Comparison of Timor dog δ 13 C and/or δ 15 N isotopic values with those of prehistoric dogs in the Cook Islands (Craig 2009), Marquesas (West 2007), New Zealand (Leach et al. 2001) and Fiji (Jones and Quinn 2009; Valentin et al. 2006). Isotopic values from Santa Rosa dogs are from ,VÊiÌÊ>°ÊÓ䣣®°Ê*>VwVÊ«}ÊÃÌ«VÊÛ>ÕiÃÊ>ÀiÊ`iÀÛi`ÊvÀÊiÊ>`Ê À>}ÊÓää®]Ê i>Û>Ì wi`ÊiÌÊ>°ÊÓään®]Êi`ÊiÌÊ>°ÊÓää®Ê>`Ê7iÃÌÊÓääÇ®° 

Contexts in source publication

Context 1

... to that of recent South Asia and Southeast Asia pariah populations, which Gonzalez (2012) divided into four types (dingo, sight hound, broad head dog and village dog). Measurements of dog remains from Neolithic levels in Niah Cave in Borneo indicate these animals had a body size smaller than the average village dog (Clutton- Brock 1959; Medway 1959). Further west, in east Java, Storm (2001) examined an incomplete skull from Hoekgrott Cave, which could represent either a large village dog, or a smaller dingo-type pariah; this specimen was tentatively dated to 2655±60 BP. In Timor, dog remains from Uai Bobo 1 (1400 BP) and Bui Ceri Uato (0–2500 BP) were described by Gollan (1982) as representing two types, one similar to the village dog, and the other larger and more dingo-like. In the east of Timor-Leste, adjacent to Lake Ira Laloro (~334 m asl), are two caves in uplifted limestone known as Matja Kuru 1 (MK1) and Matja Kura 2 (MK2). These are located approximately 9 km from the current coastline, with entrances facing south over the large freshwater lake (Figure 1). MK2 is one of several sites excavated locally which contain marine resources dating to the Pleistocene (Spriggs et al. 2003). A 1 x 1 m test pit yielded well-preserved faunal remains throughout the sequence, including abundant remains of giant rats, reptiles and freshwater turtles, from layers dating to ~36,000–30,000 cal. BP. Evidence for use of the coastal environment includes a broad range of marine shellfish, fish and turtle, which are most abundant in the earliest occupation phase. The earliest date obtained was ~36,000 cal. BP on marine shell from Spit 47, though bedrock was not reached during excavation and it is likely that older occupation deposits are present. Dates from higher in the sequence cluster into two periods: 13,000–9500 cal. BP and >3500 cal. BP. It thus appears that the deposit accumulated episodically, with an early occupation phase prior to 30,000 years ago, followed by a second occupation phase in the terminal Pleistocene-early Holocene, and a final phase in the mid- to late Holocene. MK2 currently has no evidence for human use between ~30,000 BP and 13,000 BP, the time coinciding with the Last Glacial Maximum (O’Connor and Aplin 2007; Veth et al. 2005). During excavation a dog skeleton was found in Spits 26–25 of MK2, with these remains appearing to have been interred in a pit dug from overlying levels. Pottery was present in small quantities in Spits 19–12, with most ceramics occurring in Spits 11–8. Two pot sherds in Spits 26–25 were found with the dog bone, indicating displacement of pottery from ceramic levels during digging of the burial pit. Two samples from the dog bones were prepared at the Waikato Radiocarbon Laboratory, with AMS determinations run at the Rafter Radiocarbon Laboratory on graphite targets processed by the reduction of CO with Zn in a reaction catalysed by iron 2 powder at a temperature of ~575oC. The first age determination was made on bone (L. humerus distal+midshaft) gelatin in 2003 and gave a conventional radiocarbon age (CRA) of 2967±58 BP (Wk-10051). However, the C:N ratio was 3.87 (carbon% 39.4, nitrogen% 11.87), indicating potentially poor sample quality. A new sample of ultrafiltered gelatin (L. femur prox+midshaft) was analysed in 2012 and produced a slightly younger CRA of 2867±26 BP (Wk-34931). The C:N ratio of 3.34 indicates good bone preservation (cf. Ambrose and Norr 1993:404; van Klinken 1999) and the sample produced an age of 2921–3075 cal. BP (Calib 6.1.0, two sigma, p =0.95). 13 C and/or 15 N isotopic values from bone protein are regularly used as quantitative measures of different dietary protein sources for humans and animals (Allen and Craig 2009; Field et al. 2009; Richards et al. 2009; Valentin et al. 2010). Carbon isotopes are generally considered to have a linear relationship between marine and terrestrial food sources. Stable nitrogen isotopes indicate the trophic level of an organism and have been used to determine the amounts of animal versus plant protein (Schoeninger and DeNiro 1984). On Nukuoro Atoll, Davidson (1992) dated two dog bones which had 13 C values of -15.4 and -12.5, suggesting a significant marine component in their diets. Two prehistoric dogs from New Zealand with values of 13 C values of -17.6 and -17.1 and 15 N values of 11.3 and 14.3 also consumed some marine foods (Leach et al. 2001:50). In contrast, Santa Rosa Island dogs of the late Holocene Chumash people, who relied exclusively on hunting and gathering wild foods (Rick et al. 2011), subsisted almost entirely on marine resources ( 13 C mean -1.5±0.8‰ and 15 N mean 17.9±0.5‰). The Timor dog has a 13 C value of -19.47 and a 15 N value of 9.70 (Wk-34931), which suggests a terrestrial diet similar to that of Pacific pigs, but having less marine food compared to some dogs from the Cook Islands and Marquesas, which also appear to show a difference in the trophic level of marine foods consumed (Figure 2). Several attempts were made to extract aDNA from the Timor dog remains. Both teeth (incisor, M 1 ) and bone samples (rib fragment) were processed but, while in relatively good condition macroscopically, none were well-preserved with respect to aDNA. After multiple attempts using the bone and tooth samples we amplified a very small fragment of the mitochondrial d-loop, spanning approximately 84 base pairs from site 15536–15628 (based on complete mtDNA genome reported in Bjornerfeldt et al. 2006). Over this very short region the Timor dog sequence is identical to those of Asian and Asian-derived breeds. Further analyses will focus on the 100 base pairs following site 15628 to determine whether the Timor dog has the point mutations found in dingoes and New Guinea singing dogs (Savolainen et al. 2004). The dog burial comprised 125 bone fragments, representing 92 skeletal elements. The bone was well preserved with little sign of weathering, staining, trampling, burning or gnawing. There is a high incidence of incipient breakage lines that have not developed into cracks; these likely represent the combined effects of the weight of sedimentary overburden and natural decay. The most commonly observed taphonomic feature was recent bone fragmentation which probably occurred post-excavation during transport of the MK2 assemblage to Australia. No cut marks, indicating removal of meat or skinning, were observed on any bones, and the element representation is consistent with the deliberate burial of a complete dog. Age and physical condition of the dog were evaluated using data related to epiphysial fusion, tooth eruption, pulp cavity size and limb bone thickness. The pulp cavity of the canine was assessed against previously published criteria (Kershaw et al. 2005; Knowlton and Whittemore 2001), and indicates an individual of three to four years of age. Tooth wear was, however, minimal, with dentition presenting well-defined cusps, perhaps reflecting a non-abrasive diet. The specimen was male, based on the presence of a complete os penis. Muscle attachments on the limb bones and the occipital region were deep and well-developed. A degree of pathological development was detected on some lumbar vertebrae, with slight deformation of the spinous processes. Such anomalies can be caused by early extensive muscular growth or by repetitive pack carrying/pulling. The only other anomaly was a supernumerary nutrient foramen in the left tibia. Cause of death was not evident, and the overall condition of the skeleton suggests a well fed and healthy animal. In summary, the skeletal data indicates that the Timor dog was an adult male that lived for three to four years, in which time the animal did not experience severe nutritional shortages; there is no skeletal evidence for severe trauma or disease. We have attempted to reconstruct the size and weight of the Timor dog using formulas developed by Anderson et al. (1985), Clark (1997), Hamblin (1984) and Wing (1978), which are based on measurements of length and diameter of limb bones, atlas and mandible. We also examined the degree of teeth crowding, which can help to identify wild from domesticated canids, following Davis and Valla (1978; but see Ovodov et al. 2011). Results suggest the Timor canine was a small to medium- sized dog with a body weight of 8–10 kg and a shoulder height of 39 cm (Table 1). In comparison with modern breeds, it is similar in height to a basenji, but the low body weight reflects a more gracile animal as suggested by the narrow diameter of limb bone mid-shafts. The value of the dental overlap index is 0.67, which indicates a slight degree of teeth crowding within the domesticated ...

Context 2

... Timor dog has a 13 C value of -19.47 and a 15 N value of 9.70 (Wk-34931), which suggests a terrestrial diet similar to that of Pacific pigs, but having less marine food compared to some dogs from the Cook Islands and Marquesas, which also appear to show a difference in the trophic level of marine foods consumed (Figure 2). ...