Late Quaternary Mammal Introduction and Extinction Records from Archaeological Cave Deposits in Timor-Leste

. Humans have influenced island biotas in the last few millennia through widespread forest clearances and the introduction of commensals and pest species, causing the extinction of island endemics around the world. This is particularly evident in Timor, where more than 40,000 years of human habitation produced few if any extinctions until the last few thousand years when Timor lost most of its endemic murids. We present new records and dates for endemic rodents and introduced fauna from archaeological cave deposits in Timor-Leste that captures this human-mediated transition. We discuss the chronology of faunal introductions and losses at these sites, and compare the Timor records to other records in surrounding islands. We find no directly dated evidence for significant overlap between the introduction of exotics and extinction of murid endemics at ecological timescales, although determining true extinction and introduction ages will require direct dating and modelling of taxon occurrences, which may bring extinction and introduction closer together in time. Nevertheless, we suggest that, based on current data, the almost complete loss of Timor’s endemic forests were the primary driving force in rodent extinctions.

Abstract.Humans have influenced island biotas in the last few millennia through widespread forest clearances and the introduction of commensals and pest species, causing the extinction of island endemics around the world.This is particularly evident in Timor, where more than 40,000 years of human habitation produced few if any extinctions until the last few thousand years when Timor lost most of its endemic murids.We present new records and dates for endemic rodents and introduced fauna from archaeological cave deposits in Timor-Leste that captures this human-mediated transition.We discuss the chronology of faunal introductions and losses at these sites, and compare the Timor records to other records in surrounding islands.We find no directly dated evidence for significant overlap between the introduction of exotics and extinction of murid endemics at ecological timescales, although determining true extinction and introduction ages will require direct dating and modelling of taxon occurrences, which may bring extinction and introduction closer together in time.Nevertheless, we suggest that, based on current data, the almost complete loss of Timor's endemic forests were the primary driving force in rodent extinctions.

Introduction
Humans have had a disproportional impact on island biotas over the last few millennia (Ceballos & Ehrlich, 2018;Louys et al., 2021;Nogué et al., 2021), and the widespread introduction of commensals and pest species have been implicated in the extinction of island endemics around the world (Wood et al., 2017;Castilla-Beltrán et al., 2021).However, based on current archaeological and palaeontological records, humans and other hominins may have had more limited impacts on island ecosystems prior to the widespread adoption of agriculture, maritime trade, and domestication (Leppard, 2014;Rozzi et al., 2023;Louys et al., 2021;Wood et al., 2017), but this record is not well resolved for earlier periods of the Pleistocene.On the island of Timor, more than 40,000 years of human habitation produced very few extinctions (Hawkins et al., 2017;Louys et al., 2021); although an example of the latter includes a crane (Grus sp.) that likely became extinct in the Late Pleistocene (Meijer et al., 2019).It was only in the last few thousand years that Timor lost a disproportionate amount of its endemic biota (Aplin & Helgen, 2010).
Prehistoric excavations on Timor documenting some of these losses began in rockshelters near Nikiniki, southwest Timor (Fig. 1), by Alfred Bühler in 1935 (Sarasin, 1936) (no local names of the rockshelters are provided and they are referred to as "Abri" [rockshelter] I, II, and III).Bühler uncovered pottery fragments and domestic animals that were likely Holocene in age, as well as giant rat fragments from Abri II which were subsequently described by Schaub (1937) as the extinct species Coryphomys buehleri.In 1938, Willems of the Oudheidkundige recovered relatively recent archaeological material from Ulnam Cave on the slopes of Gunung Mutis located NW of Nikiniki (Oudheidkunig Verslag, 1939: 12).Two additional caves, Liang Leluat II and Liang Djenilu, excavated by Verhoeven in 1954, produced more Holocene material, including blades, scrapers, and worked points (Verhoeven, 1959).
Older deposits on Timor were first identified at Lene Hara cave in the eastern part of the island by the Portuguese anthropologist Antonio de Almeida in 1963.The site contained an 80 cm deep cultural assemblage with marine shells and stone artefacts found throughout the sequence, but with pottery fragments restricted to the surface (Fig. 1).A brief report on the stone artefacts described them as typologically "pre-Neolithic" (Almeida & Zybszweski, 1968).However, the site was never dated, and none of the fauna was properly described.In 1966, Glover visited the site and photographed Almeida's trench, which was still open (Glover, 1969).Glover made a small cutting on the edge of the trench, which confirmed Almeida's observation that pottery was absent below the surface (Glover, 1969).
Glover conducted additional excavations between 1966-1967 at Uai Bobo 1 and 2, Lie Siri, and Bui Ceri Uato in eastern Timor and found large amounts of prehistoric terrestrial faunal remains that, together, provided a baseline cultural sequence for the island (Glover, 1986) (Fig. 1).Glover (1986: appendix 2) presented new giant murids from his excavations, with initial identification of murid cranial remains made by Dan Witter, with further work and tabulation undertaken by Jack Mahoney, as described in Glover's (1986) Appendix 2. Mammalogist Guy Musser of the American Museum of Natural History subsequently agreed to take over the study of the fossil rodents, and he later passed on the material to Kristofer Helgen, who studied the collection with mammalogist and zooarchaeologist Kenneth P. Aplin (KPA).In addition to Coryphomys, three undescribed genera of giant rats were recognized in Glover's monograph.Pending formal description, these were designated "Large murid, genus A" (Glover, 1986: plate 49), "Large murid, genus B" (Glover, 1986: plate 50), and "Large murid, genus C" (Glover, 1986: plate 51).These are hereafter referred to as Genus A, B, and C, respectively.Genus C is absent from Bui Ceri Uato but otherwise all three genera are present in all of Glover's sites.
Following the annexation of Timor-Leste by Indonesia