Acrotelsatinae, Mendes, 1991

Smith, Graeme B., 2016, Revision of the Genus Anisolepisma (Zygentoma: Lepismatidae: Acrotelsatinae), Records of the Australian Museum 68 (6), pp. 269-312 : 307-311

Acrotelsatinae redefinition

Macrochaetae feathered or smooth; antennae with basiconic sensillae; cylindrical sensilla of apex of maxillary palp distal article missing; anterior border of pronotum largely devoid of setal collar but often with 1+1 isolated pronotal tufts of macrochaetae or with a small collar occupying about half of the front margin; paramera tubuliform, long and thin, without glandular area with numerous fine setae; gonapophyses with thin setae only and sometimes apically provided with strong sclerotized cuticular teeth; all sterna not free, largely covered by coxae; 2+3 or 2+2 labial palp papillae.

Relationships

Anisolepisma seems to be very close to Primacrotelsa Mendes, 2004 with both genera showing some characters divergent to the rest of the Acrotelsatinae such as the smooth macrochaetae, the absence of bushes on the clypeus and the simple apex to the ovipositor (without cuticular spines).

The genus Heterolepisma Escherich, 1905 is generally considered to be the most primitive genus of the Lepismatidae but this “consensus” placed little emphasis on the morphology of the thoracic sterna. A review of the major morphological characters below suggests that this discussion could be re-opened in the light of our increasing knowledge of the order.

Thoracic sterna. It appears that all genera within the Acrotelsatinae lack free thoracic sternal plates which cover the medial base of the coxae. This character is not adequately described for species of Lepismina but the illustration of the ventral aspect Lepismina aurisetosa Wahlgren, 1906 in Schremmer (1964) clearly shows the coxae covering the thoracic sterna. The morphology of the meso and metathoracic sterna of the comparatively common peridomestic species Acrotelsa collaris do not appear to have been published. This arrangement is so contrary to that seen in most Lepismatidae (including Heterolepisma ) where the medial part of the sterna is developed into a free plate articulating only along the anterior margin and covering much of the internal anterior margins of the coxae, that it could be considered the most obvious defining character of the subfamily.

Mendes (1991) considered the genera Desertinoma (as Apteryskenoma Paclt without A. andersonae Womersley, 1928 ), Acrotelsa , Lepismina , Mirolepisma Silvestri 1938 , Monachina Silvestri, 1908 , and Prolepismina Silvestri 1940 as having a strongly reduced prothoracic sternum. In Prolepismina , Monachina and Mirolepisma the prosternum remains free but smaller in size and not covered by the coxae. This appears to be a reduction of the typical “free” sternite of the Lepismatidae . The meso and metasterna are also free and cover the base of the coxae. This is a fundamentally different arrangement to that seen in the Acrotelsatinae where the coxae largely cover the sterna (as in the Tricholepidiidae , Maindroniidae , Nicoletiidae and Protrinemuridae ). The medial portion of the sternum is somewhat raised between the anterior mediad corners of the coxae and that of the prosternum bears a medial tuft. In the case of Anisolepisma (and probably, judging from the published illustrations, also Primacrotelsa and Desertinoma ) the raised heart-shaped structure of the meso and metasterna have slight concave hollows below the lateral margins into which the anterior median margins of the coxae fit (see schematic representation in Fig. 190 ). The oval “sclerites” at the base of the heart shaped meso- and metasterna reported in Silvestri (1908) are a continuation of the sternum below the coxa, without any obvious suture separating the raised cordiform medial section from the rest of the sternum.

Could this condition represent a plesiomorphic state for the Lepismatidae ? Could the concave hollows beneath the raised cordiform section have become progressively deeper covering more of the coxae until the raised part of the sternum became a free sternum? If the form of the sternum in these genera represented a regression of the sternum, one would expect a decrease in size as in the prothoracic sternum of Prolepismina , Monachina and Mirolepisma , but not necessarily a fusion to the underlying sternal plates.

Macrochaetae. Both Anisolepisma and Primacrotelsa have smooth, apically bifurcate or smooth pointed macrochaetae. They do not have any feathered macrochaetae. In contrast the other genera of the Acrotelsatinae ( Acrotelsa , Apteryskenoma , Desertinoma and Lepismina ) display various degrees pectination. Smooth macrochaetae are characteristic of the Heterolepismatinae and Lepismatinae although some Ctenolepismatinae (e.g., Mormisma Silvestri, 1938 ) have “false-smooth” macrochaetae with rounded tips, seen as a secondary modification by Mendes (1988). All other genera have pectinate macrochaetae. Smooth macrochaetae are a feature of the Nicoletiidae , Protrinemuridae and Tricholepidiidae and generally considered to be plesiomorphic, suggesting Anisolepisma and Primacrotelsa are the only two known genera within the Acrotelsatinae to have retained this plesiomorphic character. At least some of the pectinate macrochaetae of Acrotelsa resemble the smooth apically bifurcate macrochaetae of Anisolepisma and Heterolepisma but with fine pectinations along the shaft (see Watson & Li, 1967, fig. 2) and different in appearance to the variety of pectinate macrochaetae seen in the Ctenolepismatinae . The pectinations in some Acrotelsatinae may be a parallel development to those of the Ctenolepismatinae .

Cephalic chaetotaxy. Most genera in the Acrotelsatinae , including Anisolepisma , have quite strongly developed anterior bushes on the frons; the exception being Primacrotelsa where the cephalic chaetotaxy is greatly reduced, anterior bushes are present but quite small. Mendes (1982) and Irish (1990) consider the plesiomorphic state as one completely lacking bushes with macrochaetae along the margins of the head as in Heterolepisma . While small anterior bristle bushes are present on the frons of at least some Maindroniidae , they are lacking in the Nicoletiidae , Protrinemuridae and Tricholepidiidae , where bristles occur scattered over the surface of the frons and often with some stronger bristles along the margins. This offers support to the argument that bristle bushes are apomorphic. However we see a reduction in density of macrochaetae within several genera of the Zygentoma (e.g., Primacrotelsa and also Qantelsella Smith, 2015 which seems to be a reduction in the much denser chaetotaxy observed in the related Acrotelsella Silvestri, 1935 ). The possibility that the chaetotaxy of Heterolepisma represents a reduction from an earlier bristle bush state should not be discarded.

Furthermore, the illustrations of cephalic chaetotaxy in Mendes (1982) show the shape of the anterior margin of the head and the arrangement of bristlecombs in Allacrotelsa kraepelini ( Escherich, 1905) (Lepismatinae) is most reminiscent of that of Anisolepisma , differing mainly in the absence of a gap in the chaetotaxy on the margin adjacent to the antennae, and the absence of 1+1 combs and scales on the clypeus. Further characters shared with this genus are discussed below.

Antennal sensillae. Anisolepisma , Heterolepisma , Allacrotelsa Silvestri, 1935 , Primacrotelsa and Acrotelsa have simple basiconic sensilla type B and type C on the annuli of the more distal intervals of the antennae. They lack specialized sensillae such as the branched asteriform sensory structures seen on the antennae of Lepismatinae ( Mendes, 1982; Molero-Baltanás et al., 2000) and the flattened circular poculiform sensillae seen in Hyperlepisma Silvestri, 1932 , Mormisma and Qantelsella Smith, 2015 . Simple sensillae are probably a plesiomorphic state with neither the Nicoletiidae , Heterolepisma , Anisolepisma nor the more ancient Lepismatinae (Allacrotelsa) having more derived antennal sensillae.

Sensillae of the maxillary palp. Anisolepisma and Primacrotelsa do not have a cylindrical sensilla at the apex of the most distal article of the maxillary palp nor do they have feathered papillae seen on the apical article of the maxillary palp of Heterolepisma , nor the asteriform sensillae seen in the Lepismatinae . They have only small basiconic sensillae types B and C. This state seems to be correlated with that of the antennal sensillae.

Papillae of the last article of the labial palp. Both Primacrotelsa and Anisolepisma have papillae of the “aufgelöst” type, which are only four in number and, at least in the case of Anisolepisma , arranged in a diamond pattern. The appearance of five papillae on just one of the two palps in the single specimen from Cobar (AMS K261044) ( Fig. 191 ) suggests that the diamond format could be an apomorphy for these genera, having evolved from the 3+2 arrangement by a fusion of the two papillae in the proximal row. The Ctenolepismatinae and Mirolepismatinae have several (three to more than ten) papillae in a single line while the Heterolepismatinae , Mirolepismatinae and Silvestrellatinae have five papillae in a 3+2 arrangement seen all subfamilies except the Ctenolepismatinae and Mirolepismatinae. The Tricholepidiidae have six papillae in two lines, the Maindroniidae have the 3+2 arrangement and the Nicoletiidae have six papillae arranged in three lines of three, two and one so it is likely that the 3+2 arrangement represents the plesiomorphic state for the Lepismatidae . The genera of the Acrotelsatinae display a wide range of states with the 3+2 arrangement in Desertinoma and Lepismina and five or six in a single row in Acrotelsa and Apteryskenoma .

Again it is interesting to note the similar character appearance in Allacrotelsa Silvestri, 1935 (Lepismatinae) with regards to the rather unusual shape of the ultimate article of the labial palp in Anisolepisma and other Acrotelsatinae ; but noting that Allacrotelsa has five papillae in two rows.

Notal chaetotaxy. Anisolepisma and Primacrotelsa lack a setal collar, a character shared with some other Acrotelsatinae ( Acrotelsa and Apteryskenoma ) however a partial collar covering 50–60% of the anterior margin is present in Desertinoma ( Kaplin, 1992) and in Lepismina there are three marginal tufts.A setal collar is absent from the Lepismatinae , the Silvestrellatinae and some Ctenolepismatinae . In contrast a setal collar appears on all Heterolepismatinae , Mirolepismatinae and some Ctenolepismatinae and strong macrochaetae are found along the anterior margin of the nota in the Tricholepidiidae , Maindroniidae (?) and most Nicoletiidae suggesting the presence of a setal collar character is plesiomorphic.

Anisolepisma , Acrotelsa and Primacrotelsa have 1+1 closed anterior tufts of radiating macrochaetae on the pronotum. Something similar is seen on the pronotum of Allacrotelsa kraepelini Escherich, 1905 (Lepismatinae) (as described in Wygodzinsky, 1942) but the tufts are located more laterally and appear to be in contact with the anterior margin. Such tufts are absent from all other genera of the Zygentoma ..

Notal trichobothria. Most genera of Zygentoma silverfish have two trichobothria (very long thin sensory hairs) on each side of the thoracic nota ( Mendes, 1986b). Anisolepisma has several very similar looking hairs on anterior lateral margins and, in the case of A. pigmentum , the posterior macrochaetae of the nota are replaced by trichobothria-like setae. These trichobothria-like setae are very long and thicker than usual (>500 µm and about 2.6 µm or greater in diameter) compared to those described in Mendes (1986b) which are up to 350 µm in length and only 1.4 µm in diameter. Mendes (2004) reports that the pronotum of Primacrotelsa also has very thin and long (trichobothria-like) macrochaetae laterally although the number is not specified. The presence of long thin trichobothria-like hairs instead of (or perhaps as well as) macrochaetae is reported by Stach (1935) in his redescription of Lepismina aurisetosa Wahlgren, 1806 (Acrotelsatinae) . In Anisolepisma these hairs look very much like typical trichobothria except for their extraordinary length but detailed examination of the insertion points and their greater thickness basally suggests that several of them are modified setae. The insertion does not appear to be the evenly rounded hole with dents on one edge as described by Kränzler & Larink (1980) but seem very similar to the insertion sockets of the macrochaetae (compare sockets in Fig. 16 ). This hypothesis is further supported by the single long thin trichobothrialike hairs on each of the posterior combs of the meso and meta nota of A. pigmentum whereas the other species have a single macrochaeta in these locations. Wygodzinsky (1942) illustrates similar supernumery trichobothria on the pronotum of Allacrotelsa kraepelini ( Escherich, 1905) (Lepismatinae) and Anallacrotelsa cricetophila Mendes, 1996 (Lepismatinae) also has three quite long trichobothria (c. 400 µm) on the pronotum.

The long thin trichobothria-like seta seen on the tibia of PIII in some species of some genera e.g., Acrotelsella (Ctenolepismatinae) , Heterolepisma (Heterolepismatinae) and Primacrotelsa (Acrotelsatinae) and Allacrotelsa (Lepismatinae) ( Wygodzinsky, 1961) but not yet seen in any species of Anisolepisma is also probably an example of the modification of a seta and such a plastic conversion in form, present in at least half the subfamilies of Lepismatidae , is probably a plesiomorphic character but at present provides little insight into the phylogeny of the family.

Urotergites I–VIII. Mendes (1982) considered the 3+3 comb arrangement to represent the most primitive condition. Reductions in the number of combs are common but more so on urotergites I and VIII. Most genera of the Acrotelsatinae have only 2+2 small combs (or single macrochaetae) on urotergite I except Desertinoma and Apteryskenoma where urotergite I is glabrous or has only 1+1 combs.

The chaetotaxy on the anterior urotergites does not appear to be very stable in Anisolepisma with regular instances of combs missing from one side. Mendes (1979) and Molero- Baltanás (2010) also report asymmetry and loss or addition of combs in some rare individual specimens of Ctenolepisma ciliata (Dufour, 1831) .

Two species of Anisolepisma have 3+3 combs on urotergite VIII while the other two have only 2+2 combs, the condition seen in all other Acrotelsatinae and all Heterolepismatinae .

Urotergite IX. Only one species of Anisolepisma ( A. pigmentum ) has chaetotaxy on urotergite IX in the form of infralateral setae. This is reminiscent of that found in species of Heterolepisma and the Nicoletiidae and therefore may be a plesiomorphic character.

Urotergite X. The shape of urotergite X is often useful at the genus and species level. The shape of this urotergite in Anisolepisma is typical of that found in species of Heterolepisma , being rounded with just 1+1 apical combs consisting of two or three macrochaetae, as well as a fringe of marginal setae. Most shapes of urotergite X are represented within the Acrotelsatinae (i.e. rounded in Anisolepisma , trapezoidal with concave posterior margin in Primacrotelsa and strongly triangular in Acrotelsa and Paracrotelsa ) and is therefore difficult to use as a character to determine the phylogeny of the subfamily.

Irish (1990) noted that the chaetotaxy of the tenth urotergite was a stronger character than its shape. Here again we see a diversity within the Acrotelsatinae with Anisolepisma and Primacrotelsa having 1+1 combs or single macrochaetae in the posterolateral corners while Acrotelsa and Paracrotelsa have several combs along each margin of their triangular tenth urotergites. Interestingly, species of Allacrotelsa Silvestri, 1935 (Lepismatinae) , including the fossil species Allacrotelsa dubia (Koch & Berendt, 1854) have a tenth urotergite very similar to Anisolepisma , in both shape and chaetotaxy.

Urosternites. Anisolepisma shares with Acrotelsa , Paracrotelsa , Primacrotelsa and Desertinoma , the presence of 2+2 combs on urosternites III–VIII. In most species of Anisolepisma there is also a small medial comb on urosternites I–VII (VIII in ♂). There is always a medial comb on urosternite I, something that is not found on other Acrotelsatinae but does occur in some species of other subfamilies. Lepismina is unique in that it is the only genus of Lepismatidae to have no urosternal chaetotaxy.

Irish (1990) considered the character state in Maindronia neotropicalis Wygodzinsky as representing the plesiomorphic state where most urosternites have 3+3 small combs. In this case the Acrotelsatinae would be closer to the ancestral state than any other subfamily where the combs have been reduced to 1+1 or 1+1+1.

Parameres. Males of Anisolepisma have tubuliform paramera similar to Acrotelsa . Those of the Heterolepismatinae are short and more bulbous, while those in the Lepismatinae are quite variable in their presentation. Paramera have been lost in the Ctenolepismatinae , Mirolepismatinae and Silvestrellatinae. They are present in the Tricholepidiidae and Nicoletiidae and Protrinemuridae as well as the Microcoryphia and should be seen as plesiomorphic.

Ovipositor. Females of Anisolepisma and Primacrotelsa have simple short ovipositors of the primary type which have only simple thin setae apically. Ovipositors in females of Acrotelsa , Desertinoma and Lepismina are also short but apically widened and with apical cuticular spines in addition to the simple setae. Females of Heterolepisma and Anallacrotelsa also have simple ovipositors of the primary type but those of Heterolepisma are very much longer, extending well beyond the apices of the inner coxal processes of urosternite IX. Ovipositors in the Nicoletiidae and Protrinemuridae are also of the primary type with simple setae and this form is generally considered as plesiomorphic. The development of cuticular spines within some genera of the Acrotelsatinae is a unique synapomorphy within the subfamily.

The position of Anisolepisma and the Acrotelsatinae within the Lepismatidae . The form of the thoracic sterna in the Acrotelsatinae would seem to be fundamentally different to the free sterna of the remaining subfamilies of the Lepismatidae . If it were to be considered as plesiomorphic state rather than an autapomorphy brought about through a reduction of the previously free sterna, then the Acrotelsatinae could be the most primitive subfamily of the Zygentoma . The subfamily possesses many primitive traits such as the presence of parameres, the primary ovipositor with simple setae, the lack of specialized sensillae on the antennae and the more numerous combs of the urosternites. Its dorsal abdominal chaetotaxy, including the presence of infralateral setae on IX, resembles that of Heterolepisma , which is considered by most workers as the most plesiomorphic. Other characters such as the occurrence of both smooth and pectinate macrochaetae, the diversity of shapes and chaetotaxy of urotergite X, do not shed any light on this question. An interpretation giving priority to the form of the thoracic sternites creates other difficulties such that the simple cephalic chaetotaxy of both the Heterolepismatinae and Lepismatinae would have to be considered as a regression from having bushes or else constitute a parallel development within the two lines of evolution with fundamentally different thoracic sternal arrangements.

Irish (1990) also commented on the difficulty of interpreting the thoracic sterna in the genera now included within the Acrotelsatinae , noting that reference to other families suggests the state with the coxa lying above the sterna as symplesiomorphic but he could not accept this on holomorphological grounds as it ran contrary to almost all other significant characters he used to create his phylogram.

Finally the number of uncommon characters shared by Anisolepisma and Allacrotelsa (Lepismatinae) is intriguing. They have similar cephalic chaetotaxy, similar unspecialized sensillae on the antennae, similar shaped labial palps, pronotal tufts, supernumery notal trichobothria and urotergite X. They differ dramatically however in the development of the thoracic sternites, but it raises the question of whether there could be an ancient relationship between these genera. Molecular data may eventually be a useful tool to help resolve this issue.

Zoogeography of the Acrotelsatinae . Discussions on the zoogeography of the Zygentoma must be considered as very hypothetical primarily because of the very uneven research activities on the order. While southern Europe, northern and southern Africa, parts of central Asia and parts of Australia have received moderate to strong attention, most of the rest of the world has been very poorly sampled judging from the literature. In addition, anthropophilic/ peridomestic species must be excluded from discussions due the presumed influence of human activities man on their wide distributions.

Being flightless the distribution is likely to be slower than for other groups although several endemic species are known from remote volcanic islands such as Hawaii. Distribution over water has therefore clearly occurred with a few Ctenolepismatinae and Nicoletiinae and several Heterolepismatinae .

The subfamily Acrotelsatinae is well known from the Mediterranean to central Asia (six species of Desertinoma is reported from Afghanistan, Mongolia, Turkmenistan and Uzbekistan, four species of Lepismina from Afghanistan, Egypt, Kuwait, Iran, Israel, Libya, Saudi Arabia, Syria and possibly Turkey? and the monotypic Primacrotelsa from Yemen). Another less well known group of genera occurs in the Australia /Pacific region with four species of Anisolepisma from Australia, one species of Paracrotelsa from Irian Jaya (eastern Indonesia) and a single species of the enigmatic Apteryskenoma from the New Hebrides ( Womersley, 1928). Given the extensive survey work of Irish in Southern Africa, it would seem that the subfamily does not occur there and it has not yet been reported from the Americas, although this could well be a sampling issue.

Silverfish, being soft bodied, do not preserve well so the fossil record is very sparse and placement of the species into subfamilies difficult. They belong to a very ancient group of insects with their origins in the Silurian or Devonian (450–400 Ma) with the Tricholepidiidae separating from the remaining Zygentoma in the late Triassic (214 Ma) based on molecular clocks ( Misof et al., 2014). At this time the continents may have been more or less united in a supercontinent. Mendes and Wunderlich (2013) summarize the fossil record to date. The oldest known Lepismatid fossils date to the Cretaceous (c. 110 Ma) a time when the continents were breaking apart. They are either unplaced (“gen. sp. from Araripe” from Brazilian sandstone) or placed within the Lepismatinae ( Myanmar amber). More recent Eocene amber specimens (50–38 Ma) from Baltic belong to the Lepismatinae and the extinct Lepidotrichidae . The first Ctenolepismatinae fossil ( Ctenolepisma electrans ) dates from the Oligocene (34–25 Ma) in Dominican amber.

It is therefore perhaps premature to explain the current distribution based on such limited data. It is possible that all Lepismatid subfamilies existed before the break-up of the continents and that the current distribution of the Acrotelsatinae represents the relic remains of a widespread subfamily, or could simply be a reflection on the incomplete collection efforts. For example, silverfish of the genus Lepisma were believed to be restricted to the Mediterranean region (except the peridomestic Lepisma saccharina ) however Smith (2015a) recently described a species collected within termite nests in Australia.

The Acrotelsatinae are showing a similar but wider distribution pattern, suggesting an earlier extensive distribution now surviving as relic populations in specialized habitats. The absence of the subfamily from southern Africa, which has been well sampled, and the Americas, which hasn’t, conflicts somewhat with any explanation of a distribution through Pangea as Australia / PNG was only joined to Eurasia via Africa and Antarctica in the Permian (c. 225 Ma). In contrast, the Heterolepismatinae have a more or less southern Gondwanan distribution, suggesting a more recent origin.

In summary, the genus Anisolepisma appears to be an endemic Australian genus, widespread through Australia in the drier regions. It belongs in the family Acrotelsatinae which has a somewhat disjunct distribution with representatives in PNG, the Solomon Islands, central Asia through the Middle East to southern and eastern Mediterranean countries.

ACKNOWLEDGMENTS. I would like to thank Dr Jürgen Deckert and Peter Schönefeld ( ZMB) for the loan of the portion of the holotype of Heterolepisma hartmeyeri preserved in alcohol and for permission to dissect and mount it onto slides, Prof. Antonio Garonna ( IEA) for the loan of the original slide mounted portion of the holotype, Dr David Britton (previously AMS) for facilitating the loan and access to the Australian Museum material, Dr John Irish for sharing his notes on the Australian silverfish, Dr. Susan Wright of the QM for the loan of this material, Ms Sue Lindsay (previously AMS) for the electron micrographs and Prof. Jonathon Major and his colleagues at Curtin University for the specimen from Barrow Island and for financial support towards the purchase of a microscope. I would especially like to acknowledge the input and encouragement of Dr Luis Mendes of the Instituto de Investigação Científica Tropical ( IICT), Jardim Botânico Tropical/Zoologia R. de Junquiera, 14, 1300-343, Lisbon, Portugal, who willingly shared his vast knowledge of the family. Finally I must thank Dr Rafael Molero-Baltanás of the University of Cordoba, for the useful comments provided when reviewing the manuscript .

Figures

Figure 190. Anisolepisma, schematic representation of mesothoracic sternum with sections at A, B and C shown on right.

Figure 191.Anisolepisma aquilonaridum n. sp. ex Cobar, aberrant labial palp.

Figures 16–17. Anisolepisma aquilonaridum n. sp. ex Bladensburg ♀ (16) pronotum, detail of insertions of trichobothria-like seta (tls) and submarginal macrochaetae (sm); (17) posterior corner of mesonotum showing long, thin trichobothria-like setae (tr0, tr1 and tr-2) and the insertion of a submarginal macrochaeta (m0).