A review of Paratranes Zimmerman, 1994, Xanthorrhoea-associated weevils of the Tranes group (Coleoptera, Curculionidae, Molytinae), with description of a new species

Grasstrees (Xanthorrhoea) are a group of monocotyledonous plants endemic to Australia, with a characteristic crown of long and narrow leaves. In the present study, the grasstree-associated weevils of the genus Paratranes Zimmerman, 1994 are reviewed and two species are recognised: P. monopticus (Pascoe, 1870) and P. zimmermani sp. nov. A lectotype for Tranes xanthorrhoeae Lea, 1898, which was recently synonymised with P. monopticus, is herein designated, and the validity of this synonymy is confirmed. Descriptions and diagnoses of the species are supplemented with photographs of their habitus and salient structures as well as a map of their distributions.


Introduction Geographical distribution
The distribution map was generated using the spatial-analysis tool of the Atlas of Living Australia (http://www.ala.org.au) and is based on the label data of all examined weevil specimens and on Paratranes occurrence data in the ALA online database. Genus Paratranes Zimmerman, 1994 Figs 1-10 Paratranes Zimmerman, 1994: 695.

Remarks
Paratranes is here classified in the tribe Orthorhinini, following Anderson et al. (2018), who included the Tranes group in this tribe based on the strongly supported relationship between Tranes Schoenherr, 1843 and the orthorhinine genus Vanapa Pouillaude, 1915 found by Shin et al. (2017) in a robust phylogenetic analysis of 522 protein-coding genes. Without considering these results, Legalov (2018) recently proposed a separate tribe for the Tranes group, Tranini (incorrectly formed as 'Tranesini'; see Hsiao & Oberprieler (2020)), but his diagnosis of Tranini did not distinguish the group from Orthorhinini, and it further contained characters not present in all the included genera. Pending further and more detailed analysis of the phylogenetic relationships of the Tranes group, it is premature to afford it tribal status in Molytinae.
The relationships of Paratranes to the other genera of the Tranes group also remain uncertain for the moment. It agrees with Tranes, Miltotranes Zimmerman, 1994 andHoweotranes Zimmerman, 1994 in lacking sclerolepidia along its metanepisternal sutures, which are present in Demyrsus Pascoe, 1872and Siraton Hustache, 1934(Oberprieler & Caldara 2012Lyal 2014;Hsiao & Oberprieler 2020), but how closely related it may be to the cycad-associated genera, Tranes and Miltotranes, is as yet unknown. The similarity of the anchor-shaped sclerite of the endophallus in Paratranes and Tranes lyterioides (Pascoe, 1875) suggests a close relationship between these taxa and that Tranes in its current concept may be nonmonophyletic. The taxonomy of Tranes is currently under investigation (unpubl. data 2020).
Colour and veStiture. Body and legs shiny black, semilustrous; body, femora and tibiae sparsely covered with very short pale setae, setae on tibiae slightly longer than on body and femora; funicles, clubs and tarsi densely covered with yellowish setae, longer than short pale setae on body, femora and tibiae.
roStrum. Moderately long, ca 1.1-1.2 × as long as pronotum in both sexes, robust, faintly evenly curved ventrad, dorsoventrally flattened, slightly broadened apically in dorsal view, coarsely punctate dorsally, punctures very fine in distal half, proximal half with paired dorsomedian and dorsolateral carinae, the latter lower than the former.

Distribution
This species is disjunctly distributed in coastal regions from northern Queensland to central New South Wales (Sydney) and in southeastern South Australia and southwestern Western Australia (Fig. 12). Based on our survey of museum collections, the Queensland population of Paratranes monopticus is the largest and eastern Australia appears to be the main distribution range of this species.

Natural history
Paratranes monopticus has been reported occurring at the base of flower stalks and leaves and amongst young leaves of Xanthorrhoea (Froggatt 1896(Froggatt , 1907Mulder 1964;Hawkeswood 1985). Mulder (1964) indicated that the species can attack the green leaves of grasstrees at the base, and the presence of a brownish, mucous secretion at the leaf bases may be linked with the occurrence of this species, although the larvae have apparently not yet been found. According to label data, specimens can be attracted by mercury-vapour lamps, but this does not necessarily indicate a nocturnal behaviour and may just relate to specimens disturbed from grasstrees growing near the lights, as happens with Tranes weevils and cycads. Based on records by Hawkeswood (1985) and label data, P monopticus is associated with at least three species of grasstrees (X. australis R.Br., X. johnsonii A.T.Lee and X. preissii Endl.), but it probably occurs on others as well and is likely not host-specific.

Remarks
The taxonomy and nomenclature of P. monopticus is complicated, because the type series of the synonymic name xanthorrhoeae comprises two different species and its author (Lea 1898) did not validly (in the description) designate a primary type, so that all specimens of the type series are syntypes. Lea also did not specify the number of specimens on which he based the description of xanthorrhoeae, but he cited the specimens as having been collected at seven localities, as "Hab.-Darling Ranges, Mt. Barker, Bridgetown, Swan River, W.A.; Galston, Gosford, Sydney, N.S.W.". There are seven specimens from these locations in Lea's collection (in SAMA) and the ANIC (see Material examined), but there are two from Sydney and none from Bridgetown, and only the three specimens from WA (Darling Ranges, Mt Barker and Swan River) are labelled as 'Co-types' by Lea. A further specimen in his collection is labelled "xanthorrheae [sic] / Lea TYPE / Pinjarrah" and also carries a large label with Lea's handwriting "Tranes / xanthorrhoeae / W Australia" and "TYPE" in red lettering along its right side, as is typical of specimens that Lea regarded as the primary types of his species names. However, as Pinjarra is not mentioned as a locality in the description of Tranes xanthorrhoeae, it appears that Lea may not have had this specimen when he described the species and it cannot therefore be regarded as belonging to the type series and be treated or designated as the name-bearing type of Tranes xanthorrhoeae.
In his description of the species, Lea (1898) noted that some of his specimens were larger and wider (especially the elytra), had a longer and thinner rostrum, the glabrous region on the prosternum wider and less noticeable and, most notable, edentate femora, and he regarded these specimens as representing the female of his new species. Zimmerman later recognised that these so-called females actually represent a different species, and he noted this on a copy of Lea's description (kept among his notes on the Tranes group in the ANIC), and later he even regarded these specimens as representing a new genus "allied to Paratranes" (Zimmerman 1994). On his copies of the descriptions of Tranes monopticus and T. xanthorrhoeae he further wrote that he had borrowed the type specimen of monopticus from Pascoe's collection in the Natural History Museum in London in 1988 and, comparing it with the "male" types of Lea's xanthorrhoeae (in particular the specimen from Pinjarra, which he accepted as the "holotype"), he concluded that the types of monopticus and xanthorrhoeae represent the same species. According to his notes, the synonymy of these two names had in fact earlier been suggested to him by John Balfour-Brown (of the Natural History Museum, London) -although a specimen so labelled by Balfour-Brown, from the Stirling Range in WA, is in fact P. zimmermani sp. nov. (see there). Zimmerman did not, however, effect this synonymy in his description of Paratranes (Zimmerman 1994), and it was only published much later by Pullen et al. (2014) (although it was actually not possible to synonymise the names without designating a lectotype for xanthorrhoeae and fixing the name to one of the two species represented in Lea's type series).
From our thorough examination of all available material of Paratranes, including of the male genitalia of many specimens, we conclude that the type series of xanthorrhoeae indeed comprises two distinct species, as Zimmerman had recognised, and furthermore that the syntypes from the Darling Range in WA and Galston in NSW represent P. monopticus (as determined by comparison with the holotype of monopticus (Fig. 11A) and specimens "c.w.t." (compared with the type) of monopticus by Zimmerman) and that those from Mt Barker, Swan River, Gosford and Sydney belong to a different species. In order to fix the name xanthorrhoeae to one of these two species, we here designate the male syntype from the Darling Range (Fig. 11B), which agrees well with Lea's description, as the lectotype of xanthorrhoeae and the other six specimens of the type series as paralectotypes. This lectotype designation and comparison of the lectotype with the holotype of monopticus enable us to validate the synonymy of monopticus and xanthorrhoeae published by Pullen et al. (2014) and to describe the other species as new, here named P. zimmermani sp. nov. The paralectotypes of xanthorrhoeae from Mt Barker, Swan River, Gosford and Sydney are further designated as paratypes of P. zimmermani sp. nov. Specimens of P. monopticus from Western Australia and South Australia differ slightly from those from Queensland (the type locality) and New South Wales by having the apical margin of the genital sclerite of the endophallus medially not straight (Fig. 9A) but slightly emarginate (WA) or lobed (SA) (Fig. 9B-C). Based on the records available to us, the WA and SA populations also appear to be geographically separated from those in eastern Australia, but in the absence of significant morphological differences we interpret all populations to represent a single, somewhat variable species.

Diagnosis
The species is readily distinguishable from P. monopticus, as given in the diagnosis of the latter above.

Etymology
The specific epithet is dedicated to the late Dr Elwood Curtin Zimmerman, former weevil taxonomist at the ANIC who is renowned for his invaluable works on Australian weevils and who was the first to recognise the distinctiveness of this species among the type series of Tranes xanthorrhoeae. Colour and veStiture. Body and legs shiny black, semilustrous; body, femora and tibiae sparsely covered with very short pale setae, setae on tibiae slightly longer than on body and femora; funicles, clubs and tarsi densely covered with pale to yellowish setae, longer than short pale setae on body, femora and tibiae.

Holotype
roStrum. Moderately long, ca 1.3-1.4 × as long as pronotum in both sexes, robust, slightly curved ventrad, dorsoventrally flattened, slightly broadened apically in dorsal view, coarsely punctate dorsally, punctures very fine in distal half, proximal half with paired dorsomedian and dorsolateral carinae, the latter lower than the former. eyeS. Subcircular in outline, slightly convex but not protruding (Fig. 4C-D).

Distribution
Based on the records we have seen, this species has two disjunct populations, one in coastal regions around central New South Wales (Sydney) and another in southwestern Western Australia (Fig. 12). Most of our examined specimens were collected in Western Australia, indicating that the species is mainly distributed there.

Natural history
Like P. monopticus, this species appears to be associated with grasstrees, having been taken from the flower stalks and green caudex of regenerating plants in a burnt area of swamp during the daytime. Adults have also been observed feeding on flower stalks of grasstrees (Fig. 13). They are cautious and quickly dropped to the ground when disturbed (J. and F. Hort, pers. obs.). The only host recorded for this species (on labels) is Xanthorrhoea preissii in Western Australia, but specimens from New South Wales imply that the species is associated with other Xanthorrhoea species in the east, such as X. australis and X. johnsonii.

Remarks
The assessment of the taxonomic status of this species varied between being considered conspecific with (females of) P. monopticus by Lea (1898) (as Tranes xanthorrhoeae) and as representing a different genus "allied to Paratranes" by Zimmerman (1994). While its status as a species distinct from P. monopticus is clear from several morphological differences (in particular in the male genitalia), we do not regard these differences as significant enough to treat it as a genus distinct from Paratranes, especially in comparison with the differences between the other genera of the Tranes group. Zimmerman (1994) did not seem convinced of a separate generic status for it either, because he did not name such a genus in his key to the genera of the Tranes group, as he did with Paratranes and other new genera.
Specimens of P. zimmermani sp. nov. in New South Wales differ slightly from those at the type locality in Western Australia (Kalbarri) in their male genitalia, by having a wider and shallower apical opening of the median dorsal groove of the penis (Fig. 9E-F), but such a wide and shallow opening also occurs in specimens of populations further south in Western Australia (e.g., Collie), and in the absence of other morphological differences we regard this difference as intraspecific variation rather than signifying different species. However, the apparent widely disjunct distribution of P. zimmermani sp. nov. between south-eastern and south-western Australia is remarkable in view of the much wider and likely continuous range of P. monopticus (see Fig. 12).

Redefinition of Paratranes Zimmerman, 1994
Zimmerman (1994) excluded Tranes monopticus from Tranes and established a new genus, Paratranes, for it, differentiating it from all other genera of the Tranes group by its slender and depressed body, the forehead slightly narrower than the basal width of the rostrum, the antennae inserted at about the distal third of the rostrum, funicular segment 2 being distinctly shorter than 1 and segment 7 incorporated into the club, confluent procoxal cavities and nearly contiguous procoxae, distinctly sulcate and dentate femora, distinct protibial uncus, distal combs of meso-and metatibiae continuing around the tibial apices and extending to the middle or slightly before the middle of the tibiae and by the rounded dorso-apical corner of the metatibiae. Zimmerman's concept of Paratranes did, however, not include P. zimmermani sp. nov., which he deemed to represent a different genus based on its longer funicular segment 2 (subequal to 1), edentate femora and very small protibial uncus. Given that such small differences also occur in other weevil genera (e.g., a femoral tooth of varying size in Siraton), together with the great overall morphological and ecological similarities of P. monopticus and P. zimmermani sp. nov., we regard both these species as representing the same genus and hence here revise the concept of Paratranes to include both these species and a more comprehensive suite of their morphological characters, including those of the terminalia (see redescription of the genus above).

Distribution and intraspecific variation of Paratranes
The distribution of Paratranes is considerably wide and broadly congruent with that of Xanthorrhoea, except that it appears much more fragmented, with disjunct populations in Western Australia, South Australia and eastern Australia and none in central Australia (as known) and seemingly also none in Victoria and Tasmania. Despite the disjunct distributions of both species in western and eastern Australia, the morphological differences are very minor among their different populations (see details in the Remarks section of the species), and, in view of the distinct interspecific differences in other genera in the Tranes group (e.g., Tranes and Miltotranes), we recognise only two species in Paratranes. The very small morphological divergence between the eastern and western Australian populations of both species indicates that both were more widely and continuously distributed in Australia in the past and have developed their disjunct distribution relatively recently. Future studies using a molecular approach are needed to explore the evolutionary history of the genus and its species.
The morphological differences between P. monopticus and P. zimmermani sp. nov. are quite conspicuous, both in external and in genital characters, and, interestingly, they include the female terminalia, which are usually similar in the same genus in weevils. Especially noteworthy is the difference in the ovipositor, which is much more pronounced than in the other genera of the Tranes group, e.g., Demyrsus and Siraton (Hsiao & Oberprieler 2020), and suggests different modes of oviposition and hence of sites of larval development between the species, such that P. monopticus may lay its eggs into deeper-lying tissues and that its larvae develop in different plant parts than those of P. zimmermani sp. nov.. This may enable the two species to co-exist in sympatry and even on the same hostplants (i.e., character displacement; Brown & Wilson 1956), as appears to be the case in locations where both have been collected (Perth, Pinjarra, Mt Barker, Bremer Bay and Stirling Range in WA and Gosford and Sydney in NSW). Although their distribution ranges overlap in Western Australia and central New South Wales, it appears that P. zimmermani sp. nov. is more common in the former and P. monopticus in the latter region.