period in Hālawa Valley, from first settlement to about AD 1400, points to just such a gradual population expansion. The fossil sequence shows a similarly gradual change in subsistence as fish, important early on, decreases as a percentage of diet relative to pigs and dogs. (The record of dental caries in these animals also shows that they were fed a mostly vegetarian diet.)34 There was a gradual increase in clearing the low Acacia koa-dominated forest inland for traditional swidden cultivation, as evidenced by an increase of charcoal in the first few hundred years of occupation and the presence of fossils of land-snail species specific to that forest.35 The forest clearance set in motion a feedback process: as the forested slopes of the watershed were burned off and tilled, erosion increased, including the collapse of the steeper valley-side slopes, shown by accumulations of layered tallus, higher up in the watershed. In turn, Hawaiians retained more of these slopes with stone terracing, at narrower contours as the slopes become steeper. In this period, the interior of Hālawa was likely not settled but was extensively exploited.36
This pattern of increasing exploitation of the valley took two forms: an expansion of the land area under cultivation and an increase of yield per unit area of land already under cultivation by shortening crop and fallow cycles—a process referred to as agricultural intensification. Intensification can proceed by two routes: one, greater inputs of labor, mostly in the form of mulching and fertilizer; and two, large investments in permanent built infrastructure such as stone walls, stone-lined ditches, and drainage structures, called by archaeologists landesque capital intensification. In Hālawa, the archaeological record shows a predominance of the latter. The historical landscape, reconstructed in the excavation series, shows a gradual progress toward greater complexity, increasingly in scale and speed during the “expansion period” from AD 1400 to 1650. In the valley floors, wet kalo pondfields were constructed with stone facing and pounded clay floors and irrigated by means of ditches and headgates linked together in elaborate, reticulate systems much like the rice terraces of Bali. Hawaiian terraces differ from the Indonesian, though, in that they are squarish in plan and do not follow slope contours so closely—instead being organized into a hierarchical sequence of ditches and drops that mirrored the social hierarchy, with the most powerful people holding rights to the head of the diversion and the least powerful people the tail.37
Archibald Menzies, the Scottish surgeon and naturalist who accompanied Captain George Vancouver’s British expedition to the Pacific in 1792, described an especially fine example of Hawaiian hydraulic engineering at Waimea on the north shore of O‘ahu:
The aqueduct which waters the whole plantation is brought with much art and labor along the bottom of the rocks from this north-west branch, for here we saw it supported in its course through a narrow pass by a piece of masonry raised from the side of the river, upwards of 20 feet and facing its bank in so neat and artful a manner as would do no discredit to more scientific builders. Indeed the whole plantation is laid out with great neatness and is intersected by small elevated banks conveying little streams from the above aqueduct to flood the distant fields on each side at pleasure, by which their esculent roots are brought to such perfection, that they are the best of every kind I ever saw.38
Farther up the valley on steeper, narrower streamside slopes, partially and intermittently irrigated and dryland, or rain-fed, terraces were built for growing varieties of kalo requiring less water. On the highest and driest terraces, sweet potato, ‘uala, would have been grown in preference to kalo. Dryland terraces averaged three meters wide, following the slope contours, unlike house terraces and wet pondfields, which were squarish in plan. In addition, numerous stone-walled pens attached to the habitation structures of fields show the growing importance of animal husbandry.39 Beyond these terraces was a third agricultural form known as colluvial slope cultivation, where a mix of crops of roots, tubers, leaves, and tree fruits and nuts were tended in eroded sediments that accumulated at the base of valley slopes. These included ‘uala; yams; bananas; pia (arrowroot); breadfruit in limited locales; nonsubsistence crops such as kukui nuts (Aleurites mollucana) for lighting, medicine, and flavoring; olonā (Touchardia latifolia) for cordage; wauke (paper mulberry) for kapa cloth; and kava (Piper mythysticum) for ritual consumption. Robert Hommon described the Hawaiians’ extensive alterations in areas receiving between five hundred and one thousand millimeters of annual rainfall—barely enough for ‘uala cultivation—as a form of surface runoff management or floodwater irrigation: “Planters subtly transformed entire local landscapes with stone structures including walls, mounds, check dams, and terraces, as well as sprawling forms defying simple description, that perhaps served as combinations of dam, lithic mulch, and barage wall.”40 But the key parameter was geochemical: success in growing Polynesian crop species in Hawai‘i’s volcanic soils is limited by the presence of necessary nutrients; new and young lava substrates are insufficiently broken down by weathering, while mid- and old-age upland soils have their minerals leached away by rainfall. Pondfield irrigation is only possible on older terrains with incised valleys and clay streambed deposits. Yet in the upper reaches of those valleys, slope erosion exposes new rock, and accumulated sediments on lower slopes and toes have a higher “base saturation” of necessary minerals such as phosphorus and nitrogen than older leached slopes.41 Though purposeful landscape alteration was part of colluvial slope cultivation, it was an extensive practice, rather than an intensive one, with much lower labor inputs required than intensive pondfield cultivation, and it was inserted into the natural community instead of replacing it—an example of what is called an agroecosystem. Such landscapes have been archaeologically described in Mākaha Valley, O‘ahu, South Kohala, and Hawai‘i and in Hālawa Valley, Kawela, and Kaulaupapa Peninsula on Molokai.42 A study by Natalie Kurashima and Patrick Kirch using GIS modeling has suggested that colluvial cultivation had the potential of doubling the amount of cultivable land on Molokai over that achieved with pondfields and dry terracing.43
In the expansion period, the growing Hawaiian population moved out of the “salubrious cores” of Hālawa and other valleys into progressively more marginal areas. Taking the extensively documented Hālawa case, by 1400 the valley’s residents had expanded out from the nucleated village of the coastal dune site (from about 100 to 350 persons over this period) and established permanent homes up the valley and in several tributaries. Their dwellings were dispersed among their terraces, in loose groupings on low ridges embedded in the cultivated landscape. From 1400 to 1500, inland sites at Hālawa proliferated, meshing well with a raft of evidence from all over Hawai‘i of a demographic explosion from 1200 to 1650. Hālawa’s population was probably between 350 and 600 persons at the end of this period. The model of exponential population growth from the first canoe loads until 1450 is borne out by the archaeological record of settlement patterns showing a clear interlinking of population growth and agricultural intensification, with neither one preceding or driving the other but each sustaining the other in a kind of symbiosis.44
The kinds of changes seen in the archaeological record in Hālawa are also seen elsewhere on Molokai—including Wailau Valley and Kalaupapa Peninsula, where sites have been excavated, and all over the Hawaiian archipelago as settlement expanded and radiated into lowland areas outside the primary windward valleys. In the expansion period, Hawaiians came to be everywhere, even inhabiting the remote islets Nihoa and Necker northwest of Kaua‘i, for several hundred years. Kirch has written: “There is scarcely an area in the lowlands (if it receives greater than 500 mm rainfall and is not a steep cliff) that upon archaeological reconnaissance does not yield evidence of indigenous Polynesian agricultural use.”45
On Molokai, Hawaiians inhabited, or at least utilized, nearly everywhere below twenty-five hundred feet elevation. Adzes from Mauna Loa quarries on the West End appear at Hālawa from the earliest period. Kalo cultivators moved into the south shore valleys, building conventionally stream-irrigated pondfields in those larger valleys where water was sufficient, such as Waialua, Puko‘o, Mapulehu, Ualapu‘e, and Kamalō. In others, terraces were carved out of the swampy coastal lowlands, kept brackish by the spring flow of water percolating down through the lava from the mountain forests and surfacing at the shore where the island’s freshwater lens
