CARPOSINA NEW SPECIES 11 (FIGS 2 E, 3E) Material examined: United States: Hawaiian Islands: Hawaii: Hawaii Volcanoes National Park, Desolation Trail, 929 m, N19.36880 W155.3674. ♂ (slide LB36♂). D Rubinoff & A Kawahara. UHIM. Remarks: This specimen has unique genitalia and a wing pattern unlike the other species near it in Figure 1, but the specimen is somewhat rubbed. Without additional material, we do not feel a full description is warranted at this time. The designation as new species 11 follows the sequence initiated by Zimmerman (1978). DISCUSSION Two new species, C. longignathosa sp. nov. and C. brevinotata sp. nov., from the Society Islands, French Polynesia, appear more closely related to the Asian C. sasakii Matsumura than the Hawaiian species (Fig. 1). Clarke (1971) named C. paracrinifera, a species from Rapa, for its superficial similarity to C. crinifera (Walsingham) from Hawaii. However, given the genetic distance and weak node support between the North and South Pacific species in this study, it is likely that Hawaiian Carposina derived from a northern temperate ancestor (Zimmerman, 1978), while the French Polynesia species appear to represent an independent incursion of the genus into the Pacific from the Austral��� Asian region. However, greater outgroup sampling is necessary to test this hypothesis. A similar pattern of multiple colonizations in the Pacific was also found in Tetragnatha spiders (Gillespie, 2002), Misumenops spiders (Garb, 2006), Ptycta bark lice (Bess et al., 2014) and Cydia moths (Oboyski, 2011). Unfortunately, there are very few well-resolved phylogenies for Polynesian arthropods that broadly sample Pacific Islands as well as potential mainland source populations to fully evaluate how widespread this pattern is. Three new species, Carposina urbanae sp. nov., C. gagneorum sp. nov. and C. new species 11, appear nested well within the Hawaiian clade, indicating they are part of the Hawaiian radiation and not recent immigrants. Carposina urbanae sp. nov. (host unknown) from Kauai appears weakly connected to an olivaceonitens clade that is distributed across the high islands feeding on Clermontia (Campanulaceae) and Pouteria (Sapotaceae). Although the olivaceonitens clade is well supported, the tentative placement of this species is likely to change with further sampling. Carposina gagneorum sp. nov. is known from only single male and single female specimens, with no sequence data. However, the wing pattern is so unique for Hawaiian Carposina (Fig. 2 C) that we are confident they represent a single species separate from C. crinifera and C. graminicolor with which it shares similar male genital morphology. C. new species 11 (host unknown) from Hawaii is in a moderately supported clade with C. graminis from Kauai which feeds on Metrosideros (Myrtaceae). The genetic and geographical distance between these specimens suggest other lineages within this clade exist on the intervening islands. The distribution and host-plant associations for Hawaiian Carposina are confusing at best (Table 1). Species descriptions (Meyrick, 1883, 1913; Walsingham, 1907) were based on short series (in some cases single specimens) of adult moths, largely collected by R. C. L. Perkins during the Fauna Hawaiiensis project (Perkins, 1913). Confusion was further compounded by the high degree of wing pattern polymorphisms in several Hawaiian microlepidoptera groups. And although male genital characters are particularly useful for Carposina, their widespread use in Lepidoptera taxonomy began after Walsingham and Meyrick���s work on Pacific Islands taxa. Larval host-plant records for several species were subsequently gained through extensive rearing efforts by O. H. Swezey (summarized in Swezey, 1954). However, Zimmerman (1978) questioned many of Swezey���s identifications and recognized Carposina new species 1 to 10 to account for discordant host and island records (Table 1). In particular, Zimmerman (1978) questioned records for C. olivaceonitens, which included plants in the distantly related families Campanulaceae and Sapotaceae. Our phylogeny shows two wellsupported clades of C. olivaceonitens that could represent cryptic species, or host races in the process of diverging. Moreover, polymorphism in this clade (compare Fig. 2 F and G) makes species assignment difficult based on superficial morphology. This uncertainty can only be resolved by comparing the morphology and molecules of specimens reared from each host across the archipelago. indicates host or island records from O. H. Swezey (1954) needing confirmation, according to Zimmerman (1978). All species described in the genus Heterocrossa by Walsingham, 1907, except for C. achroana, 1883; C. glauca, C. lacerate, C. saurates and C. benigna Meyrick, 1913; and C. gagneorum sp. nov. and C. urbanae sp. nov. Medeiros & Oboyski. New species 1���10 are those noted by Zimmerman). New species 11 (this paper). Presently, no new host associations are proposed, but some island records are confirmed or noted as new (Table 1). Carposina atronotata is reported from Maui; C. ferruginea (Walsingham), known only from Molokai, is reported from Maui; C. gemmata, known from Hawaii (and possibly Oahu), is reported from Kauai; and C. olivaceonitens, that Zimmerman (1978) restricted to Kauai, is confirmed on Maui and Hawaii. Our analyses support the monophyly of Hawaiian Carposina (Fig. 1). Using typical and accelerated mutation rates for Lepidoptera (Brower, 1994; Haines et al., 2014), our results predict a period of 2.23���8.51 Myr (95% HPD 1.59���11.75 Myr) since the arrival of Carposina in Hawaii. The current high islands were formed 0.5 Mya (Hawaii) to 5 Mya (Kauai) (Carson & Clague, 1995; Price & Clague, 2002), which places Carposina colonization sometime during the formation of Nihoa, Niihau, Kauai, Oahu or the Maui Nui complex. However, these preliminary findings are likely to change with further taxon sampling, additional molecular data and more refined estimates of mutation rates. Although the basal species in our limited sampling of the Hawaiian clade, C. semitogata, was collected from Kauai, the overall topology does not lend obvious support to a progression rule pattern of diversification (Funk & Wagner, 1995). Instead, subclades appear to include representatives feeding on the same host on both old and young islands. Similar patterns of diversification were shown for Hawaiian Cydia H��bner (Lepidoptera: Tortricidae), whereby jumps to new host genera in disparate subfamilies of Fabaceae were accompanied by filling those host niches across the archipelago (Oboyski, 2011), and for Nesophrosyne Kirkaldy (Hemiptera: Cicadellidae) with host jumps between plant families (Bennett & O���Grady, 2012, 2013). In this scenario, some species are likely to become paraphyletic as a result of differential dispersal between land masses ��� more isolated populations will develop evolutionary trajectories independent of their containing clade. This appears to be the case for C. olivaceonitens in the current study, which is rendered paraphyletic by C. gemmata and LB34 (a damaged specimen that we currently are not able to identify with certainty) (Fig. 1), both of which have distinctly different genital morphology from C. olivaceonitens. The Hawaiian Carposina clade is separated from outgroup taxa by a relatively long branch, while several interior branches have only modest support (Fig. 1). Several factors may contribute to this, including limited outgroup sampling, limited ingroup sampling, choice of genetic markers, a long period of isolation for the Hawaiian clade, extinction and/or accelerated evolutionary rates. As a result, long branches make Hawaiian Carposina difficult to place in the world fauna. Although extinction is difficult to account for in phylogenetic reconstruction (e.g. Morlon, Parsons & Plotkin, 2011), these other factors can be addressed directly with continued investigation. Carposina present an opportunity to test competing hypotheses about Hawaiian phylogeography and phyloecology. While several species are known for each island, host associations remain obscured for most (Table 1). Moreover, host/habitat loss, extinctions, climate change, and competition and predation from alien species are likely to hinder collection of essential ecological and evolutionary data (Medeiros et al., 2013). Therefore, identifying larval hosts, particularly from critically endangered habitats, and constructing a wellresolved phylogeny for the entire clade is the highest priority for this group., Published as part of Matthew J. Medeiros, Griffin L. Bianchi, Laurel R. Taschetta & Peter T. Oboyski, 2016, A review of Polynesian Carposina Herrich-Sch��ffer (Lepidoptera: Carposinidae), with descriptions of four new species, pp. 135-146 in Zoological Journal of the Linnean Society 177 on pages 142-144, DOI: 10.1111/zoj.12363, http://zenodo.org/record/270024