11.6

11.6

8

8

29

30

8

8

9

9

7

6

6

7

V. clanwilliamense

Cw7♂

?

11.5

8

8

?

30

9

8

8

10

7

6

7

8

K. biedouwense

Cw8♀

12.3

12.2

8

8

30

30

11

11

9

9

5

5

6

6

K. biedouwense

Cw9♀

10.3

10.5

7

7

27

28

10

10

10

10

5

5

7

7

K. biedouwense

Cw10♀

11.3

12.4

8

8

30

30

10

10

10

12

6

6

6

7

K. biedouwense

Cw11♀

11.2

11.5

?

?

?

?

12

9

11

?

5

5

7

6

K. biedouwense

Cw12♀

10.8

10.6

8

8

30

30

10

10

8

7

5

5

5

6

K. biedouwense

Cw8♂

11.6

11.8

10

10

28

28

10

7

8

8

6

6

7

7

K. biedouwense

Cw9♂

11.7

11.9

10

10

28

28

8

11

8

9

4

4

6

6

K. biedouwense

Cw10♂

10.4

10.5

10

10

28

28

11

?

10

?

5

5

5

6

K. biedouwense

Cw11♂

11.7

11.8

9

9

28

29

8

11

8

6

6

6

6

5

K. biedouwense

Cw12♂

10.6

11.3

23

23

10

10

8

8

5

4

6

6

basifl. div. = basalmost antennomere of basiflagellum that displays a subdivision; all antms. = total number of antennomeres (with consideration of basiflagellomere subdivisions = ‘weak articulations’)

fore a = on anterior face of fore tibia; fore p = on posterior face of fore tibia; mid a = on anterior face of mid tibia; mid p = on posterior face of mid tibia

L, R = left, right structure in a bilateral pair; ? = no data (usually due to incomplete or missing structure); – = none of the basiflagellomeres subdivided

Table 2

Some numerical ratios (for abbreviations, see text) measured on the Mantophasmatodea specimens studied here; within each species, specimens sorted by sex

Species

Specimen

prnl/msnl

prnl/prtl

prnl/prnw

antl/prnl

heaw/prnl

heaw/prnw

antl/heaw

antl/prfl

heaw/iocw

mttl/prtl

mttl/mtfl

V. clanwilliamense

Cw3♀

1.46

0.71

1.21

4.21

1.05

1.27

4.00

3.28

1.41

1.41

1.24

V. clanwilliamense

Cw4♀

1.24

0.71

0.96

4.29

1.09

1.04

3.95

3.01

1.58

1.46

1.23

V. clanwilliamense

Cw5♀

1.42

0.70

1.03

4.31

1.08

1.12

3.98

3.07

1.55

1.40

1.22

V. clanwilliamense

Cw6♀

1.58

0.70

1.01

4.33

1.07

1.09

4.04

3.13

1.41

1.39

1.22

V. clanwilliamense

Cw7♀

1.37

0.68

1.04

4.51

1.08

1.12

4.19

3.09

1.39

1.40

1.29

V. clanwilliamense

Cw3♂

1.37

0.61

1.14

4.66

1.06

1.21

4.40

3.03

1.51

1.40

1.29

V. clanwilliamense

Cw4♂

1.29

0.60

1.09

?

1.10

1.19

?

?

1.54

1.38

1.25

V. clanwilliamense

Cw5♂

1.17

0.57

1.01

5.57

1.15

1.16

4.84

3.48

1.52

1.34

1.22

V. clanwilliamense

Cw6♂

1.41

0.60

1.06

5.10

1.08

1.14

4.71

3.30

1.58

1.40

1.25

V. clanwilliamense

Cw7♂

1.35

0.65

1.09

5.30

1.13

1.22

4.71

3.42

1.46

1.48

1.33

K. biedouwense

Cw8♀

1.22

0.62

1.07

4.88

1.24

1.33

3.95

3.14

1.70

1.49

1.22

K. biedouwense

Cw9♀

1.13

0.60

1.03

4.49

1.23

1.26

3.66

2.75

1.60

1.53

1.28

K. biedouwense

Cw10♀

1.12

0.65

1.11

3.99

1.14

1.27

3.49

2.73

1.63

1.54

1.27

K. biedouwense

Cw11♀

1.13

0.67

0.94

4.49

1.27

1.18

3.55

3.12

1.98

1.52

1.16

K. biedouwense

Cw12♀

1.05

0.67

1.11

4.11

1.24

1.38

3.31

2.82

1.72

1.55

1.27

K. biedouwense

Cw8♂

1.07

0.61

1.15

5.51

1.19

1.37

4.65

3.37

1.79

1.54

1.23

K. biedouwense

Cw9♂

1.08

0.61

1.20

5.38

1.21

1.45

4.44

3.43

1.81

1.60

1.30

K. biedouwense

Cw10♂

0.95

0.54

1.04

5.08

1.26

1.32

4.02

2.87

1.85

1.54

1.26

K. biedouwense

Cw11♂

1.06

0.57

1.08

5.63

1.21

1.31

4.67

3.51

1.68

1.42

1.16

K. biedouwense

Cw12♂

1.13

0.62

1.18

4.57

1.14

1.35

3.99

2.92

1.62

1.43

1.14

? = no data (due to incomplete or deformed antenna)

Table 3

Some numerical ratios related to eye size (for abbreviations, see text) measured on the Mantophasmatodea specimens studied here; within each species, specimens sorted by sex

Species

Specimen

eyel/eyeh

eyeh/genh

eyeh/emad

eyel/genh

eyel/emad

heaw/eyeh

prfl/eyeh

prfl/eyel

prnl/eyeh

prnl/eyel

prfl/genh

prnl/genh

V. clanwilliamense

Cw3♀

1.55

0.94

0.62

1.46

0.96

4.24

5.17

3.33

4.03

2.60

4.88

3.80

V. clanwilliamense

Cw4♀

1.54

0.88

0.60

1.36

0.93

4.29

5.64

3.65

3.96

2.56

4.98

3.49

V. clanwilliamense

Cw5♀

1.84

0.83

0.57

1.52

1.05

4.85

6.30

3.43

4.49

2.45

5.23

3.73

V. clanwilliamense

Cw6♀

1.65

0.86

0.57

1.41

0.94

4.75

6.11

3.71

4.42

2.68

5.24

3.79

V. clanwilliamense

Cw7♀

1.59

0.96

0.62

1.54

1.00

4.44

6.02

3.78

4.13

2.59

5.80

3.98

V. clanwilliamense

Cw3♂

1.71

1.02

0.65

1.75

1.11

4.61

6.68

3.90

4.34

2.54

6.82

4.43

V. clanwilliamense

Cw4♂

1.71

0.93

0.63

1.58

1.07

4.78

6.95

4.07

4.35

2.54

6.44

4.03

V. clanwilliamense

Cw5♂

1.69

0.85

0.63

1.43

1.06

4.92

6.85

4.05

4.27

2.53

5.79

3.61

V. clanwilliamense

Cw6♂

1.75

0.91

0.64

1.61

1.12

4.82

6.87

3.92

4.45

2.53

6.29

4.07

V. clanwilliamense

Cw7♂

1.68

0.95

0.66

1.59

1.11

4.63

6.38

3.80

4.11

2.45

6.04

3.90

K. biedouwense

Cw8♀

1.50

1.45

0.84

2.17

1.26

3.75

4.72

3.16

3.04

2.03

6.85

4.41

K. biedouwense

Cw9♀

1.70

1.75

0.84

2.98

1.43

3.80

5.05

2.98

3.10

1.83

8.86

5.43

K. biedouwense

Cw10♀

1.59

1.55

0.81

2.47

1.29

3.93

5.02

3.15

3.43

2.15

7.77

5.31

K. biedouwense

Cw11♀

1.61

1.56

0.90

2.52

1.45

3.80

4.32

2.68

3.00

1.86

6.75

4.69

K. biedouwense

Cw12♀

1.66

1.86

0.94

3.10

1.56

3.90

4.58

2.75

3.13

1.89

8.53

5.84

K. biedouwense

Cw8♂

1.54

2.28

1.03

3.53

1.59

3.52

4.86

3.15

2.97

1.92

11.11

6.79

K. biedouwense

Cw9♂

1.67

2.12

1.06

3.54

1.77

3.78

4.89

2.92

3.11

1.86

10.35

6.59

K. biedouwense

Cw10♂

1.64

1.91

0.92

3.14

1.52

3.86

5.41

3.29

3.06

1.86

10.33

5.84

K. biedouwense

Cw11♂

1.73

1.78

0.94

3.09

1.63

3.94

5.23

3.03

3.26

1.88

9.34

5.82

K. biedouwense

Cw12♂

1.47

2.47

1.19

3.63

1.74

3.44

4.71

3.21

3.01

2.05

11.64

7.44

Table 4

Width/length ratios (for abbreviations, see text) for abdominal tergum X of the male Mantophasmatodea specimens studied here

Species

Specimen

t10w/t10l+

t10w/t10l

Categorization

V. clanwilliamense

Cw3♂

1.67

1.86

short

V. clanwilliamense

Cw4♂

1.85

2.12

short

V. clanwilliamense

Cw5♂

1.97

2.25

short

V. clanwilliamense

Cw6♂

1.82

2.21

short

V. clanwilliamense

Cw7♂

1.89

2.24

short

K. biedouwense

Cw8♂

1.84

2.57

short

K. biedouwense

Cw9♂

2.01

2.90

short

K. biedouwense

Cw10♂

1.80

2.44

short

K. biedouwense

Cw11♂

1.87

2.60

short

K. biedouwense

Cw12♂

2.03

2.66

short

Colouration

Of uniform type (Fig. 1a, c, e): Small setal spots present in all body parts (Fig. 1c, e), except on head. Dark epidermal patches, but no vermiculate lines, present in areas having setal spots (and underlying these), here displaying close correlation with setation (absent in other body parts); accordingly, 3rd and 4th maxillary palpomeres lacking extensive patch dorsodistally (Fig. 1c); scape of antenna lacking large dark patch ventrally (Fig. 1c); external face of mandible lacking well-contrasted, large dark patch (Fig. 1c); tibiae and femora without epidermal patches that are much larger than setal spots, no patches forming circumsetal rings (Fig. 1e); tibia lacking large epidermal patch externally near its tip. Cuticular midtergal and laterotergal stripes and epidermal middorsal stripes (almost) absent. Pleuromembraneal stripes and sternal patches absent. Femoral stripes and tibial stripe not darkened, conforming with general colour of femur and tibia. Cuticular basitibial patch absent, only some distinct setal spots in this area. Laterovertical pattern vestigial or absent (area consistently displaying basic colour). Basic body colour (in life) bright green, including all parts of legs. Dorsal ribbon usually well developed and continuous, reddish, brown or dark brown in males; lacking in females. Epidermal laterodorsal stripes (= lateral stripes of dorsal ribbon) distinct to a varied extent, light or dark. Ventral ribbon dark brown, ranging from very distinct to vestigial; only on abdomen in males, in females vestiges in various parts of thorax. Pleural ribbon absent. Dark median spots below bases of antennae absent. Eyes of striped type: discrete dark or reddish longitudinal ribbon through upper third of eye (Fig. 1c).

Head

Antennomeres [basalmost basiflagellomere that shows subdivision](total number of antennomeres including weak articulations): [8](30), [10](28), [11](27), or [12](26) (Table 1). Eye height slightly less than or same as gena height; eyeh/genh = 0.85–1.02 (male) or 0.83–0.96 (female) (Table 3).

Legs

Spines on fore tibia: anterior row 6–10, posterior row 7–11; on mid tibia: anterior row 5–7, posterior row 6–8 (Table 1).

Male postabdomen

Tergum X very short; t10w/t10l+ = 1.67–1.97; t10w/t10l = 1.86–2.25 (Table 4); hind margin between cerci in dorsal view with a narrow but fairly deep recess around the middle (Fig. 2g).

Cerci in dorsal view as in Fig. 2a, sclerotization moderately widened mesally at dorsal base of cercus; in lateral view narrowest around middle; only slightly widened in distal 2/5 (ventral edge only slightly convex; Fig. 2a, c). Cerci not altogether flattened, with circular cross section; external and internal faces of cerci weakly convex; well developed setation throughout whole surface of both cerci. Tips of cerci with large, finger-like dorsal projections (Fig. 2c: dpc). Projection and outermost tip of cercus with a coarse, irregular, tooth-like serration. At ventral cercal base, cercal sclerotization lacking an extension (Fig. 2d).

Vomeroid as in Fig. 2i; its body moderately wide. Ridge of vomeroid ascending in right 1/3 of vomeroid body, with leftward inclination deviating from vertical line by c. 10–20° (posterior view); in dorsal wall of vomeroid bent (c. 90–100°) to the left, continuing leftward and anteriorly as an anterodorsally projecting ridge, then curved somewhat posteriorly towards left base of vomeroid body, where it ends. Ridge lacking discrete denticles but with very fine serration. Vomeroid body weakly bulging on left flank of ridge, moderately and irregularly on right flank. Vomeroid arms strongly laterally directed; left arm altogether somewhat widened, tips without extensions.

Sclerite PH1 quite variable in shape (as in Fig. 2o), but always somewhat horseshoe-shaped with anterior arm reduced. Sclerite PH2 conspicuously large, extended in anteroposterior direction; ridge of PH2 located far to the left, its posterior portion strongly expanded into a sclerotized triangular process projecting mesally (Fig. 2k: process covered by sclerite, hence indicated by dashes). Lobe pli (Fig. 2p) strongly convex and angled, mesal tip indistinct or absent, lacking a sclerite PH7 along anteroventral edge. Hook phl unusually large compared to the other genitalic sclerites, very long, moderately and evenly curved, tip pointed, hook phs approximately the same length as hook phl (Fig. 2l).

Female postabdomen

Subgenital plate LS8 (Fig. 3a) short, its hind edge slightly and uniformly convex, not bilobate, middle part with a wide field of very weak sclerotization; sclerotization continuous but very weak across midline; no slopes between anteromedian and posterolateral parts; no particularly heavy sclerotization reaching apodemes at8 from posteromesally; apodemes at8 very small and hardly sclerotized (usually found in an irregularly folded condition). Anterior margin of LS8 and posterior margin of coxosternum CS7 closely hinged between apodemes at8 (LS8 and CS7 appear as almost fused, with fine suture in between), CS7 thus not overlapping LS8 (very slight overlap only near apodemes at8). Spermathecal bulb globular to slightly oval (Fig. 5a–e). The area between the two gonapophyses IX bears 2–3 setae per side (Fig. 4).
https://static-content.springer.com/image/art%3A10.1007%2Fs13127-010-0037-8/MediaObjects/13127_2010_37_Fig3_HTML.gif
Fig. 3

Laterocoxosternum VIII (subgenital plate, LS8) of female Viridiphasma clanwilliamense sp. n. (a) and Karoophasma biedouwense (b) in flattened, ventral view; posterior at top; setae omitted; undulating lines represent cuts through cuticle; darker/lighter grey indicates heavier/lighter sclerotization. Scale bars: 0.5 mm

https://static-content.springer.com/image/art%3A10.1007%2Fs13127-010-0037-8/MediaObjects/13127_2010_37_Fig4_HTML.gif
Fig. 4

Micrograph (a) and schematic (b) of distal abdomen in female Viridiphasma clanwilliamense sp. n., ventral view. (a) Specimen fixed in alcohol; gonapophyses VIII (gp8) out of focus; note setae (arrows) on area between gp8. (b) Distal abdomen with gonapophyses VIII removed; few setae (arrows) located in area enclosed by gonapophyseal sclerites (GP9) and posterior intervalvular sclerite (IP); undulating line represents cut through cuticle. Scale bars: 0.5 mm

Vibrational calls

Reported in Eberhard and Picker (2008; as “species A”).

Distribution

Known from the type locality (Damgaard et al. 2008: black square in fig. 1).

Revised diagnosis of Austrophasmatidae

As a consequence of the new character combination found in Viridiphasma gen. n., the diagnosis of Austrophasmatidae given in Klass et al. (2003b) has to be reconsidered as the diagnosis for ‘Austrophasmatidae without Viridiphasma’ (assuming this group to be monophyletic; see below). The inclusion of Viridiphasma requires the diagnosis of Austrophasmatidae to be modified with regard to characters (i)–(iv) in the above diagnosis of the genus, as follows. (i) In males, phallic sclerite PH2 bears a ridge and may additionally bear a process; (ii) a sclerite extension at the ventral base of the cercus is present or absent; (iii) in females, there are no, or only 2–3, setae on the membrane between the gonapophyses of segment 9 (gp9); (iv) the eye is higher than, or subequal in height to, the gena. Regarding all other characters, the diagnosis of Austrophasmatidae in Klass et al. (2003b) remains valid.

Re-examination of Karoophasma biedouwense

Material examined

5 ♂♂ (MTD; specimens Cw8♂, Cw9♂, Cw10♂, Cw11♂, Cw12♂), 5 ♀♀ (MTD; specimens Cw8♀, Cw9♀, Cw10♀, Cw11♀, Cw12♀); South Africa, Western Cape Province, at Clanwilliam Dam, 32.21°S 18.88°E, August 2007, leg. M. Eberhard & S.H. Eberhard.

Results

Specimens from Clanwillam correspond with the original description of K. biedouwense (Klass et al. 2003b) in all characters. This includes the mottled-type and loose-correlation-subtype body colouration, involving dark patches dorsally on the 3rd and 4th maxillary palpomeres and ventrally on the scape of the antenna (Fig. 1d), as well as circumsetal rings on the femur and tibia (Fig. 1f). Eye colouration is of the mosaic type (Fig. 1d). In the male postabdomen, the vomeroid shows the same characteristics in structure and shape; the phallic hook phl, the phallic lobe pli, abdominal tergum 10, and the apical parts of the cerci are of similar shape, and the basal sclerotization of the cercus forms a tongue-like extension ventrally (compare Fig. 2b, e, f, h, j, m, n, q, r with Klass et al. 2003b: fig. 16). In the female postabdomen the spermathecal bulb is of the same shape (slender and coiled) and the subgenital plate shows the same characteristics, e.g. the straight hind margin (compare Figs. 3b, 5f–j with Klass et al. 2003b: fig. 17). Almost all ratios calculated in the present study (Tables 1, 2, 3 and 4) are similar to the numbers given for K. biedouwense in Klass et al. (2003b); the only difference, concerning ratios between length of pronotum (prnl) and length of protibia (prtl), proved due to incorrect calculation in Klass et al. (2003b). Recalculation of the prnl/prtl of K. biedouwense resulted in similar numbers (around 0.6) as obtained in the present study.
https://static-content.springer.com/image/art%3A10.1007%2Fs13127-010-0037-8/MediaObjects/13127_2010_37_Fig5_HTML.gif
Fig. 5

Schematics and photographs (four females each) of spermathecae in Viridiphasma clanwilliamense sp. n. (a–e) and Karoophasma biedouwense (f–j). Spermathecal bulb (spb) connected to vagina (va) via spermathecal tube (spt); arrow points to visible connection between spermathecal tube and vagina. Scale bars: 0.5 mm

Within either sex, there is little morphological variation among the Clanwilliam specimens of K. biedouwense. One more discrete variation reported in Klass et al. (2003b) for K. biedouwense from other localities is the presence or absence of a narrow midline membrane in the female subgenital plate. The Clanwilliam specimens consistently lacked this membranous stripe. Phallic sclerite PH1 was generally found to be quite variable, both in the formerly known populations and in that from Clanwilliam.

In some characters, on the other hand, there seem to be (nearly) consistent minor differences between K. biedouwense from Clanwilliam and those from the previous localities. In the Clanwilliam males the distal part of the cerci was much less widened (but very limited distal widening was also found in one exceptional male, Bi2♂, in the earlier study; compare Fig. 2e with Klass et al. 2003b: fig. 16E). The sclerotization at the base of the cercus was slightly wider in the Clanwilliam males than in those from the previous localities (compare Fig. 2b with Klass et al. 2003b: fig. 16A, B). The mesal tip of genital lobe pli was longer in the Clanwilliam males than in other K. biedouwense (compare Fig. 2r with Klass et al. 2003b: fig. 16W, X, Y). In the female, the subgenital plate was slightly longer in the Clanwilliam specimens than in those from the previous localities (compare Fig. 3b with Klass et al. 2003b: fig. 17A, E). Studies of larger samples are needed to test the consistency of these differences.

By the possession of all the typical features of K. biedouwense, the Clanwilliam specimens of this species are clearly distinguished from the sympatric V. clanwilliamense sp. n. The two species are distinguished easily by discrete, non-overlapping character state distributions in 18 characters (Table 5). No individuals with intermediate character states were found. Additionally, principal component analysis (PCA) of measured morphometric data reveals well separated groups (Fig. 6). Principal component 1, which is strongly associated with body length, head width and leg lengths, separates males from females and can therefore be interpreted as a factor of overall body size. Principal component 2, best represented by gena height and distance between eye and mandibular articulation, distinguishes well between the two species (p < 0.001, Mann-Whitney U test). The first two principal components account for 85.3% of the total variance. PCA of the ratios gave similar results (data not shown).
Table 5

Morphological differences between Viridiphasma clanwilliamense sp. n. and Karoophasma biedouwense Klass et al. (see also Klass et al. 2003b: figs. 16, 17)

#

Character

Viridiphasma clanwilliamense

Fig.

Karoophasma biedouwense

Fig.

 

BOTH SEXES

    

1

colouration of compound eye

striped type

1c

mosaic type

1d

2

colouration of body

uniform type

1a

mottled type, loose-correlation subtype

1b

2a

circumsetal rings on legs

absent

1e

present

1f

2b

dark patches on maxillary palpomeres 3 and 4

absent

1c

present

1d

2c

dark ventral patch on antennal scape

absent

1c

present

1d

2d

conspicuous dark patch laterally on mandible

absent

1c

present

1d

3

ratio eye height/gena height

female 0.83–0.96; male 0.85–1.02

1c

female 1.36–1.86; male 1.63–2.47

1d

 

MALE

    

4

sclerotization at ventral base of cercus

without an extension

2d

with a long, tongue-like extension

2f

5

widening of distal part of cercus

moderately widened, main part gradually tapering

2c

distinctly widened, main part lobe-like and convex

2e

6

sculpture of distal part of cercus

coarsely sculptured

2c

finely sculptured

2e

7

vomeroid ridge on dorsal side of vomeroid

strongly curved and reaching far to the left

2i

weakly curved and not reaching far to the left

2j

8

denticles of vomeroid ridge

fine serration with numerous tiny denticles

2i

coarse serration with 10–13 larger denticles

2j

9

location of ridge on phallic sclerite PH2

far to the left

2k

to the right of middle

2m

10

process on phallic sclerite PH2

distinct, triangular

2k

absent

2m

11

phallic hook phl

evenly curved, pointed

2l

abruptly curved near middle, blunt

2n

12

shape of main part of lobe pli

strongly convex and distinctly angled

2p

strongly convex and rounded throughout

2r

13

mesal tip of lobe pli

indistinct or absent

2p

very distinct

2r

 

FEMALE

    

14

anterior apodemes at8 of subgenital plate

very small, hardly sclerotized

3a

quite large, distinctly sclerotized

3b

15

particularly strong sclerotization extending from at8 towards center of subgenital plate

absent

3a

present

3b

16

condition of sclerotization at midline of subgenital plate

strongly weakened

3a

uniformly strong, or slightly weakened immediately at midline

3b

17

shape of spermathecal bulb

globular to slightly oval

5a–e

slender and coiled (strongly curved)

5f–j

18

setae between gonapophyses gp9

about 3 per side

4

absent

 
https://static-content.springer.com/image/art%3A10.1007%2Fs13127-010-0037-8/MediaObjects/13127_2010_37_Fig6_HTML.gif
Fig. 6

Principal components plot from Principal Component Analysis using all measured morphometric data (except of male tergum X). Principal component 1 separates males from females; principal component 2 clearly separates Viridiphasma from Karoophasma. Table shows component loadings of the two factors; values in boldface indicate variables most strongly associated with the respective component

Discussion

Phylogenetic position of V. clanwilliamense sp. n.

According to Klass et al. (2003b) and Damgaard et al. (2008), Mantophasmatodea fall into three clades (Fig. 7): (1) Mantophasmatidae with Mantophasma and Sclerophasma (limited to Namibia); (2) a clade comprising Praedatophasma and Tyrannophasma (Namibia and northernmost South Africa); and (3) Austrophasmatidae (limited to South Africa) with Austrophasma, Karoophasma, Hemilobophasma, Lobatophasma, Namaquaphasma, Viridiphasma (formerly “Austrophasmatidae sp.n. 3”), and a clade comprising two undescribed taxa (“Austrophasmatidae sp.n. 1+ sp.n. 2”). The aberrant Tanzaniophasma subsolanum was not included in Damgaard et al.’s analysis, thus omitted in Fig. 7. Austrophasmatidae formed a strongly supported lineage. Within it, Namaquaphasma ookiepense and Viridiphasma clanwilliamense were the two most basal lineages, and a clade comprising all remaining austrophasmatids was fairly well supported. However, the relationships among N. ookiepense, V. clanwilliamense, and the remaining Austrophasmatidae were not resolved clearly. Based on a combined COI and 16S analysis the preferred hypothesis was N. ookiepense + (V. clanwilliamense + remaining Austrophasmatidae), but bootstrap support for V. clanwilliamense + remaining Austrophasmatidae was below 50. Analysing COI alone did not provide resolution to basal splitting events in Austrophasmatidae, and 16S alone returned a clade N. ookiepense + V. clanwilliamense as sister to the remaining Austrophasmatidae.
https://static-content.springer.com/image/art%3A10.1007%2Fs13127-010-0037-8/MediaObjects/13127_2010_37_Fig7_HTML.gif
Fig. 7

Phylogenetic tree of Mantophasmatodea resulting from combined phylogenetic analysis of COI and 16S sequences (Damgaard et al. 2008; tree modified here); Mantophasma zephyrum (no molecular data) added via dashed line based on morphology of genitalia (Klass et al. 2003b). Numbers above branches are bootstrap values; nodes with very low support (<50) have been collapsed. In Namaquaphasma ookiepense, “S” refers to southern, “N” to northern group of populations

Morphological characters suggest V. clanwilliamense as the sister group of the remaining Austrophasmatidae including N. ookiepense. The new species shows several features that are absent from all other Austrophasmatidae but present in the outgroup taxon Mantophasmatidae (characters not yet studied in the Praedatophasma + Tyrannophasma clade and Tanzaniophasma subsolanum), as follows.
  1. (1)

    Shape of phallic sclerite PH2: All previously described Austrophasmatidae have sclerite PH2 altogether level, except for a low longitudinal ridge (as in Fig. 2m); this also applies to N. ookiepense (Klass et al. 2003b: fig. 19G). In contrast, Sclerophasma paresisense, the only mantophasmatid for which details of the male postabdomen have been studied, lacks such a ridge but has a small process in the right posterior corner of sclerite PH2 (Klass et al. 2003b: fig. 22H). Viridiphasma clanwilliamense also has the ridge along PH2, but its posterior part is extended into a triangular process (Fig. 2k: dashed line). Assuming homology of the PH2 processes in S. paresisense and V. clanwilliamense, outgroup comparison suggests that within Austrophasmatidae the process of V. clanwilliamense is plesiomorphic, and its absence a synapomorphy for all other austrophasmatids.

     
  2. (2)

    Basal cercal sclerotization in male: In previously described Austrophasmatidae, including N. ookiepense, the sclerotization at the ventral base of the cercus forms a tongue-like extension (as in Fig. 2f and in Klass et al. 2003b: fig. 19D). This extension is absent in S. paresisense (Klass et al. 2003b: fig. 22D) and also in V. clanwilliamense (Fig. 2d). Outgroup comparison thus suggests that within Austrophasmatidae the absence of the extension in V. clanwilliamense is plesiomorphic, and its presence a synapomorphy for all other austrophasmatids.

     
  3. (3)

    Setae between gonapophyseal sclerites GP9: The membranous area enclosed by sclerites GP9 laterally and sclerite IP posteriorly is of similar structure and outline in all Austrophasmatidae and Mantophasmatidae. In Mantophasmatidae this area bears, on each side next to the midline, a longitudinal group of conspicuous setae (Klass et al. 2003b: fig. 23F), while in all previously described Austrophasmatidae, including N. ookiepense, such setae are completely absent (as in Klass et al. 2003b: fig. 7). In V. clanwilliamense there are 2–3 setae on each side (Fig. 4). Thus, outgroup comparison suggests that within Austrophasmatidae the presence of (few) setae in V. clanwilliamense is plesiomorphic, and the lack of setae a synapomorphy for all other austrophasmatids.

     
  4. (4)

    Size of compound eye: In Austrophasmatidae the gena is always distinctly lower than the eye (eyeh/genh > > 1.0, as in Fig. 1c; 1.46–1.76 in N. ookiepense, see Klass et al. 2003b: table 3), while in the mantophasmatids S. paresisense and Mantophasma zephyrum the gena is higher than the eye (eyeh/genh < 0.8). Viridiphasma clanwilliamense is intermediate in this ratio (eyeh /genh = 0.8–1.0, Fig. 1c, Table 3). This relatively low ratio (i.e. small eyes) appears as a plesiomorphy, and a higher ratio (i.e. larger eyes) as a synapomorphy for all other austrophasmatids.

     

In summary, in the characters (1)–(4) V. clanwilliamense is uniquely plesiomorphic within Austrophasmatidae, and these are the characters in which Viridiphasma does not fit the diagnosis of Austrophasmatidae as given in Klass et al. (2003b; see description of genus above). In its remaining characters V. clanwilliamense shows conditions typical for Austrophasmatidae: In the male postabdomen, the tiny phallic hooks phl and phs together with their sclerites PH3 and PH4 are present (absent in S. paresisense), a phallic sclerite PH5 opposite to PH2 is absent (present in S. paresisense; Klass et al. 2003b: fig. 22G), a phallic sclerite PH8 is absent (present in S. paresisense; Klass et al. 2003b: fig. 22J), and the shape of the vomeroid is also ‘austrophasmatoid’ (e.g. with the ridge oblique rather than straight transverse as in S. paresisense, compare Fig. 2i with Klass et al. 2003b: fig. 22E). In the female postabdomen the apodemes at8 of the subgenital plate are very short (Fig. 3a; very long in Mantophasmatidae, see Klass et al. 2003b: figs. 21A, 23A), and the vagina lacks lateral sclerotization (sclerotized in Mantophasmatidae; Klass et al. 2003b: figs. 21D, 23E). These characters still differ between all studied Austrophasmatidae and Mantophasmatidae, and remain to be studied in T. subsolanum and the Praedatophasma + Tyrannophasma clade. Polarity is currently unclear for all of them (outgroup comparison with other insects is impossible or ambiguous throughout); hence their phylogenetic implications are also unclear.

Considering the previous genus-level classification of Mantophasmatodea (Klass et al. 2003b) and the phylogenetic results in Damgaard et al. (2008) and the present study, the erection of a new genus, Viridiphasma, for V. clanwilliamense sp. n. appears appropriate.

The previous hypothesis of N. ookiepense as the sister group of the remaining Austrophasmatidae fits well with the geographic distribution of mantophasmatodean taxa (see Damgaard et al. 2008: fig. 1): Non-austrophasmatid heelwalkers— members of Mantophasmatidae and the Praedatophasma + Tyrannophasma clade— occur far north, in Namibia and the immediately neighbouring parts of South Africa; N. ookiepense occurs in northwestern South Africa; and all other austrophasmatids are found further south. This would imply the phylogenetic radiation of Austrophasmatidae to have started in the northwestern part of South Africa (where Namaquaphasma occurs) and then proceeded southwards along western South Africa. In contrast, V. clanwilliamense is placed in the geographic centre of the distribution area of Austrophasmatidae (Damgaard et al. 2008: fig. 1), which suggests that the origin of Austrophasmatidae may be further south.

Species distinction, endemism, and sympatry

The strongly supported taxon Austrophasmatidae (Damgaard et al. 2008) is limited to South Africa, where except for the northwestern-most corner no other mantophasmatodean subgroups are represented. According to taxonomic work and molecular phylogenies of Klass et al. (2003b) and Damgaard et al. (2008), altogether 12 defined OTUs of Austrophasmatidae appear to deserve the status as separate species (eight described previously, one described herein, and three still awaiting formal description). These individual units are distinguished by sufficient genetic distances as well as considerable morphological differences (mainly in the postabdomen, as typical for many insects) for them to be considered as valid biological species. Groups of specimens assigned to the same species were consistently found as distinct clades in the molecular tree and also share the same features of postabdominal morphology.

Although Mantophasmatodea have only been studied since the group’s discovery in 2002, current data suggest that each species of Austrophasmatidae is restricted geographically (Damgaard et al. 2008; Klass et al. 2003b), the widest distribution being that of Hemilobophasma montaguense across the Little Karoo (populations up to 200 km apart). However, this picture of strictly vicariance-based endemism may change when more than the currently fairly limited sampling is accomplished. So far, only two other cases of sympatric Mantophasmatodea species have been reported (see Damgaard et al. 2008: table 1, fig. 5). The first concerns Rawsonville and the species Austrophasma rawsonvillense, the common species at that locality, and Lobatophasma redelinghuysense with its main populations 100 km to the north. In this case two specimens of L. redelinghuysense were found among quite many A. rawsonvillense; however, as their 16S and COI gene sequences were similar or identical to L. redelinghuysense from Redelinghuys (far away from Rawsonville) rather than to L. redelinghuysense from, e.g., Jonkiespoort or north of Ceres (much closer to Rawsonville) (see Damgaard et al. 2008: figs. 3, 4), a confusion of samples might be suspected. The area in question should be resampled for mantophasmatodean species. The second case of sympatry, involving Austrophasma gansbaaiense and A. rawsonvillense at Gansbaai, is most likely the result of confusion of samples.

In the present study we report on two species that are regularly syntopic. Each of the two species co-occurring at Clanwilliam shows very little intraspecific variation in all the diagnostic characters introduced as taxonomically useful by Klass et al. (2003b). On the other hand, the two species were found to differ in at least 18 characters (Table 5), as well as in morphometrics (Fig. 6). The most conspicuous differences concern the condition of phallic sclerite PH2, phallic hooks phl and phs, phallic lobe pli, the vomeroid ridge and its serration, and both the tip and the base of the cerci in the males, and the shape of the spermathecal bulb and several features of the subgenital plate in the females. The shape of the spermathecal bulb (Fig. 5) is the clearest distinguishing character in the female postabdomen, being near-globular in all V. clanwilliamense but very slender and coiled in all K. biedouwense (as in all K. biedouwense from previously reported localities). Differences in basic body colouration and in the colouration and relative size of the compound eyes allow for easy distinction of the two species in the field.

Karoophasma biedouwense specimens from Clanwilliam were found to be morphologically congruent with other populations of this species. Since the diagnostic morphological features for each of the two species examined here were never mixed in any individual, and since no intermediate condition was found in any character, it would appear that these (fairly unrelated) species do not hybridize in nature. This is not unexpected, as individuals of the two species also differ in their vibrational communication signals used for mate location and recognition (Eberhard and Picker 2008). Virtually all call parameters of the male and female song differ significantly between the two species, and in playback experiments males and females of K. biedouwense responded to conspecific calls but not to those of males or females of V. clanwilliamense (Eberhard and Picker 2008).

The two species also demonstrate microhabitat selection at Clanwilliam, congruent with their differences in colouration. Viridiphasma as a rule only occurred on the small tree Euclea racemosa (Ebenaceae), often at a height of 2 metres, where its bright green colouration matched that of the large leaves of the tree. This species was not found to be polymorphic for body colour. In contrast, K. biedouwense displayed a range of colour morphs (in various shades of brown, grey and pale yellow) that matched the bare stems of the low shrubs and grasses in which they lived. The fynbos vegetation is typically microphyllous; therefore, slow-moving insects vulnerable to bird and lizard predation need to mimic stem rather than leaf colouration. This separation of the two species probably limits encounters between them, reducing inter-specific predation by the larger K. biedouwense, and potential inter-specific matings.

Of course, none of the above excludes the possibility of occasional hybridization between K. biedouwense and V. clanwilliamense or among mantophasmatodean species in general, although hybrid inviability may limit this. The taxon sampling of Damgaard et al. (2008) did not reveal any signs of introgression, although sympatry between species pairs was rarely observed in any case. It remains to be investigated if V. clanwilliamense is reproductively isolated from species potentially more closely related to it than K. biedouwense. Currently, with V. clanwilliamense most likely being sister to all other Austrophasmatidae (the hypothesis preferred here), or perhaps to all Austrophasmatidae minus N. ookiepense (preferred hypothesis in Damgaard et al. 2008), no closer relatives are known. Nonetheless the allopatric N. ookiepense, which in one phylogenetic tree in Damgaard et al. (2008) is sister to V. clanwilliamense, may appear as a candidate species for such tests. Morphological details presented here (compare Figs. 2, 3 with Klass et al. 2003b: figs. 19, 20) and characteristics of the vibrational communication signals (Eberhard, unpubl. data) indicate that these two species are reproductively isolated from each other as well.

Austrophasmatidae are restricted to the winter rainfall biomes of South Africa, viz. the Fynbos, Succulent Karoo and Nama Karoo, each of which includes a number of different vegetation types. Endemism in Austrophasmatidae seems to be at most partly correlated with the pattern of vegetation types. At least some of the species occur across several biomes and vegetation types, although particular colour morphs of a species can prevail in different kinds of habitats. For instance, Lobatophasma redelinghuysense has a brown morph in the dry, Namaqualand sand fynbos vegetation type that is replaced in the east by a green morph in populations occupying the more mesic Cederberg sandstone fynbos (Klass et al. 2003b). Such colour matching is widespread in polymorphic Polyneoptera (Bond 2007). Karoophasma biedouwense has been recorded from both the arid Succulent Karoo biome as well as from the more mesic Fynbos biome, and displays different colour morphs matching the local vegetation. Accordingly, close relatives or conspecific populations of V. clanwilliamense may well be detected in quite different habitats and might well have different colouration.

Despite much successful collecting in recent years, the South African mantophasmatodean fauna still is little explored. Numerous additional localities where heelwalkers occur surely remain to be found with further sampling, and this will most likely also lead to the discovery of currently unknown habitats and morphotypes of certain species, of additional species, and perhaps of further cases of sympatry. While the known localities appear as widely separated ‘islands’, further sampling may well show that populations within as well as across species have a fairly continuous distribution and are geographically much more closely in touch. The discreteness of species areas, the issue of hybridization along borderlines, and the extent of gene flow within species and between closely related ones remain to be examined based on more intense collecting. Presently the extent of gene flow is difficult to estimate. The presence of regionally defined within-species clades, e.g. the northern and the southern clade in N. ookiepense, the most usual presence of identical or very similar haplotypes in the same or neighbouring localities (Damgaard et al. 2008), and the slight morphological differences reported here for K. biedouwense from the two flanks of the Cederberg Mountains indicate that gene flow over longer distances is quite limited. The very few reported cases of identical haplotypes of mitochondrial genes being present in conspecific specimens from widely separated localities, e.g. L. redelinghuysense specimens from Redelinghuys and Rawsonville (Klass et al. 2003b), are all doubtful. Of course, depending on geographic or climatic barriers, there might be a pattern of a few easy and many difficult routes of gene flow, as well as a species-specific gene flow (dependent on geographic and climatic conditions). All this means that the understanding of the biodiversity of Austrophasmatidae and Mantophasmatodea as a whole is still at a very early stage.

Acknowledgements

We would like to thank S. H. Eberhard for providing photographs for Fig. 1. Cape Nature is thanked for permit AAA007-00020-0035 to collect Mantophasmatodea and undertake this work at Clanwilliam. We thank Olaf Bininda-Emonds and two anonymous referees for helpful comments on the manuscript. M.P. received financial support through a University of Cape Town URC grant, M.J.B.E. was financially supported by a doc.award (University of Vienna, Kulturabteilung Stadt Wien).

Copyright information

© Gesellschaft für Biologische Systematik 2011