Rattus rattus (Linnaeus, 1758)

Today this species is present in all Madagascar habitats from 0 to 2500 m above sea level (a.s.l.) and is relatively abundant in anthropogenic places as well as in natural forests (Soarimalala and Goodman 2011). The date of arrival of R. rattus in Madagascar is poorly known but the earliest archaeological evidence dates from the ninth to eleventh century CE at the Islamic port of Mahilaka on the north-western part of the island (Rakotozafy 1996; Radimilahy 1997, 1998). Other eastern and south-eastern coastal sites date from even earlier centuries, with archaeological evidence of well-established human settlements from the seventh to tenth century CE in southern Madagascar, and the presence of Arabic trade providing a likely explanation for the early colonisation of the island by R. rattus. This is consistent with a recent investigation showing that the arrival of the earliest R. rattus on the Swahili coast of Africa and in Pemba and Zanzibar islands dated from the seventh to eighth century CE (Prendergast et al. 2017). The scenario of a single colonisation event was suggested by successive molecular analyses of Hingston et al. (2005) and Tollenaere et al. (2010), with the Arabian Peninsula as the possible origin of the species in Madagascar. However, molecular evidence from a later study (Brouat et al. 2014), while in agreement with a single ancestral source, instead suggested two independent introduction events for R. rattus, one in the north and the other in the south of the island.

Rattus rattus was listed from Andrahomana cave, a coastal site of the south-eastern part of the island, dated between 4500 and 1500 BP (Vasey and Burney 2007; Burney et al. 2008), and is briefly listed at Ankilitelo (SW Madagascar) by Muldoon et al. (2009). The timing of colonisation of the Central Highland forests by R. rattus is uncertain. According to Goodman (1995) the earliest presence of R. rattus in the central region dates from 1916. However, this is predated by a specimen of R. rattus (NHMUK 1897.9.1.155) collected by Major from Ivohitra, a locality visited by him while working in the vicinity of Antsirabe in 1895 (Jenkins and Carleton 2005). Moroever, in the material from Lavajaza, collected around 1901–1905, Denys et al. (2021) indicated the abundant presence of the species in the cave. The lack of reference by Major may be due to his incomplete sorting of the very large quantity of material from the Children’s Cave. In fact, our study allowed sorting of the cranio-dental remains of many relatively large sized murids which display the typical morphology of R. rattus, notably three rows of three bunodont and relatively well-aligned cusps on M1.

Attributed material: all from the Upper Stratum: 12 partial skulls and maxillaries; 41 mandibular rami; 6 isolated incisors; see Appendix.

The incisive foramen is long, entering the maxillary at the level of the first root of the M1 (Figure 4c). The largest tooth of the upper molar row is M1 and M3 is the smallest, but the latter molar is not very reduced and displays two lobes of cusps with the second lobe constituted by a single round cusp. On poorly worn molars the cusps are relatively well aligned transversally and the t1 and t4 are round and as large as the t2 and t5. The lobes are well separated by shallow convex valleys. The t3, t6 and t9 are small. There is no trace of longitudinal crests. On M2 and M3 the t1 cusp is larger than on M1.

The lower molars display a classical Rattus pattern of cusps with a prelobe constituted by four round cusps of equal size on the m1. The rows of cusps are relatively aligned transversally and there is a cingulum posteriorly on the distal part of m1 and m2.

The size of the upper molar rows as well as the width of M1 fit well with modern R. rattus specimens of Madagascar (Table 3, Figure 6).