Nonmarine Mollusks from Archaeological Sites on Mo‘orea, Society Islands, French Polynesia, with Descriptions of Four New Species of Recently Extinct Land Snails (Gastropoda: Pulmonata: Endodontidae)

Carl C. Christensen, Jennifer G. Kahn, Patrick V. Kirch

Pacific Science, Volume 72, Number 1, January 2018, pp. 95-123 (Article)

Published by University of Hawai'i Press

For additional information about this article https://muse.jhu.edu/article/683172

[ This content has been declared free to read by the pubisher during the COVID-19 pandemic. ] Nonmarine Mollusks from Archaeological Sites on Mo‘orea, Society Islands, French Polynesia, with Descriptions of Four New Species of Recently Extinct Land Snails (Gastropoda: Pulmonata: Endodontidae)1

Carl C. Christensen,2,5 Jennifer G. Kahn,3 and Patrick V. Kirch4

Abstract: Nonmarine mollusks recovered during archaeological excavations on the island of Mo‘orea, Society Islands, French Polynesia, were analyzed as part of a multidisciplinary study of anthropogenic environmental change. Records of now-extinct taxa in dated archaeological contexts were combined with historic collection data from the 1830s to the present to determine the chronology of extinction of the 10 species of land snails of the family Endodontidae that for- merly inhabited the island. One species known only from a stratum antedating human settlement on the island and three known only archaeologically but from imprecisely dated strata were certainly extinct by the late nineteenth century but may have disappeared earlier. One species collected in 1838 was extinct by the late nineteenth century, and all of the described endodontid species present in the mid- to late nineteenth century are undoubtedly now extinct because none was collected by the 1934 Mangarevan Expedition or by subsequent collectors. Only a single unidentified living endodontid has been observed on Mo‘orea since the nineteenth century. Additional extinctions or extirpations have oc- curred among the Helicarionidae, terrestrial Assimineidae, and probably also in the Helicinidae. Four new species of Endodontidae are described: Libera kondoi, Minidonta opunohua, Nesodiscus nummus, and N. cookei. Libera jacquinoti, described in 1850 from poorly localized material and until now not collected subsequently, is shown to have inhabited Mo‘orea.

Recent archaeological and paleoenvi- ments and native biota both during the pre- ronmental studies have shown that human historic period following initial human settle- ­activities in the Society Islands (Central East ment of these islands ca. AD 900 – 1000 and Polynesia) and elsewhere in East Polynesia following the advent of European influence have had substantial effects on island environ- and commerce in the mid- to late eighteenth century AD (Anderson 2002, Kennet et al. 2006, Prebble and Wilmshurst 2008, Kirch et al. 2010, McWethy et al. 2010, Newell 1 This project was funded by National Science Foun- 2010, Conte and Molle 2014, Kahn et al. dation collaborative grants BCS-1029765 to J.G.K. and 2015). The investigation reported here of land BCS-1030049 to P.V.K. Manuscript accepted 13 July 2017. snails obtained during archaeological excava- 2 Natural Sciences, Bernice P. Bishop Museum, 1525 tions on the island of Mo‘orea, in the Society Bernice Street, Honolulu, Hawai‘i 96817. Islands of French Polynesia, was undertaken 3 Department of Anthropology, College of William as part of a multidisciplinary study of anthro- and Mary, Williamsburg, Virginia 23185. pogenic environmental change. 4 Department of Anthropology, University of Cali- fornia, Berkeley, California 94720. The terrestrial mollusks of the Society Is- 5 Corresponding author (email: carl@bishopmuseum lands, including Mo‘orea, came to the atten- .org). tion of the scientific community early in the nineteenth century and by 1900 were rela­ tively well studied. During the first half of the Pacific Science (2018), vol. 72, no. 1:95 – 123 doi:10.2984/72.1.7 century, the island’s land snails were collected © 2018 by University of Hawai‘i Press by members of two major exploring expedi- All rights reserved tions, that of the French vessels Astrolabe and

95 96 PACIFIC SCIENCE · January 2018

Zelee (1837 – 1840) and the United States Ex- materials and methods ploring Expedition (1838 – 1842). In the sec- ond half of the nineteenth century the pri­ Coastal Transects, Coring, and Test Excavations mary source of information on the island’s As part of a larger project investigating coastal land snails was through the work of the natu- habitations and geomorphology on the island ralist Andrew Garrett, who visited the Society of Mo‘orea, several coastal localities and allu- Islands at various times from 1857 onward vial bottomland contexts were investigated and resided on Huahine from 1870 until his ( Kahn et al. 2015). These include five coastal death in 1887 (Thomas 1979). Garrett was a localities (sites ScMo-341, -342, -343, and careful collector and published a comprehen- -349, where we completed transect coring sive review of the Islands’ terrestrial mollusks and test excavations, and Lake Temae, where (Garrett 1884); his work continues to be held we retrieved a deep pollen core) (Figure 1A, in high regard (Solem 1976, Thomas 1979). B). These locales represent three contrastive Although some taxa restricted to high eleva- depositional environments. Sites -341, -342, tions escaped his notice (Solem 1976, 1983, and -343 are low-lying coastal plains situated Gargominy 2008), we believe any taxa inhab- along the northern shore on the headlands iting low-elevation sites, such as those under between ‘Opunohu Bay and Paopao (Cook’s) study here, can safely be regarded as having Bay. They are situated across from barrier gone extinct by the 1880s if they were not reefs and major passes where there was likely ­encountered by Garrett. This includes Orobo­ to be substantial input of calcareous sediments phana cf. maugeriae, Libera jacquinoti, and the of marine origin. Site -349 is located along four endodontid species here described as the eastern shore on the headlands of Vaiare new. Henry Crampton intensively studied the Bay. It is also situated across from a barrier island’s Partulidae, and members of Bishop reef and a major pass but along a part of the Museum’s 1934 Mangarevan Expedition made island that receives less rainfall. The Lake extensive collections on the island (Cooke Temae pollen core sample was retrieved from 1935). More recently, the island’s highly en- the northeastern point of the island. This dangered Partulidae have been studied (e.g., brackish lake is now separated from the sea Murray and Clark 1980, Johnson et al. 1993), by a beach ridge of calcareous sediments. and within the last 10 yr the living nonmarine The lake is at the base of Temae Valley. The mollusk fauna of the island has been exten- main sedimentary input was from inland ter- sively sampled in the course of the Moorea rigenous sources following deforestation and Biocode Project (Moorea Biocode 2016). clearance of the adjacent hillslopes. This is the first study of land snails from pale- Abbreviations used in site descriptions are ontological or archaeological sites on Mo‘orea as follows: AH, augur hole; b.d., below datum; and thus provides the first available infor­ cmbs, centimeters below surface; L, Layer; mation about species that had gone extinct obj., object; spm., specimen; spms., speci- before the activities of nineteenth-century mens; and TP, test pit. naturalists. No comprehensive review of the island’s nonmarine mollusks has appeared Excavation and Sampling Methods since Garrett’s time, but major monographs on several important families have appeared Transects were laid out at sites from the base based on collections up to and including those of the colluvial slopes to the lagoon shore. of the Mangarevan Expedition: Baker (1940) Coring with a bucket auger proceeded at 10 on the Helicarionidae, Cooke and Kondo to 20 m intervals along each transect. Sedi- (1961) on the Achatinellidae, and Solem (1976, ments were screened through 3.175 mm 1983) on the Endodontidae and Charopidae. (1/8 inch) and 1.588 mm (1/16 inch) mesh; all Several other families, especially the Heli­ artifacts and faunal materials were quantified cinidae, Assimineidae, and Succineidae, re- on standardized forms. Based on the auger- main poorly known and are in need of further ing data, we developed predictions of areas study. with the densest subsurface cultural deposits, Figure 1. (A) Map of the South Pacific Ocean showing location of the Society Islands; B( ) Map showing location of the island of Mo‘orea (inset) and location of the sites excavated. 98 PACIFIC SCIENCE · January 2018 which were then opened up for detailed inves- introduced commensal species, Allopeas gracile tigation by excavation of 1 m by 1 m test pits ( Hutton, 1834) (21 lots, 556 specimens). A or by larger block excavations and trenches detailed analysis of the taxa represented, the via backhoe excavation. These test pits and stratigraphic distribution of extinct species backhoe trenches allowed us to view broader and those aliens introduced during the pre- stratigraphic profiles and to conduct addi­ contact and modern eras, and descriptions of tional pollen and sediment sampling. From four new species of Endodontidae are pre- these, bulk sediment samples were retained sented here. for extraction of land snails. The bulk sam- ples were floated and then wet screened and systematic review picked through to retrieve land snail samples for analysis. Snails were recovered from sedi- Family Neritidae ment samples by wet screening or flotation in the field, followed by hand picking in the Clithon sp. laboratory with the aid of a stereomicro- material: Vaipahu, ScMo-343: L II, scope. Comparisons were made with material TP1-B2, obj. 9 (2 spms.); L II, TP1-B3, obj. 8 in the malacological collections of the Ber- (3 spms.); L II, TP2-A4, obj. 6 (2 spms.); nice P. Bishop Museum, Honolulu. Speci- L IIIa, TP2-B1, obj. 6 (3 spms.); L IIIa, TP2- mens are deposited in the collections of the B2, obj. 7 (2 spms.). Bishop Museum ( bpbm) and the Muséum remarks: The freshwater Neritidae of ­National d’Histoire Naturelle, Paris (mnhn). French Polynesia have been reviewed by bpbm catalog numbers are cited in the species Starmü hlner (1976), Haynes (1988, 2001, descriptions. 2005), and Pointier and Marquet (1990). The specimens collected during this study are weathered or eroded and cannot be identi- Stratigraphic Descriptions and Radiocarbon fied with certainty but most closely resemble Dating Clithon oualaniensis as discussed in Starmü hlner Radiocarbon dates and stratigraphic descrip- (1976) and the species identified as Clithon“ tions from all sites (except -349) have been (Clithon) sp. (?cf. chlorostomus, Broderip, published in full elsewhere ( Kahn et al. 2015). 1832)” and as “Clithon chlorostoma (Broderip, Age ranges for newly dated deposits at ScMo- 1832)” by Starmü hlner (1976) and Haynes 341 and summaries of the deposits at all (2001), respectively. sites containing land snails and short strati- graphic descriptions to place the land snail – Family Helicinidae bearing deposits in context are presented in the ­Appendix. The taxonomy of Pacific island Helicinidae is confused at the generic level and chaotic at the specific level; the latest monographs results ( Wagner 1905, 1907 – 1911) are long out of Thirty-six species of nonmarine mollusks date. Richling (2009, 2011) and Richling and were recovered during this study, represented Bouchet (2013) recently assigned to Sturanya by 220 lots and 3,351 individual specimens. New Caledonian, Hawaiian, and Mangarevan Strict quantitative analysis is not possible be- species formerly assigned to the genera Orobo­ cause collecting methods were not consis- phana and Pleuropoma; Richling and Bouchet tent throughout the study ( both flotation (2013) erected a new genus, Nesiocina, for and wet screening were used), but the three ­certain other helicinids occurring on islands most abundant species were Assiminea parvula from the Cook, Society, and Austral Islands Mousson, 1865 (20 lots; 1,358 specimens), the eastward to the Gambier Islands, Pitcairn, aquatic Tarebia granifera (Lamarck, 1822) and Henderson. The smaller helicinid found (1 lot, 825 specimens), and a prehistorically on Mo‘orea during the current study is here Nonmarine Mollusks from Mo‘orea, Society Islands · Christensen et al. 99 assigned to Sturanya, although with some Family Hydrocenidae doubt because the anatomical features said to distinguish Sturanya from Nesiocina are un- Georissa striata (Pease, 1871) available for study in this subfossil material. material: Vaipahu, ScMo-343: TP2-A4, We follow Wagner (1907 – 1911) in placing obj. 7 (1 spm.). Garrett’s Helicina maugeriae in the genus remarks: Garrett (1884) stated that this Orobophana, recognizing that further ana­ was the most abundant of the three species of tomical study, not possible with the material Chondrella (= Georissa) he found to occur in available here, may result in a reassignment. the Society Islands. This species is widely dis- Garrett (1884) reported the following taxa tributed in the southern Cook Islands, the So- (all assigned by him to the genus Helicina) to ciety Islands, Austral Islands, Makatea in the occur on Mo‘orea: minuta Sowerby, 1842; Tuamotu Islands, and the Marquesas Islands flavescens­ Pease, 1865; rustica Pfeiffer, 1851; (Brook 2010 and references cited therein; inconspicua Pfeiffer, 1848; and faba “Pease” Gargominy and Fontaine 2015); archaeologi- Garrett, 1884; seven additional species, in- cally, it has been reported from Rarotonga cluding maugeriae Gray, 1825, were listed as and Mitiaro in the Cook Islands (Brook 2010, occurring elsewhere in the Society Islands. Brook et al. 2010). It survives on Mo‘orea (Moorea Biocode 2016). Orobophana sp. cf. maugeriae (Gray, 1825) material: Pihaena, ScMo-342: L VI, Family Assimineidae TP2 – B4, obj. 3 (1 spm.). remarks: A single broken specimen was Assiminea parvula Mousson, 1865 obtained of a large-shelled helicinid with a material: Pihaena, ScMo-342: L VIII, strongly angulate periphery that resembles TP 2-B3, obj. 6. 384 – 401 cm b.d. (1 spm.). O. maugeriae, a species not previously re­ Vaipahu, ScMo-343: L II, TP1-B1, obj. 6 ported to occur on Mo‘orea and known only (42 spms.); L II, TP1-B1, obj. 7 (82 spms.); from Ra‘iätea and Taha‘a (Garrett 1884). The L II, TP1-B2, obj. 3 (222 spms.); L II, TP1- level at which this specimen was found dates B2, obj. 9 (203 spms.); L II, TP1-B3, obj. 7 to the late pre-Polynesian period (2,400 – (4 spms.); L II, TP1-B3, obj. 8 (262 spms.); 4,600 yr before present). L II, TP1-B3, obj. 10 (78 spms.); L II, TP2- A4, obj. 6 (36 spms.); L II, TP2-A4, obj. 7 Sturanya sp. (20 spms.); L IIIa, TP2-B1, obj. 6 (212 spms.); material: Vaipahu, ScMo-343: L II, L IIIa, TP2-B1, obj. 7 (124 spms.); L IIIa, TP1-B1, obj. 7 (4 spms.); L II, TP1-B3, obj. 8 TP2-B2, obj. 7 (34 spms.); L IIIa, TP2-B2, (2 spms.); L II, TP2-A4, obj. 7 (3 spms.); obj. 8 (4 spms.). Teavaro, ScMo-349: AH4, L IIIa, TP2-B1, obj. 6 (2 spms.); L IIIa, TP2- obj. 1, 60 – 90 cm b.s. (6 spms.); AH4, obj. 1, B1, obj. 7 (10 spms.); L IIIa, TP2-B2, obj. 8 60 – 110 cm b.s. (1 spm.); AH4, obj. 1, 120 – (1 spm.). 150 cm b.s. (2 spms.); AH5, obj. 1, 90 – 120 cm remarks: The material studied here in- b.s. (13 spms.); AH5, obj. 1, 120 – 150 cm b.s. cludes a number of specimens of a small un- (3 spms.); AH8, obj. 1, 0 – 30 cm b.s. (9 spms.). identified helicinid with a rounded periphery. remarks: This species, including A. nitida The confused taxonomy of the group pre- (Pease, 1865), a synonym according to Cowie cludes more precise identification, but this (1997b), is widely distributed in Oceania from is probably one of the species recorded by the Philippines and Micronesia eastward to Garrett (1884) as inhabiting the island. This Hawai‘i and French Polynesia (Abbott 1958); species was present on the island in 1934 (e.g., it is extant on Mo‘orea (Moorea Biocode bpbm 150954), but no more recent records 2016). In Hawai‘i it is said to be a semiaquatic of helicinids are listed in the database of the strandline species that lives under rubble at Moorea Biocode Project (Moorea Biocode and near the shoreline ( Kay 1978), but Brook 2016). (2010:179) reported that on Rarotonga, Cook 100 PACIFIC SCIENCE · January 2018

Islands, A. parvula was “fully terrestrial and remarks: Garrett (1884) reported the widely distributed across the island.” As ­occurrence of two species of Taheitia in the Brook remarked, it appears that additional re- Society Islands, T. porrecta (Gould, 1846), and search will be necessary to determine whether T. pallida (Pease, 1867); only the latter has “A. parvula” as currently understood includes been reported from Mo‘orea. A few speci- more than a single species, a concern also mens are represented in the material studied ­expressed by Abbott (1958). Accordingly it here. would be premature to draw ecological infer- ences from the occurrence of this species in Family Thiaridae the material studied here. Melanoides tuberculata (Mü ller, 1774) Omphalotropis sp. material: Vaipahu, ScMo-343: AH9, material: Vaipahu, ScMo-343: L I, obj. 2, 0 – 30 cm b.s. (1 spm.); L I, TP2-A1, TP1-A1, obj. 2 (2 spms.); L I, TP1-A2, obj. 2 obj. 2 (2 spms.); L IIIa, TP2-B1, obj. 6 (1 spm.); L I, TP1-A3, obj. 1 (1 spm.); L II, (1 spm.). TP1-B1, obj. 6 (2 spms.); L II, TP1-B1, obj. 7 remarks: Melanoides tuberculata is a glob- (9 spms.); L II, TP1-B2, obj. 3 (2 spms.); L II, ally invasive fresh- and brackish-water snail TP1-B2, obj. 9 (1 spm.); L II, TP1-B3, obj. 8 that is now widely distributed in Oceania (2 spms.); L II, TP1-B3, obj. 10 (1 spm.); L II, (Starmü hlner 1976, 1993, Pointier and Mar- TP2-A4, obj. 6 (2 spms.); L II, TP2-A4, obj. 7 quet 1990, Cowie 1997a, Haynes 2001); it (5 spms.); L IIIa, TP2-B1, obj. 7 (20 spms.); currently inhabits Mo‘orea (Moorea Biocode L IIIa, TP2-B2, obj. 6 (12 spms.); L IIIa, 2016). It has a fossil record in Africa and In- TP2-B2, obj. 7 (1 spm.); L IIIa, TP2-B2, obj. donesia extending back to the Miocene and 8 (1 spm.). Pliocene, respectively (Oostingh 1935, van remarks: Four species of Omphalotropis Benthem Jutting 1937, Ladd 1972 and works have been reported from Mo‘orea: O. huahei­ cited therein; van Damme 1984, van Damme nensis (Pfeiffer, 1854), O. oblonga (Pfeiffer, and Pickford 2003, Joordens et al. 2009). 1854), O. scitula (Gould, 1847), and O. terebra­ ­Szabó (2009) reported M. tuberculata to be lis (Gould, 1847); an additional two species present in pre-contact (pre-European con- occur elsewhere in the Society Islands (Gar- tact) sites in Fiji. Bandel and Kowalke (1997), rett 1884). Little has been published on Soci- Athens and Ward (2006), and Athens et al. ety Island Omphalotropidinae since Garrett’s (2007) have suggested that it may be a pre- work, and species-level identifications will contact Polynesian introduction, and Kirch ­remain uncertain until the group is revised. et al. (2017) have demonstrated its pre-­ A number of specimens of an unidentified contact presence on Mangaia in the Cook species of Omphalotropis were recovered dur- ­Islands. Work in progress confirms a pre- ing the excavations reported here at levels contact presence on Ra‘iätea in the Society dating to the period AD 1521 to the present. Islands (C.C.C. and J.G.K., in prep.) and No Omphalotropis were collected in 1934 by in the Hawaiian Islands (C.C.C. and J. S. Bishop Museum’s Mangarevan Expedition, ­Athens, in prep.). The species’ presence in L nor are any current records listed in the data- IIIa of the ScMo-343 site (AD 1521 – modern, base of the Moorea Biocode (2016). most likely AD 1521 – AD 1767) is sugges- tive of a pre-contact presence on Mo‘orea Family Truncatellidae but is not conclusive because of dating uncertainties.­ Taheitia pallida (Pease, 1867) material: Vaipahu, ScMo-343: L II, Tarebia granifera (Lamarck, 1822) TP1-B1, obj. 7 (3 spms.); L II, TP1-B2, obj. 3 material: Lake Temae: Core 1, L I, (1 spm.); L II, TP1-B2, obj. 9 (1 spm.); L II, 749 – 753 cm b.s. (825 spms.). TP1-B3, obj. 8 (1 spm.); L II, TP2-A4, obj. 6 remarks: Like Melanoides tuberculata, (1 spm.); L IIIa, TP2-B1, obj. 7 (1 spm.). Tarebia granifera (also known as Thiara Nonmarine Mollusks from Mo‘orea, Society Islands · Christensen et al. 101 granifera) is a globally invasive aquatic snail to permit identification beyond the generic that now inhabits numerous Pacific islands level. (Starmü hlner 1976, Pointier and Marquet 1990, Cowie 1997a, Haynes 2001), including Family Achatinellidae Mo‘orea (Moorea Biocode 2016). Thiara granifera is native to island Southeast Asia, Elasmias sp. based on its fossil occurrence in Indonesia material: Vaipahu, ScMo-343: L II, (Oostingh 1935, van Benthem Jutting 1937). TP2-A4, obj. 7 (1 spm.). Like M. tuberculata it was a pre-contact resi- remarks: Two species of Elasmias, E. ap­ dent of the Hawaiian Islands (C.C.C. and J. S. ertum (Pease, 1864) and E. peaseanum, inhabit Athens, in prep.). The specimens recovered in Mo‘orea (Garrett 1884, Moorea Biocode this study are modern (200 yr or less before 2016). The single specimen obtained during present), so provide little information on the this study cannot be assigned to either species date of the species’ arrival on Mo‘orea. with certainty. Cooke and Kondo (1961) sug- gested that at least one member of this genus, Family Ellobiidae E. apertum, owes its wide distribution (from the Kermadec Islands and Rotuma eastward Laemodonta monilifera ( H. Adams & A. Adams,­ to Mangareva, Tubuai, and Makatea) to pre- 1854) historic translocation by the Polynesians, but material: Vaipahu, ScMo-343: L II, Brook (2010) suggested that modern com- TP1-B1, obj. 6 (1 spm.); L II, TP1-B3, obj. 8 merce has played a role as well. Because of (2 spms.); L II, TP2-A4, obj. 6 (1 spm.); its uncertain date (post AD 1521) the single L IIIa, TP2-B1, obj. 6 (4 spms.); L IIIa, TP2- specimen obtained yields no information on B2, obj. 7 (1 spm.). this question. remarks: Garrett (1884) reported Pleco­ trema mordax Dohrn, 1859, from Tahiti; he Lamellidea micropleura Cooke & Kondo, 1961 had no specimens from Mo‘orea. Plecotrema material: Vaipahu, ScMo-343: L II, mordax is a synonym of P. monilifera H. ­Adams TP1-B1, obj. 6 (1 spm.). & A. Adams, 1854, a species occurring remarks: This species occurs on Mo‘orea throughout the Indo-Pacific region from the (Cooke and Kondo 1961, Moorea Biocode Red Sea and East Africa eastward to Rapa and 2016) and has also been reported from Pit- Mangareva in French Polynesia ( Hubendick cairn and Henderson Islands (Preece 1995) 1956). The species, now assigned to the genus and from Mitiaro and other locations in the Laemodonta Philippi, 1846, is a strandline southern Cook Islands (Brook et al. 2010). It dweller and occurs “[o]n rocks in a tidal flat is presumably native to Mo‘orea and the other area, just above the normal high tide mark.” islands that it inhabits. (Raven and Vermeulen 2007:53). Lamellidea oblonga (Pease, 1865) Melampus sp. material: Vaipahu, ScMo-343: L II, material: Vaipahu, ScMo-343: L IIIa, TP1-B1, obj. 6 (4 spms.); L II, TP2-A4, obj. 7 TP2-B2, obj. 7 (1 spm.). (2 spms.); L IIIa, TP2-B1, obj. 6 (1 spm.); remarks: Garrett (1884) reported four L IIIa, TP2-B2, obj. 8 (2 spms.). species of Melampus from Mo‘orea: M. luteus remarks: Lamellidea oblonga occurs (Quoy & Gaimard, 1832), M. caffer ( Kü ster, throughout a triangular region anchored in 1843), M. fasciatus (Deshayes, 1830), and M. the southwest by Fiji and the Ellice Islands, in philippii ( Kü ster, 1845) (a synonym of M. cas­ the north by the Hawaiian Islands, and in the taneus Mü hlfeldt, 1816, according to Rehder southeast by the Pitcairn Islands, including 1980); all now occur on the island, as does Mo‘orea (Cooke and Kondo 1961, Moorea M. bidentatus Say, 1822 (Moorea Biocode Biocode 2016). “There is little doubt that 2016). The single specimen recovered dur- the wide distribution of this species is due, ing the excavations reported here is too worn mainly to the frequent voyages of the Polyne- 102 PACIFIC SCIENCE · January 2018 sians, who transported food plants on their Mo‘orea (Moorea Biocode 2016). It is be- travels, especially between islands only a few lieved to owe much of its wide distribution to hundred miles apart.” Cooke and Kondo inadvertent transport during the prehistoric (1961:201). Subsequent archaeological studies­ period (Cooke and Kondo 1961, Christensen reviewed in Christensen and Weisler (2013) and Kirch 1981, Christensen and Weisler have confirmed its pre-contact presence in 2013), and it is frequently encountered in pre- the Cook Islands (Allen 1992, 1997, 1998; contact archaeological sites within this region ­Allen and Christensen 1992, Walter 1998, (Christensen and Weisler 2013 and citations Brook 2010, Brook et al. 2010), Huahine and therein). As with L. oblonga, all dated speci- Tahiti in the Society Islands (Sinoto 1983, mens recovered here date from the period AD Orliac 1997), the Gambier Islands ( Howard 1521 to the present. and Kirch 2004, Conte and Kirch 2008), the Marquesas Islands ( Kirch 1973; Rollett 1992, Lamellidea spp. 1998), Henderson Island (Preece 1998), and material: Vaipahu, ScMo-343: L II, the Hawaiian Islands (Dixon et al. 1997, Bur- TP1-B1, obj. 6 (6 spms.); L II, TP1-B1, obj. 7 ney 2002). Lamellidea oblonga was not present (3 spms.); L II, TP1-B3, obj. 10 (2 spms.); in the Hawaiian Islands before initial human L II, TP2-A4, obj. 6 (9 spms.); L II, TP2-A4, settlement (Cooke and Kondo 1961), so it is obj. 7 (1 spm.); L IIIa, TP2-B1, obj. 6 not indigenous but is a pre-contact intro­ (9 spms.); L IIIa, TP2-B1, obj. 7 (3 spms.); duction (Christensen and Weisler 2013). It L IIIa, TP2-B2, obj. 7 (1 spm.); L IIIa, TP2- must be native to some island or islands with- B2, obj. 8 (1 spm.). in its current range in southern Polynesia, but remarks: Immature, broken, or other- its point of origin within that area is as yet wise nondiagnostic specimens of Lamellidea unknown. This is a lowland species often as- are listed as Lamellidea spp. sociated with economic plants and places of human habitation; it may be abundant under Tornatellides oblongus oblongus (Anton, 1839) dead leaves from just inland of the shore to material: Vaipahu, ScMo-343: L II, elevations of 100 m or so and rarely if ever is TP2-A4, obj. 6 (1 frag.); L II, TP2-A4, obj. 7 found in undisturbed native forest (Cooke (1 frag.). and Kondo 1961). All dated specimens ob- remarks: Tornatellides oblongus oblongus tained here from Mo‘orea are from post-AD was sparsely represented in this material; two 1521 levels. basal fragments were the only specimens ob- tained. Tornatellides o. oblongus occurs in the Lamellidea pusilla (Gould, 1847) Cook, Society, Austral, Marquesas, and Gam- material: Vaipahu, ScMo-343; L II, bier Islands and on Pitcairn Island, with an TP1-B2, obj. 9 (3 spms.); L II, TP1-B3, obj. 8 endemic subspecies, T. o. parvulus Cooke & (3 spms.); L II, TP2-A4, obj. 6 (4 spms.); L II, Kondo, 1961, on Henderson Island (Cooke TP2-A4, obj. 7 (1 spm.); L IIIa, TP2-B1, obj. and Kondo 1961, Preece 1995); it was col- 6 (1 spm.); L IIIa, TP2-B1, obj. 7 (1 spm.); lected on Mo‘orea in 1934 (e.g., bpbm 150286) L IIIa, TP2-B1, obj. 6 (3 spms.); L IIIa, TP2- but has not been encountered in recent sur- B2, obj. 7 (2 spms.). veys (Moorea Biocode 2016). “From the simi- remarks: Lamellidea pusilla is widely dis- larity of shells from different islands and from tributed in Oceania, occurring from the Mar- different island groups, some of which are shall and Solomon Islands eastward to the separated by as much as 800 miles, it seems Tuamotu and Gambier Islands (Cooke and that the distribution from its original home Kondo 1961, Kondo 1975); the western limits has been comparatively recent, probably since remain undefined because of the uncertain the advent of the Polynesians. It was un­ status of related and possibly conspecific taxa doubtedly a lowland species disseminated ac- occurring in Indonesia, the Bismarck Archi- cidentally with food plants” and “is especially pelago, and the Mariana Islands ( Kondo 1975, abundant on dead coconut, breadfruit, and Christensen and Weisler 2013). It is extant on banana leaves in plantations; on dead leaves in Nonmarine Mollusks from Mo‘orea, Society Islands · Christensen et al. 103 rather open spaces or in dense thickets of in the eastern Solomon Islands, Marshall Hibiscus­ tiliaceus and Pandanus odoratissimus; ­Islands, Fiji, Samoa, Tonga, Society Islands, and under native trees and plants.” Cooke and Cook Islands, Marquesas Islands, and Hen- Kondo (1961:251). Tornatellides o. oblongus has derson Island (Christensen and Weisler 2013 been reported from pre-contact archaeo­ and references therein). Specimens obtained logical sites in the Cook (Allen 1992, Allen in the study reported here date from the pe- and Christensen 1992, Walter 1998), Society riod AD 1521 to present but cannot conclu- (Sinoto 1983), and Marquesas Islands (Rolett sively be shown to be of pre-contact age. 1998), but no pre-Polynesian records are available to identify the islands to which Gastrocopta servilis (Gould, 1843) the species is indigenous. Preece (1998) re- material: Vaipahu, ScMo-343: L II, ported that T. o. parvulus was present in pre- TP1-B1, obj. 7 (1 spm.); L II, TP1-B3, obj. 8 Polynesian levels of sites on Henderson (3 spms.). ­Island, indicating that that subspecies occurs remarks: This species, native to the New there as a native, not as a result of prehis­ World tropics, is a modern introduction to toric or modern human commerce. All speci- the Pacific. It was first reported from New mens recovered during the study reported Guinea in 1883 (Solem 1989, Brooke 2010) here dated from the period AD 1521 to the and from Hawai‘i in 1892 [Ancey (1892), as present. Pupa lyonsiana]; it is now widely distributed in tropical Oceania (Cowie 1997a, 2001a; Brook Family Vertiginidae 2010, and references cited therein). Gastro­ copta servilis was unknown to Garrett (1884) Gastrocopta pediculus (Shuttleworth, 1852) but had become established on Mo‘orea by material: Vaipahu, ScMo-343: L II, 1934 (e.g., bpbm 150353). The few specimens TP1-B1, obj. 6 (5 spms.); L II, TP1-B1, obj. 7 of G. servilis recovered in the study reported (12 spms.); L II, TP1-B2, obj. 3 (21 spms.); here are undated but are clearly of modern L II, TP1-B2, obj. 9 (4 spms.); L II, TP1-B3, origin; the species undoubtedly currently in- obj. 8 (19 spms.); L II, TP1-B3, obj. 10 (8 habits Mo‘orea, although it is not listed in a spms.); L II, TP2-A4, obj. 6 (16 spms.); L II, recent survey of the island’s fauna (Moorea TP2-A4, obj. 7 (8 spms.); L IIIa, TP2-B1, obj. Biocode 2016). 6 (31 spms.); L IIIa, TP2-B1, obj. 7 (22 spms.); L IIIa, TP2-B2, obj. 7 (4 spms.); L IIIa, TP2- Nesopupa sp. cf. tantilla (Gould, 1847) B2, obj. 8 (1 spm.). material: Vaipahu, ScMo-343: L II, remarks: Gastrocopta pediculus occurs TP1-B1, obj. 6 (3 spms.); L II, TP1-B1, obj. 7 from the Philippines, Indonesia, and Australia (2 spms.); LII, TP1-B2, obj. 3 (1 spm.); L IIIa, eastward through Melanesia, Micronesia, and TP2-B1, obj. 6 (3 spms.); L IIIa, TP2-B1, obj. Polynesia (Pilsbry 1916 – 1918; Solem 1959, 7 (1 spm.). 1989) [but see Pokryszko (1996), arguing that remarks: Garrett (1884) reported the oc- the taxon described as Vertigo pediculus samo­ currence of Nesopupa tantilla (Gould, 1847) on ensis Mousson, 1865, is a species distinct from several of the Society Islands, though not on G. pediculus, a conclusion not adopted here]. Mo‘orea; he considered Vertigo armata Pease, It was present on Mo‘orea in 1934 (e.g., 1871, to be a synonym. Pilsbry and Cooke bpbm 150352) but has not been recorded (1918 – 1920) regarded armata as distinct and in recent surveys (Moorea Biocode 2016). added N. pleurophora (Shuttleworth, 1852) to Pilsbry (1916 – 1918) suggested that the broad the Society Islands list, though again without ­distribution of G. pediculus was due to the any specific reference to Mo‘orea. Species commerce of the Pacific islanders, and the conspecific with these taxa or closely related species is second only to Allopeas gracile in to them occur also in Fiji, Niue, and Tonga the frequency with which it has been found and in the Cook, Tuamotu, and Marquesas in archaeological sites in the Pacific islands; Islands (Garrett 1884, Pilsbry and Cooke its presence has been reported from such sites 1918 – 1920, Preece 1995, Brook 2010, Brook 104 PACIFIC SCIENCE · January 2018 et al. 2010). The material examined here authors reported it to occur also on Tahiti spans much of the range of variation said by and in the Marquesas (Baker 1940). Two ad- Pilsbry and Cooke to characterize the three ditional species of Hiona, H. scalpta (Garrett, taxa they recognized as occurring in the Soci- 1884) and H. angustivoluta (Garrett, 1884), ety Islands, but in the absence of a compre- have also been reported from Mo‘orea (Baker hensive review of their status no attempt is 1940), but no specimens identifiable as ei- made to allocate the specimens examined to ther species were represented in the material particular members of the group, if indeed studied here. L II of ScMo-343 dates to the they are distinct. Specimens recovered here period AD 1521 to the present. Hiona verticil­ are of late prehistoric to modern age. One or lata was not collected by the 1934 Mangare- more species of Nesopupa now inhabit the van Expedition or by later surveys (Moorea ­island (Moorea Biocode 2016). Biocode 2016).

Family Valloniidae Hiona sp. material: Vaipahu, ScMo-343: L II, Pupisoma (Ptychopatula) orcula (Benson, 1850) TP1-B1, obj. 7 (1 spm.); L II, TP1-B2, obj. 9 material: Vaipahu, ScMo-343: L II, (1 spm.); L II, TP1-B3, obj. 8 (1 spm.); L II, TP1-B3, obj. 8 (1 spm.); L IIIa, TP2-B1, obj. TP1-B3, obj. 19 (1 spm.); L II, TP2-A4, obj. 7 (1 spm.). 7 (1 spm.); L IIIa, TP2-B1, obj. 6 (4 spms.); remarks: Only two specimens were ob- L IIIa, TP2-B1, obj. 7 (4 spms.); L IIIa, TP2- tained. The identification of one (TP2-B1) is B2, obj. 7 (1 spm.); L IIIa, TP2-B2, obj. 1 somewhat tentative because the shell is badly (1 spm.). broken and incomplete, but the other (TP1- remarks: A number of immature speci- B3) is intact and clearly identifiable as this mens were obtained of a second species of species. Pupisoma orcula is widely distributed Hiona with a strongly angulate periphery; in the Old World tropics (Pilsbry 1922 – 1926) these cannot be assigned with certainty to any and was translocated into the Pacific islands of the members of this genus known to occur prehistorically, where it has been recorded on Mo‘orea. All date to the period AD 1521 ­archaeologically in the Cook Islands (Brook to the present. 2010), Henderson Island (Preece 1998), and probably also in the Hawaiian Islands (Chris- Family Endodontidae tensen 1984, Brook 2010, Christensen and Weisler 2013). Pupisoma orcula currently in- Libera jacquinoti (Pfeiffer, 1850) habits Mo‘orea (Moorea Biocode 2016). The material: Pihaena, ScMo-342: L VIII, specimens recovered here cannot be dated TP2-B3, obj. 5 (1 spm.); L VIII, TP2-B3, obj. with certainty; although L IIIa of ScMo-343 6 (2 spms.). Vaipahu, ScMo-343: L II, TP1- is dated as AD 1521 – modern, and most likely B1, obj. 1 (1 spm.); L II, TP1-B1, obj. 1, ca. dates to between AD 1521 – AD 1767, the 70 cm b.s. (1 spm.); L II, TP1-B1, obj. 7 presence there of Subulina octona, a modern (2 spms.); L II, TP1-B2, obj. 3 (8 spms.); L II, introduction, suggests the possibility of soil TP1-B2, obj. 9 (2 spms.); L II, TP1-B3, obj. disturbance. 10 (3 spms.); L III, TP1-C1, obj. 1 (1 spm.); L I, TP2-A1, obj. 2 (2 spms.); L I, TP2-A2, obj. Family Helicarionidae 1 (1 spm.); L IIIa, TP2-B1, obj. 7 (8 spms.); L IIIa, TP2-B2, obj. 7 (1 spm.); TP2-B2, obj. 8 Hiona sp. cf. verticillata (Pease, 1864) (3 spms.); surface near AH8 (1 spm.). material: Vaipahu, ScMo-343: L II, remarks: The descriptions of Helix caver­ TP1-B2, obj. 3 (1 spm.). nula Hombron & Jacquinot, 1852, and pre- remarks: A single specimen was found of sumably also of its senior synonym H. a species of Hiona that most closely resembles ­jacquinoti, were based on material obtained Hiona verticillata (Pease, 1864), a species ap- during the voyage of the French exploration parently endemic to Mo‘orea, although early vessels Astrolabe and Zelée, which visited the Nonmarine Mollusks from Mo‘orea, Society Islands · Christensen et al. 105

Society Islands in 1838 (Rosenman 1992). B3, obj. 6 ( bpbm 276657, 2 spms.); TP2-B4, Pfeiffer (1850:128) reported his new species obj. 3 ( bpbm 276648, 2 spms.). ScMo-343: to occur “in insula Tahiti, et in insulis Mar- TP1-B3, obj. 10 ( bpbm 276711, 2 spms.); quesas,” but Solem (1976:418) noted the un- TP2-B1, obj. 7 ( bpbm 276958, 1 spm.); TP2- reliability of early locality records from that B2, obj. 8 ( bpbm 276702, 2 spms.). region and concluded that the range of the Additional referred material: Pihaena, species was “[u]nknown, but probably Tahiti ScMo-342: TP2-B3, obj. 5 ( bpbm 276644, or Moorea, Society Islands.” Libera jacquinoti 1 spm.); TP2-B3, obj. 6, 384 – 401 cm b.s. was not observed by Garrett (1884:111), who ( bpbm 276661, 1 spm.; bpbm 276833, 1 spm.); expressed the opinion that “it probably in­ TP2-B4, obj. 6 (276647, 1 spm.). Vaipahu, habits the Austral Islands.” The material re- ScMo-343: TP2-B1, obj. 6 ( bpbm 276846, ported here is the first new collection of this 1 spm.). species in more than 150 yr and conclusively description: Shell extremely large for establishes L. jacquinoti as a former Mo‘orea the family; in Libera only L. jacquinoti [maxi- endemic, now undoubtedly extinct. mum diameter 9.28 mm according to Solem The rediscovery of L. jacquinoti is analo- (1976)] approaches it in size; diameter 6.8 – gous to the rediscovery by Orliac (1997) of L. 9.7 mm, height 3.3 – 4.7 mm, height /diameter spuria (Ancey, 1889) and L. incognata Solem, ratio 0.47 – 0.57, with from 6-7/8 to 7-7/8 1976, in an archaeological site on Tahiti; both normally coiled whorls. Spire strongly ele­ species were unknown to Garrett and were vated, broadly rounded in outline above described from poorly localized material col- ­periphery, body whorl sometimes deflected lected in the early to mid-1800s (Solem 1976). slightly below periphery of its predecessor. Three of the five species ofLibera whose is- Umbilicus secondarily narrowed to form a land of origin could not be verified by Solem brood chamber, occasionally again decoiling have now been rediscovered in the course of just before the aperture, variable in size relat- recent archaeological investigations; the ori- ing to the occasional presence of this terminal gins of L. streptaxon (Reeve, 1852) and L. decoiling, contained 4.41 – 8.33 times in the heynemanni (Pfeiffer, 1862) remain unknown. diameter. Apical surface of postnuclear whorls usually with numerous very finely spaced low Libera dubiosa Ancey, 1889 rounded riblets, too numerous and indistinct material: Pihaena, ScMo-342: L V, for accurate quantification, strongly pro­ TP2-B4, obj. 3 (1 spm.). Vaipahu, ScMo-343: tracted and weakly sinuate, spiral sculpture L IIIa, TP2-B2, obj. 8 (1 spm.). absent. Sutures very shallow, whorls rounded remarks: The collection date of Ancey’s down to broad and shallow to very shallow specimens of Libera dubiosa is unknown, al- ­supraperipheral sulcus, periphery strongly though Solem (1976) has suggested that they protruded and strongly keeled, a shallow were probably obtained during the explor­ ­subperipheral sulcus immediately below. Rib- atory expeditions of the early nineteenth cen- lets on basal surface absent or much reduced tury. The species was unknown to Garrett except immediately below periphery, when (1884), and fresh or live-collected specimens present sometimes evident only as weak have not been obtained subsequently. ­closely spaced growth wrinkles. Parietal bar- riers 2, the upper high and slender, ascend- Libera kondoi Christensen, Khan & Kirch, ing very gradually anteriorly and terminating n. sp. rather abruptly posteriorly well before the Figure 2(A – C ) limit of visibility; lower parietal barrier not material: Holotype ( bpbm 276694): So- as high, with or without a low threadlike ante- ciety Islands, Mo‘orea, Vaipahu, ScMo-343: rior extension, otherwise originating gradu- TP2-C2, obj. 2. ally about 1/8 whorl posterior to origin of Paratypes: Mo‘orea, Pihaena, ScMo-342: ­upper palatal, terminating posteriorly simul- E2, obj. 6, N102 E102 ( bpbm 276654, 1 spm.); taneously with ­upper palatal and in a similar TP2-B3, obj. 5 ( bpbm 276646, 1 spm.); TP2- manner. Columellar barrier low, recessed. Figure 2. Libera kondoi, n. sp., holotype, bpbm 276694: (A) apical view; (B) lateral view; (C ) basal view. Minidonta opunuhoa, n. sp., holotype, bpbm 277701: (D) apical view; (E ) lateral view; (F ) basal view. Nonmarine Mollusks from Mo‘orea, Society Islands · Christensen et al. 107

Palatal barriers 4, the first subperipheral, tury. Libera kondoi is represented at levels evenly spaced, bladelike,­ the second sub­ from the pre-Polynesian era to the post-AD peripheral barrier sometimes more strongly 1521 period, indicating that it survived into ­developed than others­ except much reduced the period of Polynesian occupation but be- in one specimen, all subperipheral palatals came extinct in late prehistory or soon after ­ascending gradually and terminating more initial European contact. abruptly; supraperipheral palatal situated etymology: The species is named in midway between periphery and parietal-­ memory of Yoshio Kondo, Malacologist at palatal margin, much lower than subperiph­ Bishop Museum for many years, and mentor erals but extending farther posteriorly, as- to C.C.C. and P.V.K. cending and descending very gradually. Shell with sinuous reddish-brown flammules in Minidonta opunohua Christensen, Khan, & specimens retaining color. One specimen Kirch, n. sp. ( bpbm 276647) is referred to this species with Figure 2(D – F ) some doubt because it is unique in having ra- material: Holotype ( bpbm 277071): So- dial sculpture that extends onto basal surface ciety Islands, Mo‘orea, Vaipahu, ScMo-343: of the shell, where it is crossed by fine raised TP1-B3, obj. 8. spiral striae. Measurements are based on the Paratypes: Same data as holotype (except holotype ( bpbm 276694) and eight adult test pit number), ScMo-343, TP1-B1, obj. 6 paratypes­ ( bpbm 276646, 276654, 276657/2, ( bpbm 274992, 5 spms.); TP1-B1, obj. 7 ( bpbm 276702/2, 276711, 276958), although not all 276724, 2 spms.); TP1-B2, obj. 3 ( bpbm 276740, measurements could be taken on all speci- 7 spms.); TP1-B2, obj. 9 ( bpbm 275001, 2 mens because of damage. spms.); TP1-B3, obj. 8 ( bpbm 276635, 9 spms.); remarks: Libera kondoi is very similar to TP2-A4, obj. 6 ( bpbm 274979, 4 spms.); TP2- L. jacquinoti (Pfeiffer, 1850) in size, overall A4, obj. 7 ( bpbm 276755, 1 spm.); TP2-B1, shape, and apertural barriers, but is distin- obj. 6 ( bpbm 276847, 8 spms.); TP2-B1, obj. 7 guished from it by having consistently much ( bpbm 276772, 3 spms.). finer and more numerous riblets on the apical description: Shell minute, 0.9 – 1.2 mm surface of the shell and the absence of strong in height, 2.0 – 2.6 mm in diameter, with from basal sculpture. According to the key pub- 4-3/8 to 4-7/8 normally coiled whorls. Apex lished by Solem (1976:390 – 391), L. kondoi flat to very slightly elevated, early postnuclear most resembles L. dubiosa Ancey, 1889, and whorls descending gradually, the last slightly L. gregaria Garrett, 1884, which are also more rapidly. Height /diameter ratio 0.41 – ­endemic to Mo‘orea. The large size, lack of 0.52. Umbilicus deeply cup-shaped, early ­spiral cording, and more strongly protruding whorls decoiling regularly, the last more periphery of L. kondoi distinguish it from L. ­rapidly. Diameter/umbilicus ratio 3.01 – 3.38. dubiosa; and also from L. gregaria, which is Postnuclear whorls with numerous sharply further distinguished by having only 2 instead defined regularly spaced ribs whose inter­ of 3 subperipheral palatal barriers. stices are 2 to 3 times their width, body whorl Libera kondoi is one of four previously with 88 – 117 ribs. Microsculpture of several ­unknown species of Endodontidae recovered very fine riblets between each pair of ribs. from pre-contact levels of archaeological ­Sutures impressed, periphery not angulate, ­excavations undertaken during the study regularly rounded except slightly compressed ­re­ported here. The fact that this prominent at base. Parietal barriers 4, the first 3 blade- lowland species was not encountered by the like, simple, the first the highest, extending early collectors who discovered L. jacquinoti, about one-fourth whorl into aperture, the and was unknown to Garrett (1884), strongly second and third slightly lower, weakly down- suggests that it had become extinct before turned; fourth parietal much lower, less dis- Garrett’s extensive collecting efforts in the tinct, sometimes barely indicated. Columellar 1860s through 1880s, perhaps before initial barrier absent. Palatal wall with 5 major sim- European contact in the late eighteenth cen- ple bladelike barriers, usually with a much 108 PACIFIC SCIENCE · January 2018 lower minor barrier between first and second very broadly open, diameter/umbilicus ratio palatals, similar minor barriers sometimes 1.70; on apical surface sculpture of nuclear also in interstices between remaining major whorls lost to erosion, surface of post-nuclear palatals. Color, in those specimens retain- whorls with 7 – 9 low spiral cords crossed by ing it, of regularly spaced reddish-brown low growth striae, the spiral cords first ap- flammulations. Measurements are from the pearing on third whorl and becoming obso- holotype and 10 unbroken adult paratypes lete on last 1/2 whorl, sutures moderately im- ( bpbm 274992, 275001, 276635/2, 276724, pressed; basal surface of post-nuclear whorls 276740/2, 276847/3). with sculpture similar to that of apical surface, remarks: The combination of 4 parietal sutures deeply impressed; margin of aperture barriers and 5 major palatal barriers distin- rounded at apical suture, flattened above and guishes M. opunohua from all members of the broadly rounded below angled periphery; genus reviewed by Solem (1976), including plane of aperture inclined at an angle of 30° all of the previously described taxa known to from shell axis; a single low parietal barrier, inhabit the Society Islands. Most species of receding out of sight within aperture, colu- Minidonta subsequently described from Raro- mellar and palatal barriers absent; color off- tonga, Cook Islands (Brook 2010); Rürutu, white with rather regularly spaced reddish- Austral Islands (Sartori et al. 2013); the Gam- brown flammulations. bier Archipelago (Abdou and Bouchet 2000); diagnosis: Nesodiscus nummus is distin- and Henderson Island (Preece 1998) are simi- guished from all previously described mem- larly distinguishable. Among the exceptions is bers of the genus by its depressed spire, large M. boucheti Sartori, Gargominy & Fontaine, umbilicus, with a d /u ratio of 1.70 [Solem 2013, from Rürutu, which is generally similar (1976) gave a range of 1.90 – 2.15 for species to M. opunuhoa but has a more prominent known to him; the umbilicus of N. cookei is fourth parietal barrier than the latter species, very slightly smaller than that of N. nummus, which lacks the fifth and sixth parietals some- with a d /u ratio of 1.74], and prominent sculp- times present in M. boucheti. Three Raro­ ture of fine spiral cords crossing the growth tongan species must also be discussed. In M. lines. Nesodiscus nummus is most similar to N. matavera Brook, 2010, M. pue Brook, 2010, cookei (description following), from which it and M. rutaki Brook, 2010, the shell is more differs primarily in sculpture. Spiral striae are elevated than that of M. opunohua and the um- the dominant sculptural element in N. num­ bilicus narrower; M. matavera also possesses a mus, whereas these are much reduced in N. columellar barrier, absent in M. opunohua. cookei and are limited to recessed areas at Minidonta opunohua is another previously the sutures and in the depths of the supra­ unknown species apparently endemic to peripheral and subperipheral sulci. Nesodiscus Mo‘orea and now extinct. nummus lacks the sulci so prominent in N. etymology: The species takes its name cookei. from ‘Opunohu Bay, the type locality. Solem noted that some species of Nesodis­ cus, especially N. taneae (Garrett, 1872), ex- Nesodiscus nummus Christensen, Kahn & hibit substantial intraspecific variation in shell Kirch, n. sp. form and remarked (1976:349) that “[m]ore Figure 3(A – C ) than in any other genus of Endodontidae, material: Holotype ( bpbm 276655): So- ­certain identification [withinNesodiscus ] re- ciety Islands, Mo‘orea, Pihaena, ScMo-342: quires the availability of series.” The limited E2, obj. 6, N102 E102. number of specimens of N. nummus and N. description: Shell 1.5 mm in height, cookei available to us (one shell of the former, 6.3 mm in diameter, height /diameter ratio two of the latter) prevents a proper review of 0.24, with 6-1/8 whorls; diameter at third the range of variation within each species. whorl 1.1 mm, at fourth whorl 1.6 mm, and at Nevertheless, although the two are obviously fifth whorl 2.4 mm; spire depressed, last whorl more closely related to each other than to not descending at aperture; umbilicus 3.7 mm, any non-Moorean member of the genus, the Figure 3. Nesodiscus nummus, n. sp., holotype, bpbm 276655: (A) apical view; (B) lateral view; (C ) basal view. Nesodiscus cookei, n. sp., holotype, bpbm 276658: (D) apical view; (E ) lateral view; (F ) basal view. 110 PACIFIC SCIENCE · January 2018 differences in shell sculpture and form be- in all specimens seen but showing remnants of tween N. nummus and N. cookei fully justify reddish-brown flammulations. their treatment here as two distinct species. diagnosis: Discussed under N. nummus. remarks: Nesodiscus is endemic to the remarks: Nesodiscus cookei, a species en- ­Society Islands, where seven species were re- demic to Mo‘orea, was found only in levels ported by Solem (1976) to occur in the islands dating to the period before initial human set- of Borabora, Maupiti, Huahine, Ra‘iätea, and tlement of the island. Taha‘a; none has previously been reported etymology: The species is named for from Mo‘orea. The unique specimen of N. Charles Montague Cooke, Jr., founder of the nummus was obtained from an undated strati- Bishop Museum’s vast Pacific malacology graphic level. ­collections and leader of the Museum’s 1934 etymology: The name is from the Latin Mangarevan Expedition. nummus, a small coin, and refers to the ap- pearance of the shell; it is treated as a noun in Family Charopidae apposition. Sinployea modicella (Férussac in Deshayes, Nesodiscus cookei Christensen, Kahn & Kirch, 1840) n. sp. material: Vaipahu, ScMo-343: L II, Figure 3(D – F ) TP1-B1, obj. 7 (1 spm.); L II, TP1-B3, obj. 8 material: Holotype ( bpbm 276658): So- (2 spms.); L II, TP2-A4, obj. 6 (2 spms.); ciety Islands, Mo‘orea, Pihaena, ScMo-342: L IIIa, TP2-B1, obj. 6 (8 spms.); L IIIa, TP2- TP2-B3, obj. 6. B1, obj. 7 (6 spms.). Paratypes: TP2-B3, obj. 6, 384 – 401 cm remarks: Sinployea modicella is endemic b.d. ( bpbm 276660, 1 spm.); TP2-B4, obj. 3 to Mo‘orea (Solem 1983); it was collected on ( bpbm 276650, 1 spm.). the island in 1934 (Solem 1983) and is still ex- description: Shell 6.8 mm in diameter, tant at high elevation ( J. D. Slapcinsky, pers. 1.6 mm in height, height /diameter ratio 0.24, comm., 20 June 2016). with 6-1/4 whorls; diameter at third whorl 1.1 mm, at fourth whorl 1.6 mm, and at fifth Sinployea sp. whorl 2.6 mm; spire depressed, last whorl not material: Vaipahu, ScMo-343: L IIIa, descending at aperture; umbilicus 3.9 mm, TP2-B1, obj. 7 (1 spm.). very broadly open, diameter/umbilicus ratio remarks: A single badly worn specimen 1.74; apical sculpture of nuclear whorls lost to was found of a species of Sinployea that is erosion, post-nuclear whorls with a deep su- much more widely umbilicate than is S. modi­ praperipheral sulcus and rough growth striae, cella. It has some resemblance to Discocharopa recess of sulcus marked with 2 to 3 weak spiral aperta (Mo¨llendorff, 1888) but is larger than cords, sutures deeply impressed; basal surface that species [shell diameter 2.2 mm versus nuclear whorls without visible sculpture, <2.0 mm in D. aperta (fide Solem 1983)]. The post-nuclear whorls with a weak subperiph- single specimen dates from the period AD eral sulcus and several weak spiral cords in 1521 to the present. recess of sulcus and adjacent to suture; suture of last whorl deeply impressed, of preceding Family Subulinidae whorls less so; margin of aperture rounded at apical suture, strongly angulate at periphery, Allopeas gracile ( Hutton, 1834) weakly angulate below subperipheral sulcus; material: Pihaena, ScMo-341: L I, TP1- plane of aperture inclined at an angle of 40° A2 (4 spms,); L I, TP2-A4, obj. 7 (1 spm.); L from shell axis; a single moderately strong I, TP3-A1, obj. 4 (3 spms.). Vaipahu, ScMo- ­parietal barrier receding out of sight within 343: L I, TP1-A1, obj. 1 (2 spms.); L I, TP1- aperture, a single low but long subperipheral A3, obj. 1 (1 spm.); L II, TP1-B1, obj. 6 (14 palatal barrier deeply recessed within aper- spms.); L II, TP1-B1, obj. 7 (44 spms.); L II, ture, columellar barriers absent; color leached TP1-B2, obj. 3 (42 spms.); L II, TP1-B2, obj. Nonmarine Mollusks from Mo‘orea, Society Islands · Christensen et al. 111

9 (15 spms.); L II, TP1-B3, obj. 7 (1 spm.); L L II, TP1-B2, obj. 3 (3 spms.); L II, TP1-B3, II, TP1-B3, obj. 8 (37 spms.); L II, TP1-B3, obj. 8 (4 spms.); L II, TP1-B3, obj. 10 (3 obj. 10 (15 spms.); L II, TP2-A4, obj. 6 (51 spms.); L II, TP2-A4, obj. 6 (1 spm.); L II, spms.); L II, TP2-A4, obj. 7 (47 spms.); L IIIa, TP2-A4, obj. 7 (4 spms.); L IIIa, TP2-B1, obj. TP2-B1, obj. 7 (91 spms.); L IIIa, TP2-B1, 6 (4 spms.); L IIIa, TP2-B1, obj. 7 (3 spms.). obj. 6 (86 spms.); L IIIa, TP2-B1, obj. 7 (91 remarks: This species, formerly known spms.); L IIIa, TP2-B2, obj. 7 (16 spms.); L as Opeas pumilum (Pfeiffer, 1840), has been IIIa, TP2-B2, obj. 8 (8 spms.). Teavaro, carried throughout the tropics by modern ScMo-349: AH5, obj. 1, 90 – 120 cm b.s. commerce and is widely distributed in the (1 spm.); L III, TP1-B1, obj. 3 (1 spm.). ­islands of the Pacific (Pilsbry 1906 – 1907; remarks: Allopeas gracile (= Lamellaxis Cowie 1997a, 2001a; Brook 2010). It was not gracile and Opeas oparanum) is a circumtropi- recorded by Garrett (1884) but had become cal species widely distributed by human com- established on Mangareva in the Gambier Is- merce (Pilsbry 1906 – 1907; Cowie 1997a, lands by the mid- to late nineteenth century 2001a). It was the only subulinid reported and on Mo‘orea by 1925 (Brook 2010). It is from the Society Islands by Garrett (1884), still extant on Mo‘orea (Moorea Biocode who referred to it as Stenogyra tuckeri (Pfeif­ 2016). Fresh shells have been found in ar­ fer, 1846). It is the anthropophilic land snail chaeological excavations on Rarotonga, Cook most often recovered from pre-contact ar- Islands, at depths of up to 0.25 m, suggesting chaeological sites in Oceania; records include that the species is partly subterranean in habi- the eastern Solomon Islands, New Caledonia, tat (Brook 2010), and thus that its presence in the Marshall Islands, Fiji, the Cook Islands, deeper stratigraphic levels has little strati- Samoa, Tonga, and the Society, Gambier, graphic significance. Marquesas, and Hawaiian Islands (Chris- tensen and Weisler 2013, 2017, and refer- Paropeas achatinaceum (Pfeiffer, 1846) ences cited therein). On Mo‘orea it was well material: Pihaena, ScMo-341: L I, TP1- represented at sites dating from the late pre- A1, obj. 7, 0 – 10 cm b.s. (4 spms.); L III, TP1- historic period or later, but its presence (with B3, obj. 2 (1 spm.); L I, TP2-A1, obj. 4 (4 Lissachatina fulica, a modern introduction) in spms.). Pihaena, ScMo-342: L I, TP1-A1, obj. the pre-Polynesian L III of ScMo-349 sug- 3, 0 – 10 cm b.s. (25 spms.); L I, TP1-A2, obj. gests contamination or soil disturbance. 6, 10 – 20 cm. b.s. (1 spm.); L I, TP1-A3, obj. 6, 20 – 30 cm b.s. (1 spm.); L III, TP1-C7, Leptinaria unilamellata (d’Orbigny, 1837) obj. 3 (1 spm.). Vaipahu, ScMo-343: L I, material: Pihaena, ScMo-341: L I, TP1- TP2-A1, obj. 2 (1 spm.). A2 (2 spms.); L I, TP2-A3, obj. 4 (2 spms.); L remarks: This is another species that has I, TP3-A2, obj. 6 (1 spm.). Vaipahu, ScMo- become widely distributed in the islands of 343: L II, TP1-B1, obj. 7 (4 spms.). the Pacific as a result of modern commerce remarks: This species occurs in the West (Pilsbry 1906 – 1907; Cowie 1997a, 2001a). Indies, Central America, Venezuela, Peru, The earliest record from Polynesia is from Bolivia, and Brazil (Pilsbry 1906 – 1907, Dutra Hawai‘i in 1904; P. achatinaceum had become 1988, Robinson et al. 2009). Leptinaria unila­ established on Tahiti by 1925 (Brook 2010) mellata was first reported from French Poly- and was present on Mo‘orea by 1934 (e.g., nesia (without further details) by Soubeyran bpbm 150364) and currently inhabits the is- (2008) and more recently has been found to land (Moorea Biocode 2016). The presence of inhabit Mo‘orea (Moorea Biocode 2016) and this modern immigrant in L III of ScMo-342 Rürutu and Raivavae in the Austral Islands suggests contamination. (Gargominy and Fontaine 2015). Subulina octona (Bruguière, 1789) Opeas hannense (Rang, 1831) material: Pihaena, ScMo-341: L I, TP1- material: Pihaena, ScMo-342: AH3, A2 (8 spms.); L I, TP2-A1, obj. 4 (1 spm.); L obj. 13, 330 – 360 cm b.s. (1 spm.); ScMo-343: I, TP2-A3, obj. 4 (2 spms.); L I, TP2-A4, obj. 112 PACIFIC SCIENCE · January 2018

7 (1 spm.). Vaipahu, ScMo-343: L II, TP1- effect on the native land snails inhabiting the B1, obj. 6 (2 spms.); L II, TP1-B1, obj. 7 high islands of tropical Polynesia (e.g., Chris- (2 spms.); L II, TP1-B2, obj. 9 (1 spm.); L I, tensen and Kirch 1986, Cowie 2001b, Bouchet TP2-A1, obj. 2 (1 spm.); L II, TP2-A4, obj. 6 and Abdou 2003, Lydeard et al. 2004, Rich- (2 spm.); L II, TP2-A4, obj. 7; L IIIa, TP2- ling and Bouchet 2013; Regnier, Achaz, et al. B1, obj. 6 (2 spms.); L IIIa, TP2-B1, obj. 7 2015; Regnier, Bouchet, et al. 2015; Brodie (8 spms.). et al. 2016). This process began with initial remarks: This is yet another invasive cir- human settlement and has only accelerated cumtropical species now widely distributed since the initiation of Western contact in the in Oceania (Pilsbry 1906 – 1907; Cowie 1997a, eighteenth century (Solem 1976, 1990; Cowie 2001a). The view of Solem (1964) that it is of 1992; Regnier, Bouchet, et al. 2015). The dis- tropical American origin is confirmed by its appearance of many of the region’s native presence in a pre-Columbian archaeologi- snails has been accompanied by their partial cal site in the U.S. Virgin Islands (Quitmyer replacement with assemblages of nonnative 2003). It is a modern introduction to the Pa- species, the earliest of them arriving with the cific islands, having reached New Caledonia prehistoric voyages of the Pacific islanders by 1888, the Hawaiian Islands by 1903, and (Christensen and Kirch 1981, Christensen the Society Islands by 1925 (Brook 2010); it and Weisler 2013). The rate of introduction was present on Mo‘orea in 1934 (e.g., bpbm increased with the vast expansion of inter­ 150328) and still inhabits the island (Moorea island, interarchipelago, and now intercon­ Biocode 2016). Its presence in ScMo-342: L tinental commerce during the modern era II indicates either a modern age for this level (Cowie 2001a, 2001b). The study reported or contamination. here sought to investigate the chronology of this process of land snail extinction and re- Subulinidae ( Unidentified) placement on the island of Mo‘orea. material: Vaipahu, ScMo-343: L II, We focus especially on the endemic TP1-B2, obj. 3 (1 spm.). Endodontidae of Mo‘orea, because this group remarks: This specimen, too immature is one of the most species-rich elements of for precise identification, undoubtedly be- the island’s nonmarine molluscan fauna and longs to one of the species of Subulinidae because prior studies (Garrett 1884, Solem ­reported here. 1976) have provided data on historical col­ lections of these snails essential to a chrono- Family Achatinidae logical analysis of the extinction process. The land snail family Endodontidae, used here in Lissachatina fulica (Bowdich, 1822) the restricted sense of Solem (1976), is en- material: Teavaro, ScMo-349: L III, demic to the islands of Remote Oceania. Rep- TP1-B1, obj. 3 (1 spm.). resentatives of the family occur from Palau remarks: The Giant African Snail, for- eastward through the Ellice Islands, Lau merly placed in the genus Achatina, is a noto- ­Archipelago of Fiji, Tonga, Samoa, Cook rious invasive species (Mead 1961, 1979) na- ­Islands, Society Islands, Austral Islands, Tua- tive to East Africa that first became established motu Archipelago, and Gambier Archipelago on Tahiti in 1967 and soon spread to Mo‘orea to Henderson Island and northward to the and other nearby islands (Clarke et al. 1984). Hawaiian Islands; fossil taxa extend the range It is represented in the material reported here of the family to the Marshall Islands and Mid- by a single apical fragment found at a sup­ way Island in the northwestern Hawaiian posedly pre-Polynesian level of ScMo-349. ­Islands (Solem 1976, 1977). Solem’s 1976 monograph was based primarily on live-­ discussion collected or recently dead specimens obtained from the early nineteenth century to the It is now clear that both prehistoric and mod- ­mid-twentieth century. Since the publication ern human activities have had a devastating of Solem’s work, material from paleontologi- Nonmarine Mollusks from Mo‘orea, Society Islands · Christensen et al. 113 cal and archaeological excavations has dem- by the time of Garrett’s activities in the late onstrated that the Polynesian endodontid nineteenth century. The precise timing of the fauna was much more diverse than previously extinction of the two new species of Nesodiscus known. Numerous recently extinct taxa from is unclear because one is known only from such sites have been described from the pre-Polynesian material and the other only ­Hawaiian Islands (Christensen 1982), Cook from an undated specimen; neither was en- Islands (Brook 2010), Austral Islands (Zim- countered by nineteenth-century collectors, mermann et al. 2009, Sartori et al. 2013), and they were undoubtedly extinct by Gar- Gambier Islands (Abdou and Bouchet 2000), rett’s time. The helicinid here identified as Makatea in the Tuamotu Archipelago (Sartori Orobophana sp. aff. maugeriae appears to have et al. 2014), and Henderson Island (Preece had a similar history, because it was not known 1998). The study reported here adds four to inhabit Mo‘orea before the study reported ­previously unknown endodontid species to here. Libera jacquinoti, described in the early the Moorean fauna and demonstrates that one to mid-nineteenth century but not collected ­additional species, described early in the nine- subsequently until now, must have become teenth century from poorly localized material, extinct after 1838 but before Garrett’s time. was in fact endemic to Mo‘orea. Solem (1976) concluded that collections of Solem (1976) and Bouchet and Abdou Libera dubiosa also date from the early nine- (2003) have shown how the review of his­ teenth century, and thus the time frame of torical records and museum collections can that species’ extinction was probably similar document the chronology of the extinction to that of L. jacquinoti, because both had dis- of insular land snail faunas, the former in the appeared by Garrett’s time. Four species col- context of the Endodontidae of the Society lected by Garrett but not seen subsequently, Islands. The study reported here extends that Mautodontha (Garrettoconcha) punctiperforata, methodology by incorporating records of M. (G.) perforata, Libera gregaria, and L. now-extinct native taxa and prehistorically recedens,­ are undoubtedly now extinct and had ­introduced alien taxa from dated pre-contact probably disappeared by the time of the 1934 archaeological contexts to provide a more Mangarevan Expedition, which failed to find comprehensive understanding of this process them. None of these is listed as extant in of faunal extinction and replacement. the database of the Moorea Biocode Project (Moorea Biocode 2016), although a single ­unidentified species of endodontid was found Extinction of Mo‘orea’s Native Nonmarine at high elevation in 2010 ( J. D. Slapcinsky, Mollusks pers. comm., 20 June 2016). By contrast, the Solem (1976) identified six endodontid spe- island’s diverse fauna of Partulidae apparently cies endemic to Mo‘orea and discussed the suffered relatively little from prehistoric and chronology of their disappearance. The study early modern human activities, because those reported here adds one species known only snails were still abundant in the early twen­ from the period before initial Polynesian tieth century (Crampton 1932). The partu- ­settlement, two species known only from pre- lids persisted in good numbers (Murray and historic or early historic era archaeological Clarke 1980, Murray et al. 1982, Johnson contexts, one species from undated archaeo- et al. 1986) until the misguided introduction logical material, and one species described in 1977 of the carnivorous snail Euglandina early in the nineteenth century from unlocal- rosea (Férussac, 1821), in an unsuccessful bio- ized material but now shown to have inhab­ logical control program targeting the Giant ited Mo‘orea. Two of the four previously African Snail, Lissachatina fulica. The intro- ­unknown endodontid species recovered from duction of Euglandina resulted in the rapid archaeological contexts (Libera kondoi and extermination of most of the island’s partulids Minidonta opunohua) must have become ex- (Tillier and Clark 1983, Clarke et al. 1984, tinct during the pre-contact period or early in Murray et al. 1988, Cowie 1992), although a the historic period; they were certainly extinct few scattered populations apparently persist 114 PACIFIC SCIENCE · January 2018

(Lee et al. 2009). Species of Omphalotropis Project (2016) recorded the existence of a and Hiona known formerly to have inhabited possible new species of charopid on the is- the island also appear to have become extinct land’s peaks, where an unidentified endodon- since the late nineteenth century. tid also occurs ( J. D. Slapcinsky, pers. comm., It appears that the extinction of Mo‘orea’s 20 June 2016). In any event, it is clear that the endemic land snails has been a continuing native land snail faunas that once inhabited process since the initial human settlement of the coastal lowlands of Mo‘orea have been the island (or perhaps earlier, given the un­ devastated by ecological changes that are the certain timing of the disappearance of the direct and indirect result of human activities. two Nesodiscus species), although difficulties in resolving the pre-contact /post-contact inter- Comparison with Tahiti face make it impossible to ascribe any extinc- tion events to the earlier period conclusively. Archaeological studies and museum collec- The timing of that process has varied among tions from Tahiti present a similar picture the different taxonomic elements of the fauna. of continuing extinctions among the native With regard to Polynesian endodontids, Endodontidae, although again there the na- Solem attributed their extinction to habitat tive partulids persisted in good numbers until disturbance and predation by nonnative ants, the introduction of Euglandina. Among the noting (1976:100) that with only two excep- endodontids, Solem (1976) identified three tions “no endodontids have been taken from species of Mautodontha and six of Libera in- disturbed primary forest or secondary vegeta- habiting Tahiti, suggesting that four or five tion zones.” He attributed the relatively low additional species described in the nineteenth diversity of native endodontids in Samoa to century from poorly localized material might the presence there of endemic ants. More re- also have come from Tahiti. Garrett en­ cently (Athens 2009, Drake and Hunt 2009), countered four species, Mautodontha parvidens biologists have become aware of the major (Pease, 1861), Libera bursatella (Gould, 1846), role of the prehistorically introduced Pacific L. garrettiana Solem, 1976, and L. retunsa Rat, Rattus exulans, as an agent of ecological (Pease, 1864), but of these only L. bursatella change on oceanic islands. The native land has been collected subsequently. Specimens mollusks of Mo‘orea, especially the ground- of subspecies L. b. orofenensis Solem, 1976, dwelling endodontids, undoubtedly suffered were collected at high altitude by Mangarevan from the influence of rats both through direct Expedition personnel, and a single specimen predation (not detectable in the fossil record of the nominate subspecies was collected in of the small to minute species represented 1946 by a collector for Field Museum. Four of here because their shells would be entirely de- the species recognized by Solem, Mautodontha stroyed when eaten by rats) and as a result of zimmermani Solem, 1976; M. aoraiensis Solem, rat-induced changes in the lowland vegeta- 1976; L. micrasoma Solem, 1976; and L. tion. As in Hawai‘i ( Hadfield and Sauffler cookeana Solem, 1976, were high-elevation 2009), predation on arboreal snails by climb- species collected by members of the 1934 ing rats has also been a factor (Cowie 1992), Mangarevan Expedition. As on Mo‘orea, re- but Moorean partulids nevertheless con­tinued cent archaeological investigations supplement to thrive until the recent introduction of existing museum collections by providing evi- Euglandina,­ a highly efficient alien predator dence of species that had disappeared before against which they had no defense. the earliest biological collections on the is- It has been shown elsewhere in Polynesia land. Orliac (1997) excavated sites in Tahiti’s that endemic land snails may survive at the Papeno‘o Valley where he obtained speci- higher elevations of tropical Polynesian is- mens of a number of native land snails from lands (Solem 1972, Gargominy 2008), includ- pre-contact contexts. As with Libera jacquinoti ing as yet undescribed species of Endodon­ and the study reported here, Orliac confirmed tidae (Brook 2012). This may well be true of that two species previously known only from Mo‘orea as well, because the Moorea Biocode unlocalized material collected in the first Nonmarine Mollusks from Mo‘orea, Society Islands · Christensen et al. 115 half of the nineteenth century, Libera spuria copta servilis, Opeas hannense, Paropeas achatina­ (Ancey, 1889) and L. incognata Solem, 1976, ceum, Subulina octona, Leptinaria unilamellata, were in fact native to Tahiti; he also recorded and Lissachatina fulica) are historic-era intro- four previously unknown and undescribed ductions, the latest of them, L. unilamellata, taxa that he identified asMautodontha aff. first being reported from Oceania in 2008. parvidens (Pease, 1861), Nesodiscus aff. cretaceus Two aquatic species, Melanoides tuberculata (Garrett, 1884), Libera aff. dubiosa (Ancey, and Tarebia granifera, will probably prove to 1889), and L. aff. incognata Solem, 1976. To have been prehistoric introductions to the summarize, Tahiti is now known to have been ­Society Islands; the material reported here inhabited by four species of endodontids that suggests but does not prove a pre-contact became extinct during the prehistoric or early arrival­ for M. tuberculata. Determination of post-contact period, two species that sur- the dates of introduction of certain species is vived until the nineteenth century but that made more difficult by anomalous occur­ had evidently gone extinct by the late 1800s, rences at improbably early levels due to sedi- and five species that survived until the period ment disturbance, possibly as a result of crab 1934 – 1946, although only at high elevations. burrowing (Specht 1985). Again as on Mo‘orea, Tahiti’s native Par- tulidae survived well into the twentieth cen- conclusion tury but have been virtually extirpated since the introduction of Euglandina (Crampton Mo‘orea provides yet another example of the 1916, Coote et al. 1999), though again as on extreme vulnerability of the native land snail Mo‘orea a few populations have thus far sur- faunas of Polynesian islands. As with the land vived this new onslaught of predation (Coote birds of the region (Olson and James 1982; 2007). Steadman 1995, 2006; Duncan et al. 2013), the native terrestrial mollusks of the high ­islands of Polynesia are rapidly disappearing Introduction of Alien Nonmarine Mollusk Species as a result of the direct and indirect effects of The pre-contact voyages of the Pacific island- human activities, both pre- and post-contact. ers resulted in the interisland translocation of Although examination of historical museum at least 10 species of land snails into or within collections can yield much information about tropical Polynesia; some were introduced into the extent and chronology of modern mol­ the region from Island Southeast Asia, and luscan extinctions (Solem 1976, Bouchet and others that were originally native to one or Abdou 2003), analysis of materials from pale- more islands within the region became much ontological and archaeological investigations more generally distributed (Christensen and will be essential to determine the true extent Weisler 2013). The speed of the process ac- of the original diversity of the terrestrial celerated greatly with the advent of European ­mollusk faunas of these islands and the full influence in the region in the sixteenth cen- chronology of the extinction process. Fur- tury and in the Society Islands post-AD 1767 thermore, the presence of alien land snail with the ever-increasing frequency of out- ­species introduced during both the pre- and side contact with once-isolated island com- post-contact eras provides a method of dating munities (Cowie 2001a, 2001b). The study similar to the use that has been made of evi- reported here documented the presence on dence of the presence of the Polynesian rat, Mo‘orea of five species previously shown to Rattus exulans (e.g., Wilmshurst and Higham have been transported into and /or within 2004). Unfortunately, relatively few studies Polynesia (Lamellidea oblonga, L. pusilla, Gas­ of native land snails from dated pre-contact trocopta pediculus, Pupisoma orcula, and Allopeas sites in tropical Oceania are yet available (e.g., gracile), although limitations on chronological Christensen and Kirch 1986, Orliac 1997, precision make it impossible to conclusively Preece 1998), and as the study reported here demonstrate their pre-contact presence on shows, it can be difficult to conclusively Mo‘orea. An additional six species (Gastro­ ­distinguish pre-settlement, post-settlement / 116 PACIFIC SCIENCE · January 2018 pre-contact, and early post-contact extinction Hunt, eds. Historical ecology in the Pacific events. Nevertheless, nonmarine mollusks islands: Prehistoric environmental and are often well represented in sediments from landscape change. Yale University Press, Pacific sites, and their study is an important New Haven. complement to paleoenvironmental studies in ——— . 1998. Holocene sea-level change on the region. Aitutaki, Cook Islands: Landscape change and human response. J. Coastal Res. 14:10 – acknowledgments 22. Allen, M. S., and C. C. Christensen. 1992. We express our appreciation to the following Landsnail analysis. Pages 490 – 497 in M. S. individuals who assisted or facilitated our re- Allen. Dynamic landscapes and human search: Chantal Tahiata, Ministèrè de la Cul- subsistence: Archaeological investigations ture, de L’Artisianat, et de la Famille; Teddy on Aitutaki Island, southern Cook Islands. Tehei, Chef de Service, Service de la Cul- Ph.D. diss., University of Washington, ture et du Patrimoine; Priscille Frogier, Chef Seattle.­ de Service, Délégation à la Recherche de la Ancey, C. F. 1892. Études sur la faune mala- Polynésie Française; M. Van Bastolaer, le cologique des Iles Sandwich. Mem. Soc. Maire de Mo‘orea-Maiao. Belona Mou and Zool. Fr. 5:708 – 722. Tamara Maric of the Service de la Culture et Anderson, A. 2002. Faunal collapse, landscape du Patrimoine facilitated the permit process. change and settlement history in Remote Diana Izdebski drafted the figures. We thank Oceania. World Archaeol. 33:375 – 390. the following property owners for permission Athens, J. S. 2009. Rattus exulans and the cata- to work on their lands: the Wilder family, strophic disappearance of Hawai‘i’s native Mari Mari Kellum, Eric de Bovis, Aimata lowland forest. Biol. Invasions 11:1489 – Teariki, Nathalie Terai Richmond, and Adol- 1501. phe Frogier, and Neil Davies, Director, Athens, J. S., and J. V. Ward. 2006. ‘Aimakapä Gump Research Station. Permission to work Fishpond: Paleoenvironment at Kaloko- in ‘Opunohu Valley was granted by Louis Honoköhau National Historic Park, North Sandford, Director, Domaine d’Opunohu. Kona, Hawai‘i Island. Report prepared John D. Slapcinsky provided information on for the National Park Service, Pacific West recent field surveys on Mo‘orea. Photographs Region. Pacific Islands Support Office, of shells are courtesy of Holly Bolick. Haleakalä National Park, Makawao, Maui, Hawai‘i, by International Archaeological Research Institute, Inc. Literature Cited Athens, J. S., J. V. Ward, and D. W. Blinn. Abbott, R. T. 1958. The gastropod genus 2007. Paleoenvironment of Pu‘uhonua o Assiminea­ in the Philippines. Proc. Acad. Hönaunau National Historical Park, South Nat. Sci. Phila. 110:213 – 278, pls. 15 – 25. Kona, Hawai‘i Island. Report prepared Abdou, A., and P. Bouchet. 2000. Nouveaux for the National Park Service by Inter- gastéropodes Endodontidae et Punctidae national Archaeological Research Institute, (Mollusca, Pulmonata) récemment étients Inc. de l’archipel des Gambier (Polynésie). Baker, H. B. 1940. Zonitid snails from Pacific Zoosystema 22:689 – 707. islands. Part 2. Hawaiian genera of Micro- Allen, M. S. 1992. Dynamic landscapes and cystinae. Bernice P. Bishop Mus. Bull. 165. human subsistence: Archaeological investi- Bandel, K., and T. Kowalke. 1997. Eocene gations on Aitutaki Island, southern Cook Melanotarebia n. g. and its relations among Islands. Ph.D. diss., University of Wash- modern Thiaridae (Caenogastropoda: ington, Seattle. Cerithioidea). Neues Jahrb. Geol. Palae- ——— . 1997. Coastal morphogenesis, cli- ontol. Monatsh. 11:683 – 695. matic trends, and Cook Island prehistory. Bouchet, P., and A. Abdou. 2003. Endemic Pages 124 – 146 in P. V. Kirch and T. L. landsnails from the Pacific islands and the Nonmarine Mollusks from Mo‘orea, Society Islands · Christensen et al. 117

museum record: Documenting and dating ——— . 2017. A record of the invasive land the extinction of the terrestrial Assiminei- snail Allopeas gracile ( Hutton, 1834) (Pul- dae of the Gambier Islands. J. Molluscan monata: Subulinidae) from a First Mil­ Stud. 69:165 – 170. lennium BC archaeological site in the Brodie, G., G. M. Barker, H. Pippard, C. S. southeastern Solomon Islands. Archaeo + Bick, and D. Ó. Foighil. 2016. Disappear- Malacology Group Newsl. 27:2 – 4. ing jewels: An urgent need for conserva- Clarke, B., J. Murray, and M. S. Johnson. tion of Fiji’s partulid tree snail fauna. Pac. 1984. The extinction of endemic species by Conserv. Biol. 22:249 – 261. a program of biological control. Pac. Sci. Brook, F. J. 2010. Coastal landsnail fauna 38:97 – 104. of Rarotonga, Cook Islands: Systematics, Conte, E., and P. V. Kirch. 2008. One diversity, biogeography, faunal history, ­thousand years of human environmental and environmental influences. Tuhinga transformation in the Gambier Islands 21:161 – 252. (French Polynesia). Terra Austral. 29:253 – ——— . 2012. Report on the landsnail fauna 264. of Upland Savai‘i. Pages 140 – 150 in J. Conte, E., and G. Molle. 2014. Reinvestigat- Atherton and B. Jeffries, eds. Rapid bio­ ing a key site for Polynesian prehistory: diversity assessment of Upland Savai‘i, New results from the Hane dune site, Ua ­Samoa. Secretariat of the Pacific Regional Huka (Marquesas). Archaeol. Oceania Environment Programme, Apia, Samoa. 49:121 – 136. Brook, F. J., R. K. Walter, and J. A. Craig. Cooke, C. M., Jr. 1935. Report of C. Mon- 2010. Changes in the terrestrial molluscan tague Cooke Jr., Malacologist and leader. fauna of Miti‘aro, southern Cook Islands. Pages 36 – 56 in H. E. Gregory, Report of Tuhinga 21:75 – 98. the Director for 1934, Bernice P. Bishop Burney, D. A. 2002. Late Quaternary chro- Mus. Bull. 133. nology and stratigraphy of twelve sites on Cooke, C. M., Jr., and Y. Kondo. 1961. Re­ Kaua‘i. Radiocarbon 44:13 – 44. vision of Tornatellinidae and Achatinelli- Christensen, C. C. 1982. A new species of dae (Gastropoda, Pulmonata). Bernice P. Endodonta (Pulmonata, Endodontidae) Bishop Mus. Bull. 221. from Oahu, Hawaii. Malacol. Rev. 15:135 – Coote, T. 2007. Partulids on Tahiti: Differ- 136. ential persistence of a minority of endemic ——— . 1984. Analysis of nonmarine mol- taxa among relict populations. Am. Mala- lusks. Pages 355 – 376 in R. Schilt, Subsis- col. Bull. 22:83 – 87. tence and conflict in Kona, Hawai‘i: An Coote, T., E. Loewe, J.-Y. Meyers, and D. archaeological study of the Kuakini High- Clarke. 1999. Extant populations of en- way Realignment Corridor. Department of demic partulids on Tahiti, French Polyne- Anthropology, Bernice P. Bishop Museum, sia. Oryx 33:215 – 222. Departmental Report Series 84-1. Cowie, R. H. 1992. Evolution and extinction Christensen, C. C., and P. V. Kirch. 1981. of Partulidae, endemic Pacific island land Nonmarine mollusks from archaeological snails. Philos. Trans. R. Soc. Lond. B Biol. sites on Tikopia, southeastern Solomon Sci. 335:167 – 191. ­Islands. Pac. Sci. 35:75 – 88. ——— . 1997a. Catalog and bibliography of ——— . 1986. Nonmarine mollusks and eco- the nonindigenous nonmarine snails and logical change at Barbers Point, O‘ahu, slugs of the Hawaiian Islands. Bishop Mus. Hawai‘i. Bishop Mus. Occas. Pap. 26:52 – Occas. Pap. 50:1 – 66. 80. ——— . 1997b. Catalog of the nonmarine Christensen, C. C., and M. I. Weisler. 2013. snails and slugs of the Samoan Islands. Land snails from archaeological sites in the Bishop Mus. Bull. Zool. 3. Marshall Islands, with remarks on prehis- ——— . 2001a. Invertebrate invasions on Pa- toric translocations in tropical Oceania. cific islands and the replacement of unique Pac. Sci. 67:81 – 104. native faunas: A synthesis of the land and 118 PACIFIC SCIENCE · January 2018

freshwater snails. Biol. Invasions 3:119 – predation by rats on the island of Moloka‘i 136. 1982 – 2006. Biol. Invasions 11:1595 – 1609. ——— . 2001b. Decline and homogenization Haynes, A. 1988. Notes on the stream ne­ of Pacific faunas: The land snails of Ameri- ritids (Gastropoda; Prosobranchia) of Oce- can Samoa. Biol. Conserv. 99:207 – 222. ania. Micronesica 21:93 – 102. Crampton, H. E. 1916. Studies on the varia- ——— . 2001. Freshwater snails of the tropi- tion, distribution and evolution of the ge- cal Pacific islands. Institute of Applied Sci- nus Partula: The species inhabiting Tahiti. ences, Suva. Carnegie Inst. Wash. Publ. 228:1 – 311. ——— . 2005. An evaluation of members of ——— . 1932. Studies on the variation, distri- the genera Clithon Montfort, 1810 and bution, and evolution of the genus Partula. Neritina­ Lamarck, 1816 (Gastropoda: Ne- The species inhabiting Moorea. Carnegie ritidae). Molluscan Res. 25:75 – 84. Inst. Washington Publ. 410. Howard, N. M., and P. V. Kirch. 2004. Zoo- Dixon, B., D. Soldo, and C[ harles] C. Chris- archaeological analysis of faunal assem- tensen. 1997. Radiocarbon dating land blages. Pages 106 – 121 in E. Conte and snails and Polynesian land use on the island P. V. Kirch, eds. Archaeological investiga- of Kaua‘i, Hawai‘i. Hawaii. Archaeol. tions in the Mangareva Islands (Gambier 5:52 – 62. Archipelago), French Polynesia. Univer­ Drake, D. R., and T. L. Hunt. 2009. Invasive sity of California, Berkeley, Archaeological rodents on islands: Integrating historical Research Facility, Contribution No. 62. and contemporary ecology. Biol. Invasions Hubendick, B. 1956. A conchological survey 11:1483 – 1487. of the Genus Plecotrema (Gastropoda, Duncan, R. P., A. G. Boyer, and T. M. Black- ­Ellobiidae). Proc. Malacol. Soc. Lond. burn. 2013. Magnitude and variation of 32:110 – 126. prehistoric bird extinctions in the Pacific. Johnson, M. S., J. Murray, and B. C. Clarke. Proc. Natl. Acad. Sci. U.S.A. 110:6436 – 1986. An electrophoretic analysis of phy- 6441. logeny and evolutionary rates in the genus Dutra, A. V. C. 1988. Aspectos da ecologia e Partula in the Society Islands. Proc. R. Soc. da reprodução de Leptinaria unilamellata Lond. B Biol. Sci. 227:161 – 177. (Orbigny, 1835) (Gastropoda, Subulini- ——— . 1993. The ecological genetics and dae). Rev. Bras. Zool. 5:581 – 591. adaptive radiation of Partula on Moorea. Gargominy, O. 2008. Beyond the alien inva- Oxford Surv. Evol. Biol. 9:167 – 238. sion: A recently discovered radiation of Joordens, J. C. A., F. P. Wesselingh, J. de Nesopupinae (Gastropoda: Pulmonata: Vos, H. B. Vonhof, and D. Kroon. 2009. Vertiginidae) from the summits of Tahiti Relevance of aquatic environments for (Society Islands, French Polynesia). J. hominins: A case study from Trinil ( Java). Conchol. 39:517 – 536. J. Hum. Evol. 57:656 – 671. Gargominy, O., and B. Fontaine. 2015. Over- Kahn, J. G. 2012. Coastal occupation at the view of the terrestrial and freshwater GS-1 Site, Cook’s Bay, Mo‘orea, Society ­molluscs. Pages 54 – 91 in J.-Y. Meyer Islands. J. Pac. Archaeol. 3:52 – 61. and E. M. Claridge, eds. Terrestrial bio­ Kahn, J. G., C. Nickelsen, J. Stephenson, N. diversity of the Austral Islands, French Porch, E. Dotte, C. C. Christensen, L. Polynesia. Edition: Patrimoines naturels, May, J. S. Athens, and P. V. Kirch. 2015. 72, Muséum national d’Histoire naturelle, Late Holocene coastal morphogenesis and Paris. human ecodynamics on Mo‘orea, Society Garrett, A. 1884. The terrestrial Mollusca in- Islands (French Polynesia). The Holocene habiting the Society Islands. J. Acad. Nat. 25:333 – 347. Sci. Phila. (2) 9:17 – 114, pls. 2 – 3. Kay, E. A. 1978. Hawaiian marine shells. Hadfield, M. G., and J. E. Sauffler. 2009. The Bishop Museum Press, Honolulu. demographics of destruction: Isolated pop- Kennet, D. A., A. J. Anderson, M. Prebble, E. ulations of arboreal snails and sustained Conte, and J. Southon. 2006. Prehistoric Nonmarine Mollusks from Mo‘orea, Society Islands · Christensen et al. 119

human impacts on Rapa, French Polynesia. ——— . 1979. Pulmonates, Vol. 2B. Econom- Antiquity 80:340 – 354. ic malacology with particular reference to Kirch, P. V. 1973. Prehistoric subsistence Achatina fulica. Academic Press, London. patterns in the northern Marquesas Moorea Biocode. 2016. Moorea Biocode Islands, French Polynesia. Archaeol. Phys. Project database. http:// biocode.berkeley Anthropol. Oceania 8:24 – 40. .edu /. Accessed June 2016. Kirch, P. V., E. Conte, W. Sharp, and D. Murray, J., and B. Clarke. 1980. The genus Nickelsen. 2010. The Onomea site (Tara- Partula in Moorea: Speciation in prog­ vai Island, Mangareva) and the human ress. Proc. R. Soc. Lond. B Biol. Sci. ­colonization of southeastern Polynesia. 211:83 – 117. ­Archaeol. Oceania 45:66 – 79. Murray, J., M. S. Johnson, and B. Clarke. Kirch, P. V., M. V. Gonzalez, and A. M. 1982. Microhabitat differences among Plourde. 2017. Invertebrate faunal remains ­genetically similar species of Partula. Evo- from Tangatatau Rockshelter. Pages 139 – lution 36:316 – 325. 155 in P. V. Kirch. ed. Tangatatau Rock- Murray, J., E. Murray, M. S. Johnson, and B. shelter (Mangaia, Cook Islands): The Clark. 1988. The extinction of Partula on ­evolution of an eastern Polynesian socio- Moorea. Pac. Sci. 42:150 – 153. ecosystem. Monumenta Archaeologica. Newell, J. 2010. Trading nature: Tahitians, Cotsen Institute of Archaeology, Univer- Europeans and ecological exchange. Uni- sity of California, Los Angeles. versity of Hawai‘i Press, Honolulu. Kondo, Y. 1975. Status of Lamellidea solomon­ Olson, S. L., and H. F. James. 1982. Fossil ensis Dell (Stylommatophora: Achatinel­ birds from the Hawaiian Islands: Evidence lidae: Lamellideinae). Malacol. Rev. 8:114. for wholesale extinction by Man before Ladd, H. S. 1972. Cenozoic fossil mollusks Western contact. Science ( Washington, from western Pacific islands: Gastropods D.C. 217:633 – 635. (Turritellidae through Strombidae). U. S. Oostingh, C. H. 1935. Die Mollusken des Geol. Surv. Prof. Pap. 532:1 – 79. Plioza¨ns von Boemiajoe ( Java). Wet. Lee, T., J. B. Burch, T. Coote, P. Pearce-­ ­Meded. Dienst Mijnbouw Ned.-Indie¨ Kelly, C. Hickman, J.-Y. Meyer, and D. Ó. 26:1 – 247. Foighil. 2009. Moorean tree snail survival Orliac, M. 1997. Human occupation and en- revisited: A multi-island genealogical per- vironmental modifications in the Papeno‘o spective. BMC Evol. Biol. 9:204. Valley, Tahiti. Pages 200 – 229 in P. V. Lydeard, C., R. H. Cowie, W. F. Ponder, Kirch and T. L. Hunt, eds. Historical ecol- A. E. Bogan, P. Bouchet, S. A. Clark, K. S. ogy in the Pacific islands: Prehistoric Cummings, T. J. Frest, O. Gargominy, ­environmental and landscape change. Yale D. H. Herbert, R. Hershler, K. E. Perez, University Press, New Haven, Connect­ B. Roth, M. Seddon, E. E. Strong, and icut. F. G. Thompson. 2004. The global de- Parkes, A. 1997. Environmental change and cline of nonmarine mollusks. BioScience the impact of Polynesian colonization: 54:321 – 330. Sedimentary records from central Polyne- McWethy, D. B., C. Whitlock, J. M. sia. Pages 166 – 199 in P. V. Kirch and T. L. Wilmshurst, M. S. McGlone, M. Fromont, Hunt, eds. Historical ecology in the Pacific X. Li, A. Dieffenbacher-Krall, W. O. islands: Prehistoric environmental land- Hobbs, S. C. Fritz, and E. R. Cook. 2010. scape change. Yale University Press, New Rapid landscape transformation in South Haven, Connecticut. Island, New Zealand, following initial Pfeiffer, L. 1850. Descriptions of twenty-four Polynesian settlement. Proc. Natl. Acad. new species of Helicea, from the collection Sci. U.S.A. 107:21343 – 21348. of H. Cuming, Esq. Proc. Zool. Soc. Lond. Mead, A. R. 1961. The giant African snail: A 17:126 – 131. problem in economic malacology. Univer- Pilsbry, H. A. 1906 – 1907. Manual of con- sity of Chicago Press, Chicago. chology. 2nd ser. Vol. XVIII. Achatinidae: 120 PACIFIC SCIENCE · January 2018

Stenogyrinae and Coeliaxinae. Academy of Regnier, C., P. Bouchet, K. A. Hayes, N. Y. Natural Sciences, Philadelphia. Yeung, C. C. Christensen, D. J. D. Chung, ——— . 1916 – 1918. Manual of conchology. B. Fontaine, and R. H. Cowie. 2015. Ex- 2nd ser. Vol. XXIV. Pupillidae (Gastro- tinction in a hyperdiverse endemic Hawai- coptinae). Academy of Natural Sciences, ian land snail family and implications for Philadelphia. the underestimation of invertebrate extinc- ——— . 1922 – 1926. Manual of conchology. tion. Conserv. Biol. 29:1715 – 1723. 2nd ser. Vol. XXVII. Pupillidae (Orculi- Rehder, H. A. 1980. The marine mollusks of nae, Pagodulinae, Acanthinulinae, etc.). Easter Island ( Isla de Pascua) and Sala y Academy of Natural Sciences, Philadel- Gómez. Smithson. Contrib. Zool. 289:1 – phia. 167. Pilsbry, H. A., and C. M. Cooke Jr. 1918 – Richling, I. 2009. The radiation of the Heli- 1920. Manual of conchology, 2nd ser. cinidae in New Caledonia (Mollusca: Vol. XXV. Pupillidae (Gastrocoptinae, ­Gastropoda: Neritopsina) including zoo- Vertigininae). Academy of Natural Sci- geographic considerations. Zoologia Neo- ences, Philadelphia. caledonica 7. Biodiversity Studies in New Pointier, J.-P., and G. Marquet. 1990. Tax- Caledonia 198:247 – 372. onomy and distribution of freshwater mol- ——— . 2011. Comment on the Hawaiian lusks of French Polynesia. Venus Jpn. J. species of Helicinidae. Pages 40 – 63 (incl. Malacol. 49:215 – 231. 11 pls.) in M. Severns, Shells of the Ha­ Pokryszko, B. M. 1996. The Gastrocoptinae waiian Islands [Part 2]. The land snails. of Australia (Gastropoda: Pulmonata: Pu- The endemic land and freshwater species pilloidea). Invertebr. Taxon. 10:1085 – 1150. and their described variants. Conchbooks, Prebble, M., and J. M. Wilmshurst. 2008. De- Hackenheim. tecting the initial impact of humans and Richling, I., and P. Bouchet. 2013. Extinct introduced species on island environments even before scientific recognition: A re- in Remote Oceania using paleoecology. markable radiation of helicinid snails ( He- Biol. Invasions 11:1529 – 1556. licinidae) on the Gambier Islands, French Preece, R. C. 1995. Systematic review of the Polynesia. Biodivers. Conserv. 22:2433 – land snails of the Pitcairn Islands. Biol. J. 2468. Linn. Soc. 56:273 – 307. Robinson, D. G., A. Hovestadt, A. Fields, and ——— . 1998. Impact of early Polynesian A. S. H. Breure. 2009. The land Mollusca ­occupation on the land snail fauna of Hen- of Dominica (Lesser Antilles), with notes derson Island, Pitcairn Group (South Pa- on some enigmatic or rare species. Zool. cific). Philos. Trans. R. Soc. Lond. B Biol. Meded. (Leiden) 83:615 – 650. Sci. 353:347 – 368. Rolett, B. V. 1992. Faunal extinctions and Quitmyer, I. 2003. Zooarchaeology of Cin­ depletions linked with prehistory and en- namon Bay, St. John, U.S. Virgin Islands: vironmental change in the Marquesas Pre-Columbian overexploitation of animal ­Islands (French Polynesia). J. Polynesian resources. Bull. Fla. Mus. Nat. Hist. Soc. 101:86 – 94. 44:131 – 138. ——— . 1998. Hanamiai: Prehistoric coloni- Raven, H., and J. J. Vermeulen. 2007. Notes zation and cultural change in the Mar­ on mollusks from NW Borneo and Singa- quesas Islands, East Polynesia. Yale Univ. pore. 2. Synopsis of the Ellobiidae (Gas- Publ. Anthropol. 81. tropoda, Pulmonata). Vita Malacol. 4:29 – Rosenman, H. 1992. Two voyages to the 62. South Seas/ by Captain Jules S.-C. Du- Regnier, C., G. Achaz, A. Lambert, R. H. mont d’Urville: Translated from the Cowie, P. Bouchet, and B. Fontaine. 2015. French and retold. University of Hawai‘i Mass extinction in poorly known taxa. Press, Honolulu. Proc. Natl. Acad. Sci. U.S.A. 112:7761 – Sartori, A. F., O. Gargominy, and B. Fon- 7766. taine. 2013. Anthropogenic extinction of Nonmarine Mollusks from Mo‘orea, Society Islands · Christensen et al. 121

Pacific land snails: A case study of Rurutu, Inseln. Ann. Naturhist. Mus. Wien French Polynesia, with description of eight 80:473 – 656. new species of endodontids. Zootaxa ——— . 1993. Ergebnisse der o¨sterreichischen 3640:343 – 372. Tonga-Samoa Expedition 1985 des Insti- ——— . 2014. Radiation and decline of tuts fü r Zoologie der Universita¨t Wien: endodontid land snails in Makatea, French Beitra¨ge zur Kenntnis der Sü ß- und Polynesia. Zootaxa 3772:1 – 68. ­Brackwasser-Gastropoden der Tonga- und Sinoto, Y. H. 1983. Archaeological excava- Samoa-Inseln. Ann. Naturhist. Mus. Wien tions of the Vaito‘otia and Fa‘ahia sites on 94/95:217 – 306. Huahine Island, French Polynesia. Natl. Steadman, D. W. 1995. Prehistoric extinc- Geogr. Soc. Res. Rep. 15:583 – 599. tions of Pacific island birds: Biodiversity Solem, A. 1959. Systematics and zoogeogra- meets zooarchaeology. Science ( Washing- phy of the land and fresh-water Mollusca ton, D.C.) 267:1123 – 1131. of the New Hebrides. Fieldiana Zool. 43. ——— . 2006. Extinction and biogeography ——— . 1964. New records of New Caledo- of tropical Pacific birds. University of Chi- nian nonmarine mollusks and an analysis cago Press, Chicago. of the introduced mollusks. Pac. Sci. Stevenson, J., A. Benson, J. S. Athens, J. G. 18:130 – 137. Kahn, and P. V. Kirch. ( In press.) Polyne- ——— . 1972. Tekoulina, a new viviparous sian colonization and landscape change ­tornatellinid land snail from Rarotonga, on Mo‘orea, French Polynesia: The Lake Cook Islands. Proc. Malacol. Soc. Lond. Temae pollen record. The Holocene. 40:93 – 114. Szabó, K. 2009. Molluscan remains from Fiji. ——— . 1976. Endodontoid land snails from Terra Aust. 31:183 – 211. Pacific islands (Mollusca: Pulmonata: Sig- Thomas, W. S. 1979. A biography of Andrew murethra). Part I. Family Endodontidae. Garrett, early naturalist of Polynesia. Part Field Museum Press, Chicago. 1. Nautilus 93:15 – 28. ——— . 1977. Fossil endodontid land snails Tillier, S., and B. C. Clarke. 1983. Lutte from Midway Atoll. J. Paleontol. 51:902 – ­biologique et destruction du patromoine 911. génétique: Le cas des mollusques gastéro- ——— . 1983. Endodontoid land snails from podes pulmonés dans les territoires fran- Pacific islands (Mollusca: Pulmonata: Sig- çais du Pacifique. Genet. Sel. Evol. 15:559 – murethra). Part II. Families Punctidae and 566. Charopidae, zoogeography. Field Museum van Benthem Jutting, T. 1937. Non marine Press, Chicago. Mollusca from fossil horizons in Java with ——— . 1989. Non-camaenid land snails of special reference to the Trinil fauna. Zool. the Kimberley and Northern Territory, Meded. (Leiden) 20:83 – 180. Australia. I. Systematics, affinities and van Damme, D. 1984. The freshwater Mol- ranges. Invertebr. Taxon. 2:455 – 604. lusca of northern Africa: Distribution, bio- ——— . 1990. How many Hawaiian land snail geography, and palaeoecology. W. Junk, species are left? And what can we do for Dordrecht. them. Bishop Mus. Occas. Pap. 30:27 – 40. van Damme, D., and M. Pickford. 2003. The Soubeyran, Y. 2008. Espèces exotiques enva- late Cenozoic Thiaridae (Mollusca, Gas- hissantes dans les collectivités françaises tropoda, Cerithioidea) of the ­Albertine d’outre-mer. Etat des lieux et recomman- Rift Valley ( Uganda-Congo) and their dations. Collection Planète Nature. Co- bearing on the origin and evolution of mité française de l’IUCN. Paris, France. the Tanganyikan thalassoid malaco­fauna. Specht, J. 1985. Crabs as disturbance factors Hydrobiologia 498:1 – 83. in tropical archaeological sites. Aust. Wagner, A. J. 1905. Helicinenstudien. Denksch. ­Archaeol. 21:11 – 18. Kaiserl. Akad. Wiss., mathematisch-natur­ ­ Starmü hlner, F. 1976. Beitra¨ge zur Kenntnis wissenschaftliche Klasse 77:357 – 450, pls. der Sü ßwasser-Gastropoden pazifischer I – IX. 122 PACIFIC SCIENCE · January 2018

——— . 1907 – 1911. Die Familie der Helicini- clay largely devoid of volcanic clasts, and L VI (273 – dae. Neue Folge. In: Martini and Chem- 282 cmbs) was a gleyed, compact, sticky clay, entirely nitz, eds., Systematiasches Conchylien- lacking volcanic clasts with a thin lens of compressed peat at its upper limit. L VII (282 – 322 cmbs), the Cabinet 1 (18) [(2)]:1 – 391, pls. 1 – 70. Bauer basal deposit, was a calcareous marine sand contain- and Raspe, Nü rnberg. [published as ing shells of marine gastropods that was coarser- “1911”; pages 1 – 72, pls. 1 – 12: 1907; pages grained and included water-rolled pieces of Acropora 73 – 160, pls. 13 – 30: 1908; pages 161 – 216, branch coral in its lower part. Land snail – bearing deposits at -341 include those found pls. 31 – 42: 1909; pages 217 – 328, pls. in TP1 L I (19 – 36 cmbs) and TP3 L I (0 – 17 cmbs) 43 – 66: 1910; pages 3229 – 391, pls. 67 – 70: ( Kahn et al. 2015), a dark brown (10YR4/3) deposit 1911]. of fluvially transported sand and gravel incorporating Walter, R. 1998. Anai‘o: The archaeology of larger subrounded to rounded volcanic clasts. This a fourteenth century Polynesian commu- most likely represents a flood event from the nearby intermittent stream. Earlier work at the same site, but nity in the Cook Islands. N. Z. Archaeol. at its southern extent, dated the upper part of the Assoc. Monogr. 22. ­fluvial deposit (presumably the same as our L II ) to Wilmshurst, J. M., and T. F. G. Higham. between the late seventeenth – twentieth centuries AD 2004. Using rat-gnawed seeds to indepen- ( Kahn 2012), which provides a terminus post quem date for L I. Two new dates on short-lived species dently date the arrival of rats and humans from L III provide calibrated age ranges of AD 1410 – in New Zealand. The Holocene 14:801 – 1455 (Beta-411454, 510 ± 30 BP) and AD 1710 – 1950, 806. respectively (Beta-411453, 80 ± BP) at 2σ. The latter Zimmermann, G., O. Gargominy, and B. date has multiple intercepts; given that we did not re- Fontaine. 2009. Quatre espèces nouvelles cover historic artifacts in the deposit, the most likely age range is AD 1690 – 1730. d’Endodontidae (Mollusca: Pulmonata) Current data suggest that L I most likely postdates the éteintes de Rurutu ( Iles Australes, eighteenth century. A cultural deposit underlying this Polynésie francaise. Zoosystema 31:791 – fluvial deposit (and apparently not extending to the 805. northern portion of the site) was dated AD 1031 – 1210 ( Kahn 2012) and provides a best estimate for the ini- tial settlement of ScMo-341. This sample dates to the Appendix early Polynesian period (after initial colonization of the island, as opposed to pre-Polynesian, before set- ScMo-341 (Lat. 17° 29′ 24″ S, Long. 149° 49′ 37″ W ): tlement of the island). The -341 transect ran from near the base of a steep ScMo-342 (Lat. 17° 29′ 14″ S, Long. 149° 49′ 52″ W ): ridge (rising to an elevation of 140 m) across the gent­ This transect is situated 500 m west of -341, near the ly sloping coastal plain to the shoreline, a total dis- western edge of an alluvial fan that emanates from tance of 100 m. Three 1 m by 1 m test pits were exca- the mouth of Vaioma Valley. The transect ran 300 m vated along this transect. TP1 was located ca. 50 m from the shore across the gently sloping coastal inland of the circle-island road and had the deepest flat to the ephemeral Vaioma Stream. Two 1 m by stratification where seven depositional units were en- 1 m test units were excavated, in addition to a 4 m by countered ( Kahn et al. 2015). The uppermost deposit, 3 m block excavation. The stratigraphy encountered L I (L I, 0 – 50 cm below surface), was a massive de- in TP1 was similar to that found at -341. L I (0 – posit of sandy-clay loam, well sorted with occasional 38 cmbs) had a blocky ped structure, and the upper angular to subangular volcanic clasts and occasional 15 cm is the A horizon. L II (38 – 62 cmbs) was a charcoal flecks. This was the only land snail – bearing ­highly compact sandy-clay loam with rounded to sub- deposit found at the site; artifact recovery and geo- rounded volcanic clasts and charcoal flecking. Large morphology suggest that it represents modern and basalt cobbles forming a pavement were associated historic fill used to level the coastal plain. Below with dense charcoal concentrations in a hearth, indi- this, L II (50 – 65 cmbs) included a deposit of fluvially cating the presence of a former occupation and pos- transported sand and gravel incorporating larger sibly a cookhouse. L III (62 – 184 cmbs) was a massive, ­subrounded to rounded volcanic clasts; it most likely structureless, silty-clay loam with occasional char- represents a flood event from the nearby ephemeral coal flecking, the lowest 15 cm becoming strongly stream. L III (65 – 145 cmbs) was a deposit of massive, mottled. L IV (184 – 200 cmbs) was a gleyed, compact, structureless, well-sorted, highly compact silty clay structureless, and well-sorted clay, containing some with occasional subangular volcanic clasts and char- fibrous plant material. The basal deposit, L V (200 – coal flecks. L IV (145 – 150 cm) was a dark grayish- 240 cmbs), included a calcareous marine sand, fine- to brown (10YR4/3-2) lens of silt and sand with rounded medium-grained, with fragments of marine shell. and subrounded volcanic clasts (1 – 5 cm size range), Land snail – bearing deposits at -342 include those re­ representing a single fluvial depositional event. L V covered in TP2 and excavation unit N102 E102 in (150 – 273 cmbs) represented a massive deposit of silty L V, the very dark gray (5YR3/1) medium- to coarse- Nonmarine Mollusks from Mo‘orea, Society Islands · Christensen et al. 123

grained calcareous sand with marine shells and abun- ScMo-349 (17° 30′ 49″ S, 149° 46′ 15″ W ): This site is dant anaerobically preserved plant remains. Seed situated on the northeastern side of Mo‘orea Island in samples dated from these sediments yielded ages of Teavaro, where there is a wide coastal plain (240 m). 3030 ± 30, 3120 ± 30, and 3740 ± 30 BP, consistent Four stratigraphic deposits were located during the with a pre-Polynesian period dating to between 4600 site excavations. L I (0 – 44 cmbs) was a structureless, and 2400 BP ( Kahn et al. 2015). medium brown deposit of silty clay with scattered ScMo-343 (Lat. 17° 29′ 49″ S, Long. 149° 51′ 03″ W ): subangular volcanic clasts. This deposit had high fre- This site is situated near the eastern headland of quencies of basalt flakes and debitage and moderate ‘Opunohu Bay ( Vaipahu Point) (Figure 1B). Here, amounts of charcoal. L II is a dark orange-brown the coastal plain is approximately 170 m wide and ter- ­deposit of fluvially and alluvially transported sand, minates at a steep ridge. We excavated two 1 m test gravel, and clay incorporating larger subrounded to pits in swampy ground. Four stratigraphic deposits subangular volcanic clasts. L III represents a compact, were exposed in TP1 and TP2. L I (0 – 6 cmbs) was a structureless, gleyed, and well-sorted clay containing silty-clay loam with a few scattered subrounded volca- fibrous plant material. The basal deposit, L IV, is a nic clasts. L II (6 – 96 cmbs) was a massive, structure- medium- to coarse-grained calcareous sand with less silty-clay loam with scattered charcoal flecks ­marine shells. and a few scattered subrounded volcanic clasts. L IIIa Land snail – bearing deposits at -349 include those from (96 – ca. 140 cmbs) included a compact, structureless, L I and those from L II. Although samples from -349 gleyed, and well-sorted clay containing some fibrous have not yet been dated, excavations and artifact re- plant material and with abundant land snails at the covery from these deposits are indicative of Polyne- base. L IIIb (ca. 140 – 190 cmbs) was a lens of dark sian-era deposits. gray, compact, structureless, gleyed, and well-sorted Lake Temae: This is a brackish-water lake situated in clay, containing some fibrous plant material. L IV the basin of a former lagoon (17° 29′ 05″ S, 149° 46′ (140 – 200 cmbs), the basal deposit, was a medium- to 08″ W ). Parkes (1997) analyzed sediment cores from coarse-grained calcareous sand with marine shells. Lake Temae and dated changes in charcoal content At -343, the land snail – bearing deposits include TP2, and pollen spectra thought to represent initial island L II, a dark brown (7.5YR3-4/2), massive, structure- colonization to AD 690 – 990. Our project is analyzing less deposit of silty-clay loam with scattered charcoal and dating newly taken cores from Lake Temae that flecks and a few scattered subrounded volcanic clasts. suggest initial anthropogenic impact on sedimenta- Land snails were also recovered from TP1 and TP2 L tion rates and vegetation change at or around AD III deposits, a dark gray, gleyed ( N4/ ), very compact, 1000 ( J. Stevenson, pers. comm.). Aquatic snails re- structureless, well-sorted clay, containing some covered from L I at a depth of 115 – 166 cm below the ­fibrous plant material. L III was not dated, but it is lake bottom are in a deposit with a terminus quem of bracketed by dates of 4610 ± 30 BP from L IV (pre- modern to ca. 200 yr BP, signaling that these samples Polynesian) and 250 ± 30 BP from L II (highest prob- derive from the post-contact or historic period (Ste- ability intercept of cal. AD 1626 – 1679, Polynesian venson et al., in press). period).