Taxonomic Revision of the Genus Miltotranes Zimmerman, 1994 (Coleoptera: Curculionidae: Molytinae), the Bowenia-Pollinating Cycad Weevils in Australia, with Description of a New Species and Implications for the Systematics of Bowenia

Simple Summary Miltotranes Zimmerman, 1994 is a genus of weevils pollinating Bowenia cycads belonging to two small endemic Australian species included in the IUCN Red List of Threatened Species and CITES Appendix II. We provide the first taxonomic revision of Miltotranes resulting in the identification of two previously described species and a newly recognised species, M. wilsoni sp. n. from the McIlwraith Range of the Cape York Peninsula. Morphological comparison reveals its affinity to Tranes, whose species are the pollinators of Macrozamia and Lepidozamia cycads. It appears that the association of Miltotranes with Bowenia may represent a secondary host switch in the Tranes group due to a closer relationship between Macrozamia, Lepidozamia and African Encephalartos than with Bowenia. The coincidence of the geographic ranges of Miltotranes weevils with distribution of their cycad hosts indicates that the isolated Bowenia population in the McIlwraith Range may represent a distinct, third species of Bowenia. Likely, the conspecificity of the Miltotranes weevils occurring in the Wet Tropics also suggests that several morphologically enigmatic localised populations represent B. spectabilis, confirming earlier botanical conclusions. The present study highlights the significance of systematic study of tightly plant-associated insects and its relevance for the taxonomy of their host plants. Abstract The Australian endemic weevils of the genus Miltotranes Zimmerman, 1994 (Curculionidae: Molytinae: Tranes group), comprising two species, M. prosternalis (Lea, 1929) and M. subopacus (Lea, 1929), are highly host-specific and the only known pollinators of Bowenia cycads, which comprise two CITES-protected species restricted to Tropical Queensland in Australia. In the present study, the taxonomy of Miltotranes is reviewed, a lectotype for the name Tranes prosternalis Lea, 1929 is designated and a new species associated with the Bowenia population in the McIlwraith Range is described as M. wilsoni sp. n. The descriptions and diagnoses of all species are supplemented with illustrations of their habitus and salient structures, and an identification key to all species and a distribution map are provided. Potential implications of the new species and of the taxonomy and biogeography of Miltotranes overall on the systematics and conservation of Bowenia are discussed.


Introduction
Miltotranes Zimmerman, 1994 (Curculionidae: Molytinae: Tranes group) is a genus of weevil endemic to north-eastern continental Australia, currently comprising two named

Material Examined
Types. Lectotype ( Figure

Diagnosis
Adults of this species are externally extremely similar to M. wilsoni but distinguishable from it by the following characters (states of M. wilsoni in parentheses): pronotum with a pair of triangular black marks on posterior margin and elytra with interval 1 alternating black and orange along basal half and a broken and mosaic median black macula ( Figure 1A,B) (pronotum unicolorous, without black marks, and elytra with interval 1 uniformly black along basal half and an entire median black macula; Figure 1E,F); vestiture on pronotum and elytra longer, dense, clustered in parts to somewhat obscure derm ( Figure 1A,B and Figure 4A) (shorter, sparsely distributed; Figure 1E,F and Figure 4C); pronotum broader, ca. 0.8-0.9× as broad as elytra, lateral margins distinctly rounded in male ( Figure 1A,B and Figure 4A) (narrower, ca. 0.7-0.8× as broad as elytra, lateral margins weakly rounded in male; Figure 1E,F and Figure 4C); protibiae thicker in male, ca. 6.0× as long as wide ( Figure 6A) (more elongate and slender in male, ca. 7.5× as wide; Figure 6C); penis more elongate, ca. 1.9-2.1× longer than wide ( Figure 9A,B,D,E) (penis thicker, ca. 1.6-1.7× longer than wide; Figure 9M,N,P,Q).
Colour and vestiture. Head dark red to black, antennae reddish brown, thorax orange to dark red, pronotum with a pair of triangular black marks at base, elytra orange to dark red, with black anterior margin and anterior half of elytral intervals 1 and an irregular black mark medially, abdomen dark red, coxae, trochanters, femora and tibiae black, tarsi reddish brown, semilustrous ( Figure 1A,B); body and legs covered with coarse, sublanceolate and subsquamiform, yellowish setae, clustered in some parts to somewhat obscure derm, especially on pronotum, prosternal elevated process in male, scutellar shield, intermesocoxal process in male and elytra, setae longer on pronotum, prosternal elevated process in male, intermesocoxal process in male and elytra, setae denser in margins of prosternal elevated process in male.

Distribution
Miltotranes prosternalis occurs in the coastal regions of far northern Queensland, recorded from the Endeavour River at Cooktown in the north to Townsville in the south ( Figure 14).

Natural History
Miltotranes prosternalis is exclusively associated with Bowenia spectabilis (except for the McIlwraith Range population, doubtfully regarded as B. spectabilis) and is the apparent sole pollinator of its host, developing in the male cones of the plants mainly from October to December. Its habits and interaction with B. spectabilis have been summarised by Hsiao & Oberprieler [6].

Remarks
Lea [24] described Tranes prosternalis based on specimens from the Endeavour River in the far north of Queensland. In his description he wrote that he had "… two specimens

Distribution
Miltotranes prosternalis occurs in the coastal regions of far northern Queensland, recorded from the Endeavour River at Cooktown in the north to Townsville in the south ( Figure 14).

Natural History
Miltotranes prosternalis is exclusively associated with Bowenia spectabilis (except for the McIlwraith Range population, doubtfully regarded as B. spectabilis) and is the apparent sole pollinator of its host, developing in the male cones of the plants mainly from October to December. Its habits and interaction with B. spectabilis have been summarised by Hsiao & Oberprieler [6].

Remarks
Lea [24] described Tranes prosternalis based on specimens from the Endeavour River in the far north of Queensland. In his description he wrote that he had " . . . two specimens before me", indicating that the type series only comprises two specimens, and in his collection (in SAMA) there is a male labelled "Type" and another male labelled "co-type". However, as he did not designate a primary (name-bearing) type specimen in his description, both his "type" and "co-type" specimens are syntypes of equal nomenclatural status. In order to fix the name prosternalis to a single, name-bearing type, we here designate the male syntype labelled as "type" (Figure 13A), which is well prepared and agrees well with Lea's description, as the lectotype of Tranes prosternalis and the "co-type" specimen as a paralectotype.
Insects 2022, 13, x FOR PEER REVIEW 18 of 27 before me", indicating that the type series only comprises two specimens, and in his collection (in SAMA) there is a male labelled "Type" and another male labelled "co-type". However, as he did not designate a primary (name-bearing) type specimen in his description, both his "type" and "co-type" specimens are syntypes of equal nomenclatural status. In order to fix the name prosternalis to a single, name-bearing type, we here designate the male syntype labelled as "type" (Figure 13A), which is well prepared and agrees well with Lea's description, as the lectotype of Tranes prosternalis and the "co-type" specimen as a paralectotype.  Specimens of M. prosternalis from north of Cairns differ slightly from those from south of Cairns by having the anterior margin of the rhombic sclerite of the endophallus broadly truncate and the apex narrowly rounded ( Figure 9A-C, see blue arrows, Figure 10A), whereas in specimens from south of Cairns the anterior margin is narrowly rounded and the apex is broadly rounded to subtruncate ( Figure 9D-F, see blue arrows, Figure 10B). However, the endophallic sclerite of some specimens is intermediate between these conditions, and in the absence of other significant morphological differences in both external and genital characters we interpret all populations as representing a single, somewhat variable species. More comprehensive study of specimens from all populations and the addition of genomic data should be able to refine the species delimitation. (Lea, 1929) ( Figure 1C,D, Figure 2C,D, Figure 3D, Figure 4B,E, Figure 5B,E, Figure 6B,E, Figure 7B, Figure 8B,E,H, Figure 9G-L, Figure 10C, Figure 11B,E, Figures 13B and 14 Type locality: Byfield, Queensland, Australia.
Colour and vestiture. Body and legs dark brown, without lustre ( Figure 1C,D); body and legs covered with coarse, sublanceolate and subsquamiform, yellowish setae, clustered in parts to somewhat obscure derm, especially on pronotum, prosternal elevated process in male, scutellar shield, intermesocoxal process in male and elytra, setae longer on pronotum, prosternal elevated process in male, intermesocoxal process in male and elytra, setae denser at margins of prosternal elevated process in male.

Distribution
Miltotranes subopacus only occurs in the Byfield area of central-eastern Queensland ( Figure 14).

Natural History
Miltotranes subopacus is a host-specific pollinator of Bowenia spectabilis. Its life history is seemingly similar to that of M. prosternalis, according to the literature [1,2,4,5].

Remarks
Lea [24] described Tranes subopacus based on a single specimen from Byfield in Central Queensland, explicitly stating that "…the type appears to be a female" and "…the type is probably a female". There is also only such a single specimen in his collection (in SAMA) ( Figure 13B), indicating that this specimen is the holotype.

Natural History
Miltotranes subopacus is a host-specific pollinator of Bowenia spectabilis. Its life history is seemingly similar to that of M. prosternalis, according to the literature [1,2,4,5].

Remarks
Lea [24] described Tranes subopacus based on a single specimen from Byfield in Central Queensland, explicitly stating that " . . . the type appears to be a female" and " . . . the type is probably a female". There is also only such a single specimen in his collection (in SAMA) ( Figure 13B), indicating that this specimen is the holotype. ( Figure 1E,F, Figure 2E,F, Figure 3E, Figure 4C,F, Figure 5C,F, Figure 6C,F, Figure 7C, Figure 8C,F,I, Figure 9M-R, Figure 10D, Figure  Colour and vestiture. Head dark red, rostrum black, antennae reddish brown, thorax orange to dark red, elytra orange to dark red with black anterior margin and anterior half of elytral interval 1 and an irregular black macula medially, abdomen dark red, coxae, trochanters, femora and tibiae black, tarsi reddish brown, semilustrous ( Figure 1E,F); body and legs covered with coarse, sublanceolate and subsquamiform, yellowish setae, clustered on prosternal elevated process in male and intermesocoxal process in male, somewhat obscuring derm, setae longer on pronotum, prosternal elevated process in male, intermesocoxal process in male and elytra, setae denser in margins of prosternal elevated process in male.

Derivation of Name
The species is named for Gary Whittaker Wilson, botanist at the Australian Tropical Herbarium at James Cook University, who undertook a significant major study of the pollination and systematics of Bowenia.

Distribution
Miltotranes wilsoni is known from only two localities in the Mcllwraith Range in northern Queensland ( Figure 14) and appears to be restricted to a small area east of the Mungkan Kandju National Park, where its Bowenia hostplant occurs.

Natural History
No specific information is available about the life history of the species, but it is presumably very similar to those of the other two Miltotranes species, especially that of M. prosternalis. Most of the specimens were collected from young expanding fronds or foliage at the end of June, which is the cooler season, rather than the summer coning period. This suggests that the species may overwinter in the adult stage, which agrees with a similar presumption made for M. prosternalis by Hsiao & Oberprieler [6].

Remarks
The taxonomic status of M. wilsoni as a species distinct from M. prosternalis was first noted by Elwood Zimmerman in 1994 (see Material Examined above), and Oberprieler later provisionally supported it (pers. comm. 2000 to G. Wilson [5]). These observations are confirmed in the present study, based on numerous differences from M. prosternalis in the characters of both external and genital structures. These differences are in line with species differentiations in other genera of the Tranes group [12,26].

Systematic Placement of Miltotranes and Evolution of Cycad Pollination in the Tranes Group
Miltotranes belongs to the systematically enigmatic Tranes group of genera [14], which has been recently placed in the tribe Orthorhinini [21], based on the phylogenomic study of Shin et al. [27], or in its own tribe, Tranini, based on morphological characteristics [22]. Miltotranes is one of two genera of the Tranes group pollinating their cycad hosts, and its species are the sole obligate pollinators of Bowenia. Despite the as yet unresolved phylogenetic relationships in the Tranes group, Miltotranes has a close affinity to Tranes based on morphological characters including colour pattern (dark brown to reddish brown; Figure 1), position of the scrobes (parallel to rostrum in lateral view; Figure 2) and sexually dimorphic prosternum (prominently protuberant in the males; Figure 4) and protibiae (with a large brush in the males; Figure 6), which suggests that the obligate cycad pollination systems of Tranes and Miltotranes may represent a single evolutionary event in the Tranes group. The cycad host genera of Tranes, Macrozamia and Lepidozamia, are not closely related to Bowenia [28][29][30], and Macrozamia and Lepidozamia are also not each other's closest relatives (the latter being more closely related to the African genus Encephalartos), so that evolutionary host shifts are indicated to have occurred both in Tranes and in the Tranes group. Given the distant and evidently much older phylogenetic origin of Bowenia [28][29][30], Oberprieler [11] earlier already suggested that the association of Miltotranes with Bowenia represents a secondary colonisation of this genus by the Tranes group. A phylogenetic analysis using genomic data is currently in preparation to resolve the systematic position and phylogenetic relationships of Miltotranes and the evolution of its cycad pollination.

Implications of Miltotranes Systematics on Bowenia Taxonomy
The genus Bowenia is endemic to tropical Queensland and currently contains two described extant species, B. spectabilis, occurring mainly in the Wet Tropics bioregion of northeast Queensland (from Cape Melville in the north to Cardwell in the south), and B. serrulata, restricted to the Byfield area of the Central Queensland Coast bioregion (north of Rockhampton). Whereas the species limits of B. serrulata have not been contentious, the taxonomic status of a number of populations of B. spectabilis has been uncertain, in particular an isolated northern one in the McIlwraith Range of the Cape York Peninsula but also another seemingly isolated one in the Starcke National Park north of Cooktown and two on the Atherton Tablelands, one at Kuranda and the other at Tinaroo. All have been treated as "putative B. spectabilis" by Wilson [5], and the Tinaroo population has even been regarded as a distinct species [8,31]. As Miltotranes weevils are known from all these populations and act as the pollinators of the plants, their species identities have implications for the taxonomic status of these cycad populations.
The localised northern Bowenia population in the McIlwraith Range is geographically separated from the B. spectabilis populations further south by the Normanby Gap, also known as the Laura Gap or Laura Basin, a drier, alluvial-lowland river catchment area that separates the Iron Range and McIlwraith Range rainforest area of the Cape York Peninsula from the Wet Tropics [32]. This geographical disjunction is believed to be of Late Miocene age [5], but its exact age and progression over time requires further investigation. The McIlwraith Bowenia population is usually treated as belonging to B. spectabilis [7,33], but Wilson [5], having had insufficient samples of it available for his morphological study, only provisionally regarded it as belonging to this species. Kokubugata et al. [34,35], using a cytotaxonomic approach to address the taxonomic issues of Bowenia, were also unable to include this population in their karyotype analyses, and no rigid genetic analysis has been conducted of it to date. Its taxonomic status therefore remains somewhat uncertain.
Wilson [5] also regarded the Bowenia population in the Starcke National Park as an isolated one, separated from populations further south by the Black Mountain Divide or Corridor (located just north of Cairns [32]). There are, however, other Bowenia populations north of this corridor (e.g., at Cape Tribulation, Cooktown and Cape Melville), and it does not appear to be a barrier of significance for Bowenia.
The status of the Bowenia populations at Kuranda and Tinaroo has been uncertain due to their seemingly intermediate morphological characteristics between B. serrulata (pinnules with serrate margin; caudex large and branched) and B. spectabilis (pinnules with entire margin; caudex small and sparsely branched). The Tinaroo population has been treated as a distinct, undescribed species [8,31], whereas Norstog & Nicholls [36] regarded it only as an infraspecific variety of B. spectabilis. Hill & Osborne [7] found that serrate pinnules occur in all populations of B. spectabilis, and Wilson [5] concluded from his comprehensive morphological study that pinnule and caudex structure are phenotypically plastic characters determined by ecological factors (especially temperature) and that the Kuranda and Tinaroo populations are likely only ecotypes rather than subspecies of B. spectabilis. This conclusion is supported by the karyotype analyses of Kokubugata et al. [34,35], which revealed that these two populations have the same number of median-centromeric chromosomes as B. spectabilis and are thus cytotaxonomically closer to this species than to B. serrulata.
From our study of Miltotranes weevils from these three populations and also from the Endeavour River at Cooktown (ca. 60 km south of the Starcke population) we conclude that the specimens from the Kuranda, Tinaroo and Cooktown populations of Bowenia are conspecific and represent M. prosternalis (the Endeavour River being its type locality) and also that specimens from all populations of B. spectabilis further south (from Cape Tribulation to Townsville) represent the same species. Only the specimens from the isolated northern McIlwraith population represent a different species, here described as M. wilsoni and being distinguishable from M. prosternalis on several morphological features, including of the genitalia.
The conspecificity of Miltotranes weevils from the Kuranda, Tinaroo and Cooktown populations (and thus likely also the Starcke population) of Bowenia indicates that these plant populations are also conspecific and all represent B. spectabilis. In contrast, the recognition of a different species of Miltotranes occurring in the McIlwraith population of Bowenia (M. wilsoni) supports the notion that this population may not represent B. spectabilis [5] but a different species as well. It further suggests that this cycad population and its weevil pollinator may have become isolated together from their congeneric populations further south by the development of the Normanby Gap and that their concomitant differentiation may represent a case of co-speciation mediated by vicariance. Likewise, the (evidently older) evolutionary divergence of the species pair of M. prosternalis and M. wilsoni from M. subopacus in the south may be concomitant with that of B. spectabilis from B. serrulata and could also be due to vicariance (the development of the dry-savanna Burdekin and Saint Lawrence Gaps [32]). However, analysis of the phylogeny of the Tranes group and circumspect research into the timing of the diversification events in both the weevils and the plants is required to explore these scenarios.

Conclusions
This first systematic revision of the weevil genus Miltotranes, whose species are the sole known pollinators of the small endemic Australian cycad genus Bowenia, results in the identification and delimitation of three species, the previously described M. prosternalis and M. subopacus and a newly recognised and described species, M. wilsoni. Several morphological characters of Miltotranes shared with Tranes indicate that these two genera are closely related (evidently sister taxa) and, given the widespread association of Tranes with the cycad genus Macrozamia but also with the two species of Lepidozamia (which is more closely related to the African Encephalartos), it appears that the association of Miltotranes with Bowenia may represent an evolutionary host shift in the Tranes group from Macrozamia to Bowenia.
The ranges of the three Miltotranes species coincide well with those of their cycad hosts, M. subopacus only occurring on the southern species B. serrulata, M. prosternalis on the northern B. spectabilis and M. wilsoni on the northern-most and isolated Bowenia population in the McIlwraith Range of the Cape York Peninsula, which is thus indicated to represent a distinct, third species of Bowenia. Similarly, the conspecificity of the Miltotranes weevils occurring in the Cooktown/Starcke, Kuranda and Tinaroo populations with M. prosternalis suggest that these populations also represent B. spectabilis, confirming earlier botanical conclusions that the latter two only represent morphologically slightly different ecotypes [5,7]. Furthermore, the taxonomic and geographical congruence between the three species of Miltotranes and their Bowenia hosts suggests that their evolutionary differentiations may also be concomitant.
The This study lastly highlights the relevance of systematic study of tightly plant-associated insects for the taxonomy of their hosts. As the only known pollinators of Bowenia cycads, Miltotranes weevils are evidently instrumental in maintaining the reproductive integrity of their hosts, and their species identities are therefore also highly relevant for the species identities of their hosts. The case of Bowenia cycads and their Miltotranes pollinators emphasises the need for more comprehensive and congruent taxonomic and phylogenetic studies of the plants and their associated weevils.