Early Preserved Polynesian Kumara Cultivations in New Zealand
Higham, T. F. G., Gumbley, W. J., Antiquity
Archaeological evidence for prehistoric gardening practices in Polynesia includes stone boundary walls, storage pits and structures, drainage systems and evidence for the modification of soil, but often the remains of horticultural practise are ephemeral. Maori developed a range of novel modifications to their traditional horticultural methods which enabled the successful introduction of the range of Polynesian cultigens into the temperate New Zealand environment, the furthest southwards these crops were introduced. They modified the soil by adding charcoal, shell and alluvial gravels to change the friability and temperature retention, and stored tubers in semi-subterannean pits for the next growing season (Jones 1991: 14-8; Challis 1976). Here, we report what we believe is the first direct archaeological evidence for the actual layout of prehistoric kumara gardens in New Zealand. Our interpretation receives support from the accounts of early Europeans in New Zealand, including Joseph Banks and William Colenso.
During archaeological excavations at the S14/ 201 site near Hamilton City, in New Zealand, we identified an area of c. 1.2 ha of modifed garden soils (originally, the garden covered 6.7 ha but has been reduced in area by residential housing). The land was designated for a new road and the topsoil was removed using supervised mechanical excavators. In one area, we discovered the intact remains of a Maori garden. We found a group of small hollows which measured 0.26-0.34 m diameter and 0.03-0.1 m depth. The hollows were identified at the interface between the A and B soil horizons and were filled with sand which had been obtained from the Hinuera Formation volcaniclastic alluvium c. 1.5-2.0 m beneath the ground surface (FIGURES 2 & 3). Three large 'borrow pits' remain of an original seven which evidence this extraction process (we estimate 550-720 cu. m of alluvium and gravels were removed from these pits). We think the hollows were probably puke (small mounds which Walsh described as `about 9 inches high and 20 in. to 24 in. in diameter, set quite close together' (in Best 1976: 149)), used for planting kumara tubers.
Two sets of hollows were identified and these abutted each other; one northern and one southern (FIGURE 1). The northern set measured 16.5x4.4 m on 8-9 rows (north-south). In this set there was an average of 0.5 m between the centres of each hollow. The southern set measured 9x5.8 m and was 14 rows wide. In this set there was an average of 0.43 m between the centres of each hollow and 0.33 m between rows. The northern set had an approximate density of 100-125 hollows per 10 sq. …