The Tongoleleka Lapita site is located in the village of Hihifo 250 m inland from the southwestern shore of Lifuka Island. Initial occupation occurred during a period of relatively higher sea level, and original site context is thought to have been in a back beach setting (Dickinson et al. 1994). Bordering the site on its inland eastern side is a low lying-swale. Formerly this would have been a wetland area.
The Tongoleleka site has been long known, first recorded in the 1960s and subsequently excavated by Dye (1988: 122-147) in 1984. The site was returned to in 1995 to acquire a data base consistent with earlier defined objectives for the Lapita project. This work first involved an auger test program to determine Lapita site boundaries and concentration areas. One location was then selected for the excavation of a 3 x 3 m block and taken to a depth of between 1.3 and 1.5 m (Figure 6). The site has an extremely dense Polynesian plain ware component in upper and middle strata that is underlain by early Lapita materials. As tallied in the report on 1995 field activities (Burley 1996a), a total of 38,492 undecorated sherds were recovered as well as 968 decorated pieces. Decorated rims account for 10.5 % of the overall rim assemblage. A total of 309 nonceramic artifacts were excavated, the majority (51.5 %) being shell valuables such as beads, bracelets, rings, pendants and disks. A large volume of faunal remains, including several species of extinct bird, were also collected. For select excavation units, a 20% proportionate sample of the 3.2 mm sieve matrix was retained for washing and intensive faunal collection. A 100 % shell fish sample was collected from three 50 x 50 cm subunits within the block excavation.
At the conclusion of the 1995 excavation, the Hihifo district officer asked that the excavation be left open for use as a septic field. The 3 x 3 m unit continued to remain open (and unused) in 1997, providing an opportunity for further stratigraphic excavations. Carried out from 9 - 17 June, two additional 1 x 1 m units off of the southeast corner were completed.
Objectives for the 1997 project at Tongoleleka were two-fold. First, David Steadman's preliminary analysis of the Tongoleleka bird fauna indicated a transition from upper to lower strata. To enhance his ability to define this transition, he proposed to carefully excavate a single 1 x 1 m unit in microstratigraphic fashion, with particular concern over the isolation of even the most minor features, disturbances, and stratigraphic anomalies. The exposed profile from the 1995 block facilitated this exercise, giving an opportunity to select a unit (No. 10) with few apparent disturbances. A faunal control sample for 3.2 and 1.6 mm sieves consistent with other sites was also to be collected. The second objective resulted from the first. Steadman's excavation revealed a distinct stratum (IIc) that was absent or unrecognized in the 1995 block. To further assess the nature and extent of this stratum, an adjacent 1 x 1 m unit (No. 11) was excavated to the southeast (Figure 6).
Steadman's excavation (Unit 10) employed the use of arbitrary levels of no more than 5 cm thickness bounded by principal stratigraphic breaks or feature outlines. All features and other disturbances were excavated as separate operations. Matrices were screened through nested sieves of 6.4, 3.2 and 1.6 mm size, and a 10% proportionate sample of the 3.2 and 1.6 sieved matrices were retained; the remaining 3.2 mm matrix fraction was then field-picked for bone and other small artifacts. The second unit (No. 11), excavated by myself, employed 10 cm arbitrary spits bounded by stratigraphic breaks. Matrices were screened through 6.4 and 3.2 mm sieves only.
The stratigraphy of the two 1 x 1 m units conforms to previous descriptions of the 3 x 3 m block with the one exception already noted. From top to bottom, this consisted of the following (Figure 7):
A small number of features were isolated in the 1997 excavations. Most notable of these are a 50 cm diameter post hole associated with Stratum IIb (Figure 7) and a buried hearth intrusive from Stratum III deep into Stratum IV. The hearth is of particular importance for it is associated with a well-preserved collection of extinct bird bones and turtle remains.
The Tongoleleka artifact and faunal assemblage recovered from the 1995 excavation remains the largest of the five Lapita sites. Now added to these totals are 5,418 ceramic pieces as well 43 nonceramic artifacts (Tables 6 and 7). Of the ceramic assemblage, only a very small number (n=104) are decorated. However, the percentage of decorated rims is 9.8% of the total rim assemblage, a figure quite comparable to that of the earlier project. The nonceramic artifact assemblage is similar to existing collections by including a high proportion of shell valuables.
A total of 28 charcoal samples was collected from the 1997 excavations. One sample each from Stratum IIa and IIb of Unit 11 and another from the early Lapita hearth of Stratum III, Unit 10, were submitted for radiocarbon dating. Both the IIa (2490 + 50 BP) and IIb (2430 + 50 BP) dates are consistent with a Polynesian plain ware phase occupation. The Stratum III date of 2690 + 50 BP is also consistent with other Lapita dates from Tongoleleka (see Dye 1990; Appendix A).
Research objectives for excavations at the Tongoleleka site in 1997 were specific, and required only two additional 1 x 1 m units. These unit excavations, nevertheless, were important for they provide a microstratigraphic data set for analysis, and faunal control samples that are directly comparable to other samples in Ha`apai. Radiocarbon dates from the unit excavations further support the radiocarbon chronology from 1995 excavations. Specifically they date initial site occupation to approximately 2700 BP with the plain ware site occupation continuing for another three or four centuries.