Chrysorthenches callibrya

Sohn, Jae-Cheon, Kobayashi, Shigeki & Yoshiyasu, Yutaka, 2020, Beyond Wallace: a new lineage of Chrysorthenches (Lepidoptera: Yponomeutoidea: Glyphipterigidae) reveals a journey tracking its host-plants, Podocarpus (Pinopsida: Podocarpaceae), Zoological Journal of the Linnean Society 190, pp. 709-736 : 730-734

Chrysorthenches callibrya was originally combined with Diathryptica (type species: D. proterva ) by

Turner (1923). In the present study, we examined the type species of Diathryptica and found three major differences: the chorda of the forewing, the division of the male valva and the sclerotization of the ductus bursae in the female genitalia. The C. callibrya species-group includes C. callibrya and two allied new species. These three species share the diagnostic characteristics of Chrysorthenches as proposed by Dugdale (1996), including, in both sexes, the lack of a dense awning of scales on the antennal scape, the long apical segment of the labial palpus, the lack of socii and gnathos, the presence of a V-shaped mesal lobe on the male sternum VIII, and the absence of a sclerotized costa on the membranous distal part of the valva in the male genitalia. In addition, our examination of the larvae of C. muraseae revealed four characteristics associating it with Chrysorthenches : (1) seta V1 on the meso- and metathorax present at the ventral edge of each coxa, (2) a spiracle on abdominal segment VIII on the SD1 pinaculum, posterior to the SD1 seta, (3) abdominal segments VII and VIII with one SV seta and (4) abdominal segment IX with setae D1 and D2 on the same broad pinaculum. These shared characteristics justify a generic transfer of Diathryptica callibrya to Chrysorthenches and the assignment of the two new species mentioned in this study to the same genus. The former finding was also consistent with our COI phylogeny result (Supporting Information, Fig. S1).

Our cladistic analysis supported the monophyly of the C. callibrya based on one synapomorphy, the entirely sclerotized ductus bursae (24: 2 in Table 1), and three homoplastic characters: the bifid uncus (6:0 in Table 1), the ventrally arising ductus seminalis (22: 1 in Table 1) and an enception of the ductus seminalis on the corpus bursae (26: 1 in Table 1). The fast and slow character optimizations in the cladistics study recognized two additional synapomorphies: the presence of a short chorda on the forewing (4: 1 in Table 1) and larval thoracic L1 and L2 setae arising on separate pinacula (30: 1 in Table 1) (Supporting Information, Fig. S2). The larval features of C. muraseae differed from those of the C. argentea and the C. porphyritis species-groups in the presence of thin SD1 setae on the mesothorax and the abdominal segment VIII, and an SV setal group on abdominal segments I and II bisetose. These characters can also serve as synapomorphies of the C. callibrya species-group, but more information on larval characters is needed to confirm their phylogenetic value.

Our cladogram ( Fig. 15A) for 12 species of Chrysorthenches differed from that presented by Dugdale (1996). The most critical dissimilarity was in the position of C. polita (Philpott, 1918) , which was placed in the C. argentea species-group in our study but in the C. porphyritis species-group by Dugdale (1996). The positions of C. glypharcha (Meyrick, 1919) and C. phyllocladi Dugale, 1996 were also discordant between the two studies. All these differences may be the result of our modifications and additions of characteristics to the data matrix presented by Dugdale (1996). Thus, we analysed another data matrix (J. -C. Sohn, unpublished) that included the same characterset and coding as that used by Dugdale (1996). The analysis still resulted in a different cladogram from that described by Dugdale (1996), possibly due to the additions of C. callibrya and C. muraseae . In fact, the relationships among the species-groups in Chrysorthenches are ambiguous, because those depend on the characteristics of the ductus bursae and the ductus seminalis, which are membranous and thus versatile. In accordance with this ambiguity, the backbone relationships of Chrysorthenches were poorly supported by the results of our study (1–2 range in Bremer supports: Fig. 16). Thus, the phylogenetic relationships within Chrysorthenches need further attention.

PODOCARPACEAE ASSOCIATION

Chrysorthenches is distinguished from other lineages of the Orthenches -group by a trophic association with conifers. Larval host-plants are known for only nine of the 12 species in Chrysorthenches and for two congeners whose larval hosts were inferred from vegetation in which the adult moths were observed ( Dugdale, 1996). These records indicate that all species of Chrysorthenches , except for C. virgata (Philpott, 1920) , which feeds on Cupressaceae , and C. smaragdina , whose larval hosts are unknown, are associated with Podocarpaceae . Among the Podocarpaceae , the majority of Chrysorthenches species utilize the largest genus of that family, Podocarpus . The members of the C. callibrya species-group seem also to be associated with Podocarpus . The host-plants of C. muraseae are reported from the present study, while an association of C. callibrya with Podocarpus could be inferred from Dugdale’s (1996) field observation at Charlotte Pass, New South Wales, Australia. It is likely that C. smaragdina also feeds on Podocarpus , given the host associations of two other species in the same species-group and the occurrence of Podocarpus in Thailand.

Chrysorthenches utilize seven genera of Podocarpaceae and those genera are not necessarily closely related ( Fig. 16). This may suggest that most, if not all, of their host associations have resulted from sequential colonization, not co-evolution, as Dugdale (1996) has already pointed out. Podocarpaceae-feeding species of Chrysorthenches were associated with only one or two plant genera, while Podocarpus was the genus on which most species of Chrysorthenches feed ( Fig. 16). The C. callibrya species-group, earliest diverging in Chrysorthenches , also uses Podocarpus as a larval host. Taken together, these observations may suggest that ancestral Chrysorthenches colonized Podocarpus and later shifted to other podocarp genera. Among the Podocarpus -feeding Chrysorthenches , the New Zealand species are associated exclusively with the Australis subclade in the subgenus Podocarpus . On the other hand, the host-plants of the C. callibrya species-group belong to two Podocarpus subgenera ( Fig. 17).

The trophic associations between Chrysorthenches and Podocarpaceae are noteworthy, given the limited numbers of insects that utilize these plants. Other than Chrysorthenches , few lepidopterans feed on Podocarpaceae and they include macroheterocerans such as Erebidae ( Lymantriinae ), Geometridae and Lasiocampidae and some microlepidopterans ( Tortricidae , Gracillariidae , Lecithoceridae and Pyralidae ) worldwide ( Okelo, 1972; Singh et al., 1978; Oku, 1979; Murase, 2005; Costa & Boscardin, 2014; Liu et al., 2018). Most of these moths are generalist larval feeders, but Makivora hagiyai Oku, 1979 ( Tortricidae ) is a specialist on Podocarpus . Chrysorthenches are comparable to Milionia Walker, 1854 ( Geometridae ) in that all or nearly all members are associated with Podocarpaceae . Yasui (2001) found that Milionia were able to sequester the phytochemicals of Podocarpus for protection against predatory stink bugs. Like Milionia , the adults of Chrysorthenches are colourful, but it is unknown if they can also take advantage of a chemical defence system.

BIOGEOGRAPHY AND HOST–PLANT TRACKING

The high trophic fidelity of Chrysorthenches with Podocarpaceae hints that the radiation of Chrysorthenches may have been affected by the host-plants. Recent studies have suggested that Podocarpaceae originated in Gondwana during the Triassic–Jurassic periods ( Biffin et al., 2011; Rothwell et al., 2012; Escapa et al., 2013). Furthermore, Lu et al. (2014) estimated the origination of the extant podocarp genera to be in the Early Cretaceous. The largest genus of Podocarpaceae , Podocarpus , is one of the representative groups in the Antarctic flora that originated in the cold and wet climate of southern Gondwana ( Page, 1990; Mill, 2003). Quiroga et al. (2016) dated the divergence of two subgenera of Podocarpus as within the Late Cretaceous–Early Palaeogene. The surviving lineages of Podocarpaceae radiated into the tropical regions, not earlier than 30 million years ago or the Late Eocene ( Cernusak et al., 2011).

Extant species of Chrysorthenches occur only in New Zealand, eastern Australia, Tasmania, South-East Asia and Japan ( Fig. 17). The highest diversity among the Chrysorthenches species (eight of 13 total species) is observed in New Zealand. This, from the viewpoint of traditional dispersal biogeography, would suggest that New Zealand is the centre of origin for Chrysorthenches . However, the result of our DIVA analysis ( Fig. 15B) favoured a broad distribution of ancestral Chrysorthenches that subsequently split according to palaeogeographical changes. Regarding their presence in Tasmania and Australia, the Chrysorthenches – conifer association may pre-date the opening of the Tasman Sea, which began about 80 million years ago ( Molnar et al., 1975). The distributional range of Chrysorthenches occupies only a small proportion of the distribution of Podocarpaceae . This difference may indicate that Chrysorthenches evolved long after the Podocarpaceae radiation that pre-dated the splitting of the Gondwanan subcontinents. Another, less plausible, explanation would be the extensive extinction of Chrysorthenches , except in the Australasian region. Direct evidence for this hypothesis does not exist to our knowledge, but a leaf-mine trace left by a larva that was presumed to belong to Chrysorthenches in Wilf et al. (2005) may indicate their existence on other Gondwana subcontinents until at least 52 million years ago.

The Chrysorthenches callibrya species-group differs from the other two congeneric species-groups as the distribution of the former is not restricted to the Australasian region ( Fig. 17). Moreover, three species of the species-group have disjunctive distributions: eastern Australia for C. callibrya , Thailand for C. smaragdina and Japan for C. muraseae . Our cladogram recovered this species-group as the earliest diverging with respect to other Chrysorthenches ( Fig. 17). This poses questions, such as: why have no members of the species-group been reported from west and north Australia, Papua New Guinea and other islands spanning the Wallacea zone? Further inventory of Chrysorthenches in the Australasian region may help to fill these gaps.

The collective distributional range of the C. callibrya species-group corresponds to that of the island arc system connecting Australia and East Asia. This island system has facilitated trans-Wallacean radiation in many organisms through faunal exchanges between Australia and Asia during 15–20 million years ago ( Sklenarova et al., 2013). The C. callibrya species-group may have followed this route, but their direction was distinctively northward, as reconstructed from our DIVA analysis ( Fig. 15B). Most biogeographic studies in Australia and Asia have suggested southward dispersals ( De Jong, 2001), although there are a few examples indicating northward radiations; for example, the plant family Proteaceae ( Truswell et al., 1987) and skipper butterflies of the Taractroceragroup ( De Jong, 2001).

The Chrysorthenches callibrya species-group may have evolved as a result of their colonization of Podocarpus in the Cenozoic Era. Given the distributions of their sister groups, it would seem plausible that ancestors of the C. callibrya species-group evolved as one of the lineages resulting from the radiation of Chrysorthenches before the separation of New Zealand and Australia in the Middle–Late Cretaceous. In such a scenario, this lineage would have dispersed toward South-East Asia, as represented by the occurrence of C. smaragdina in Thailand. Such an event could have happened only after the Podocarpus species had radiated into tropical Asia in the Late Eocene, about 30 million years ago ( Cernusak et al., 2011) and after the first opportunity for faunal exchange between Australia and Asia approximately 25 million years ago ( De Jong, 2001). Consistent with these requirements, the emergence of host-plant clades for the C. callibrya species-group were estimated to occur in the Eocene–Oligocene periods ( Quiroga et al., 2016). Pre-existence of host-plants was a prerequisite for the C. callibrya species-group crossing the Wallacea zone, like other lepidopteran examples ( Beck et al., 2006). Chrysorthenches muraseae may represent the currently understood terminus of the sequential radiation for the C. callibrya species-group. It is known that the island arc system crossing the Wallacea zone reached Japan in the Late Miocene or in the Pliocene ( De Jong, 2001).

Recent advances in biogeography allow the differentiation of dispersal from vicariance and the determination of possible divergence dates using molecular data ( Trewick, 2000; Trewick & Wallis, 2001; Waters & Roy, 2004; de Queiroz, 2005). This type of approach would be necessary to better explain the curious distribution of Chrysorthenches .