Volume 78 1969 > Volume 78, No. 1 > Economic change in New Zealand prehistory, by D. R. Simmons, p 3 - 34
ECONOMIC CHANGE IN NEW ZEALAND PREHISTORY
This paper attempts to show that the economy of prehistoric man in New Zealand underwent a series of changes which can be described in terms of response to, or control of, the ecology according to the following principles. Hunting-gathering groups are dependent on the natural resources available to them; selective utilisation and control of activities can harvest food sources but do not radically alter the ecological balance; agricultural or pastoral groups either introduce new crops or animals and alter or take advantage of the existing ecological situation to accommodate them or, by selection, give an endemic species dominance; changes in vegetation-cover or animal populations are critical to hunting-gathering groups but their effect is not so important on non-hunting groups.
The main islands of New Zealand offer a range of land form and climate stretching over some fourteen degrees of latitude between 48° S and 34° S in a fairly narrow strip rarely exceeding 250 miles in width. The main mountain range is situated down the western side of the South Island and the continuation of this in broken form in the lower part of the North Island, combined with a predominantly south westerly wind belt, makes for a range of regional variation in climate as well as restricted sub-regional and local topographical variation. The changing archipelagic stages in New Zealand's geological history, 1 the glaciations, volcanism, and plant succession, have produced local and regional variation in plant cover and animal distribution.
The ecological situations into which prehistoric man came in New Zealand were thus many and various. Whether he could control it or not, - 4 his first response was to the local situation. There are evident differences to be expected in the exploitation of inland plains or valleys, broken hill country, enclosed bays, open sandy shores, boulder banks or rocky coasts. The response, even from settlers with a similar cultural background, must have varied. No real progress can be made towards a general theory of economic change in prehistoric New Zealand unless detailed regional economic sequences have been correlated as the basis of wider analysis of variations in time and space and ideally, this should be done. In this study, however, the fine details of regional and local sequences have been largely ignored in favour of a more general characterisation in order to suggest the types of economic change and to bring out some of the dynamics affecting these changes.
In the South and most of the North Island, archaeological evidence indicates that by 1000 A.D. man was largely orientated to a coastal forest hunting existence with accessory fishing. As Green has suggested 2 the initial Polynesian settlers must have undergone a period of adaptation during which the agricultural and hunting techniques of tropical Polynesia were adapted to the new conditions, and during which the exploration of local resources was undertaken. Such an initial adaptation period is likely to have been fairly short, as the hunting-gathering techniques found to be practicable were already an integral part of the agriculture-hunting complex practised in Polynesia. If such initial settlers came with a full complement of tuber and cucurbit crops then adaptation to a seasonal cycle which included storage must have posed serious difficulties in the less favourable areas. Even in the areas not subject to severe frosts 3 it is likely that a period of adaptation was necessary before the full benefit of a predominantly agricultural economy was possible. In tropical Polynesia, ethnographic reports of the 17th, 18th and 19th centuries, and what little evidence can be inferred from food preparation tools found in archaeological sites indicates that the dominant crops were taro, breadfruit and coconuts 4 with little emphasis on kumara (or sweet potato). In New Zealand the dominant crop at the time of European contact was kumara with accessory taro, gourds and harvesting of fern root. The ubiquitous store-pit found in most North Island archaeological contexts correlates broadly with ethnographically recorded kumara storage. 5 No storage of this type is recorded either for taro or fern root. Harvesting of bracken fern root (Pteridium esculentum) is ethnographically reported as a concomitant of kumara agriculture, that is as a second growth following the clearing and use of bush land for kumara. 6 Only in the Chatham Islands 7 does clearing of bush for fern root appear to have been unaccompanied by kumara agriculture. The conversion of large areas of the Chathams from bush to fern some time after the tenth - 5 century is indicated by the pollen profiles from peat deposits. 8 The southern half of the South Island of New Zealand where kumara could not be grown could be expected to demonstrate a similar pollen profile but in fact does not do so. 9 The importance of cordyline root at least in the late prehistoric period in Otago and Southland, is documented from ethnographic reports 10 and field evidence in the form of large ovens situated on bush edges or near gullies where cordyline occurred, and in the occurrence of clumps of cordyline (ti or whanake) near settlement sites where they would not be expected to grow naturally. The use and even cultivation of ti or whanake plants for underground sucker stems is well recorded for the northern kumara areas as an accessory luxury valued for its high saccharine content. 11 In the southern area, in the late period, the three months of the year, December, January and February, were spent in cooking and drying the underground sucker stems. Both the Chathams and the southern New Zealand areas were non-agricultural until the introduction of European potatoes, yet were not without resources which ensured a carbohydrate supply. In the Chathams the clearing of karaka bush for fern must have been deliberate as karaka bush will not burn in a green state. The recorded replanting of large fern roots 12 can be characterised as semi-cropping. Similarly, the planting of ti and whanake or even the intensive harvesting of the roots in Murihiku can also be regarded as semi-cropping. In these areas, the lack, in archaeological contexts, of food pounders, which could have been used in preparing these semi-crops, makes it difficult to confirm that semi-cropping was a prehistoric practice, but few archaeological sites are favourable for the preservation of wooden objects such as fern root beaters.
A further factor to be taken into account in discussing economy is the effect of trade. At all periods in South Island prehistory, obsidian from North Island sources, such as Mayor Island in the Bay of Plenty and later, Taupo in the central North Island, was reaching even Stewart Island. In the late period (c. 1800-1840) dried kumara, cloaks, and canoes from the East Coast of the North Island were being exchanged for greenstone, white heron feathers, scented oils and preserved mutton birds. 13 The effect of these practices on the total economy was possibly not very great, yet in terms of the inter-relationship of differing ecological situations their mere occurrence is significant. Separate niches cannot be firmly established as completely independent units though they may have been relatively self contained.
Despite the variation evident in the type of basic agriculture or semi-agricultural economy possible in different parts of New Zealand, a fairly definite line can be drawn in the eighteenth century between the full agricultural areas where accessory harvesting, hunting and gathering were used and the hunting-gathering areas which used harvesting or - 6 semi-cropping as accessory practices. Except for local differences, all of the North Island, the northern South Island and Banks Peninsula were in the zone of a basic kumara agriculture economy. The southern South Island, Chatham Islands and probably most of the West Coast belonged to the zone of basic hunting-gathering. This division corresponds to the kumara growth limit which is reported as Taumutu, twenty miles south of Banks Peninsula. 14 The importance of kumara and the development of specialised storage to cope with the crop could have begun only in areas where survival of the live tuber in the winter period did not require sophisticated storage. The Auckland area, where tubers can be over-wintered in the ground in sheltered spots, sheltered coastal areas further north and certainly the Kaitaia region where plants can grow wild on sheltered northern slopes, are the most likely places for this adaptation to have taken place. The ultimate extent of kumara agriculture in the 18th century means that at some time there was an active extension of this technique to other areas. That such did take place may be inferred from the material culture of the Classic Maori culture normally associated with fortified villages, pit complexes and kumara agriculture. In 1921 Skinner 15 was able to divide New Zealand into a number of cultural areas which fell into two main groups, Northern and Southern. This basic division, while owing much to the age area theory, nevertheless is an acute appraisal of the then available ethnographic data. The area of the North Island and Skinner's Kaiapoi and Wakatu (Nelson) regions actually share a fairly homogeneous material culture. Pa, pit, adze, fish-hook, ornament, musical instrument and carving styles are all very similar. There are some regional differences but none of these are great enough to hide the close similarities which together distinguish Classic Maori culture from all other Polynesian cultures. If we assume a common point of origin for the agricultural techniques, then it is logical to assume a common point of origin in the north for the associated material culture.
Tribes still inhabiting Classic Maori areas ascribe their local origin to migratory canoes from Hawaiki. Almost all of the migration traditions refer to kumara and often taro crops being planted in the new areas, crops which were successful and continued to be planted up to the present. One tradition of the Bay of Plenty refers to an agricultural traveller who returned to Hawaiki when he found the indigenous people did not know kumara, to fetch kumara seed for them. The local food was said to be fern root. These traditions of bringing crops also mention native New Zealand trees used for food which were said to have been introduced into the area at the same time as kumara. One of these is the karaka (Corynocarpus laevigata). Much ingenuity has been used in explaining these apparent anomalies with the result that some of the canoes are now said to have called at Raoul Island in the Kermadecs on their way from Polynesia to pick up karaka. 16
A thorough check of the authenticity of the records of traditions leaves a solid core of authentic tribal traditions referring to the origin of the - 7 tribes inhabiting Bay of Plenty—Rotorua, Waikato, Hawke's Bay and Southern Taranaki—all of which state that kumara has been cultivated in those areas since they were first settled. None of these areas is suitable for growing kumara without storage in the winter. 17 This suggests that while the original ancestors of the Maori came from Polynesia, many, if not all of the canoe traditions referring to agricultural groups settling south of Auckland, actually refer to movement from Northland, which can thus be regarded as the last Hawaiki. Absolute dates obtained from traditional sources by European computation techniques are always suspect but may serve as a broad guide if subject to a number of checks. Roberton, 18 by applying fairly stringent criteria, reached a suggested series of dates for the arrival of original ancestors at the canoe landing points and their spread inland. The Arawa canoe is said, by Roberton, to have landed in the Bay of Plenty in the 13th century and to have colonised the central area by the 15th century. Likewise the Tainui people landed at Kawhia in the 13th century and spread to inland Waikato by the 15th century. The Takitimu people landed at Wairoa north of Gisborne in the late 14th century and finally conquered Hawke's Bay by the 16th-17th centuries. Similar computations were made by Roberton for other areas indicating that “major moves which set the pattern for the modern distribution of population all took place about the 16th century”. 19 While traditional information cannot be used to explain archaeological data it is significant that a logical postulate arising out of archaeological and ethnographic evidence should correlate reasonably well with authentic traditional records.
Origin traditions of tribal groups still resident in an area cannot of course be relied upon for information about any previous inhabitants in that area unless these have become an integral part of the population and tracing descent from them is important. The Takitimu immigrants into Poverty Bay in the late 14th or 15th century, probably emigrated from near Kaitaia. Movement further south was halted, as Hawke's Bay was already occupied by Ngati Awa peoples who are also said to have moved from Northland. Pa and kumara cultivation are specifically mentioned in the traditions of Ngati Kahungunu as belonging to these earlier people who were later absorbed by intermarriage and conquest into the Ngati Kahungunu in the 17th century. Close relations of the Hawke's Bay Ngati Awa still live in the Bay of Plenty. Another branch of Ngati Awa, calling themselves Ati Awa, is resident in Taranaki on the West Coast. In the South Island the Classic Maori Ngai Tahu of Canterbury, trace their origin to an early split from Ngati Kahungunu in the 17th century, a story confirmed by Ngati Kahungunu sources. They took with them all the elements of Classic Maori culture but, probably because of their isolation, their cultural heritage varies slightly from the northern forms.
Origin traditions of the Northland tribes refer to associations with various of the well known canoes of tribes living further south such as - 8 Tainui, Te Arawa, Mataatua and Horouta, but their own origin is usually traced to different canoes, Mamari, Matawhaorua, Mahuhu and others. Two accounts written down from informants in Hokianga and Kaitaia before 1855 specifically trace the origin of the southern tribes and their canoes to Northland. These were manuscripts written for Judge Martin and copied by E. Shortland. 20 They appear to be reliable evidence for a type of tradition not elsewhere recorded, in that none of the other areas have any traditions in which they claim that the origin of the various tribes of New Zealand is in their own area. A comparison of the genealogies of the North shows that the present-day tribes Ngapuhi, Rarawa, Aupouri, Ngati Whatua, all trace back to pivotal ancestors who lived about ten to twelve generations ago. Further back than this, their genealogical lines tend to coalesce not only with each other but also with the immediate pre-canoe, or in some cases post-canoe, ancestors of the more southern “fleet” canoe tribes. 21 Traditional information as to tribal movements suggests strongly that the present-day Northland tribes were able to develop and move into their late 18th century positions because of areas becoming available to them. An example of this is provided by the Ngati Whatua movement from Kaitaia to Auckland. 22 Ngati Whatua themselves probably had their origin among the ancestors of the Ngati Awa and Ati Awa groups who now occupy areas in the Bay of Plenty and Taranaki. 23 Ngati Whatua conquered Hokianga and Kaipara then moved down as far as Auckland, moving from Hokianga to Kaipara in the mid 17th century, then conquering Auckland in the late 18th century. 24 A small North Hokianga tribe, the Aupouri, then moved into the area north of Kaitaia. Field evidence for this move is that the type of pa named by Ngati Whatua as having been built by them is a quite distinctive form of ring ditch which is also associated with the Ngati Awa and Ati Awa areas. In the north, the Aupouri elders still point to the ring ditch pa in their area as former Ngati Whatua strongholds.
From the standpoint of tradition then, Northland is the only area where it is claimed that a number of the tribes originated. All the more southern areas claim descent from particular migratory canoes. Genealogically there would seem to be some substance in the Northland claim which is made more plausible by the pattern of later movement from the area.
Traditional information does point to a pattern of movement culminating in the settlement of tribal areas mainly between say the 13th and 16th centuries. These dates, of course, are only rough approximations. Archaeological documentation of the origins of these agricultural settlers is not yet available but on logical grounds one can suggest an immediate origin within New Zealand in the Northland area.
Archaeological information and C14 dating have shown that by about 1000 A.D. or soon after, Polynesians bearing a culture of Eastern Polynesian origin had explored the main coastal areas of both North and - 9 South Islands. Initial settlement of New Zealand may have preceded this exploration by a considerable time but the close similarity between the material culture of early New Zealand and that of Eastern Polynesia indicates a fairly close relationship. A speculative date of the eighth or ninth century for the first settlement seems reasonable. 25 Whether there was one, or more than one, original movement from Polynesia has always been a debatable point in New Zealand prehistory. The very similarity of the early period material culture in all areas points to a single point of origin if not to a single immigration. However, it is simpler to assume a single rather small immigration. 26 The original settlers brought with them dogs, rats and tuber crops but as far as is known, no pigs, poultry or breadfruit. Except in the far north planting or keeping tubers alive is likely to have been somewhat chancy, firstly because of the difficulties of mastering a cooler environment, and secondly because agricultural settlers are likely to have come at a time of the year when sufficient food and seed tubers were available to ensure survival and continuation. They are therefore likely to have arrived in New Zealand in the high summer period about December or January. If New Zealand had been in the tropics this would have enabled them to plant and harvest a crop before the cooler weather set in. But New Zealand conditions normally require tuber planting to be made in October or early November. Thus, there were many difficulties facing agricultural groups. From the dynamics of the later suggested spread of kumara agriculture, it would seem unlikely that agriculture played any part at all in the early exploration and colonisation of most of New Zealand. Even taro plants would have had difficulty in surviving their first winter if planted late, south of Auckland.
REVIEW OF ARCHAEOLOGICAL EVIDENCE
These divisions are neither absolute nor do they necessarily occur in all areas in this order or at the dates suggested; rather the developments outlined are thought of as having reached a peak 27 between the dates suggested.- 10 - 11
In the southern South Island an initial hunting-gathering economy was probably practised during the period of adjustment after the arrival of settlers. Easily accessible sea shore estuaries or river situations would have attracted sea-borne migrants. The first type of economy for which there is definite evidence was orientated toward the coastal forest with accessory fishing and sea mammal hunting. The necessary knowledge and techniques for hunting birds developed. The most important birds hunted were a full range of moa genera and species but many other genera and species of forest, water or sea birds were also hunted. There is little doubt that at this time, about 1000 A.D., the whole of the South Island was heavily forested with either podocarp or beech. 29 Fishing gear, while it has important resemblances to Island Polynesia, was not nearly as important as would be expected in coastal sites. There are increasing differences between regions evident in the material culture of the latter part of the early period when a restricted range of moa and other birds were being hunted, but coastal forest hunting was still the basic economy.
The discovery and exploitation of raw materials for flake knives, one of the specialised tools developed in the early and middle periods, necessitated some exploration inland, up the main rivers or valleys, with food being obtained from forest hunting and eeling.
Between A.D. 1300 and 1450 there was a marked change in ecology in the eastern South Island. The area of forest began to shrink and the forest started to withdraw from the coast. Many birds including moa became scarce and hunting activities were concentrated in the surviving areas of forest in Western Southland, South Otago, Otago Peninsula and coastal North Otago. The birds were hunted in the restricted forest areas but quickly became non-viable as a food source so that restricted forest hunting became counter-balanced by sea-shore resources, fish, some shellfish and seals. Large numbers of seal were hunted, indicating a fairly large population turning to them, but soon they in turn became scarce and increasing reliance was placed on fish and shellfish with greater emphasis on fishing gear.
After about 1450 A.D. the economy was almost entirely orientated to shore fishing and gathering. Local resources, that is within a ten mile radius of a site, were utilised for tools and food. Layers and sites of this period (intermediate) are characterised by a build-up of shell midden, but the sites are fairly small indicating either small family groups or a reduced total population. Early period sites are often very large, middle period sites still big, but intermediate period ones are often quite small. This difference in size is probably a fairly simple reflection of the availability of food. When food is plentiful, hunting groups can come together in large agglomerations but must scatter over a wide territory when it is not. Scarcity or otherwise may be seasonal. There is little evidence available on the seasonal exploitation cycle in southern New Zealand. Judging by the finds of eggs used as water bottles and from scattered egg- - 12 shell, large sites where moa were hunted seem to be associated with nesting areas. However, in the southern South Island it does seem possible to suggest a correlation between period and abundance of food. The early and middle periods before the withdrawal of the forest from the eastern coast were periods of abundance. The intermediate period with restricted forest hunting was one of economic hardship. A large population turning to seals quickly reduces the supply of these animals so that they remain economically significant only for rather short periods. The large amount of shell midden and the extreme regionality of site areas and culture 30 indicates fairly small local populations utilising each area and unable to penetrate outside them. Sites of this period are not numerous and everything points to a much smaller total population. In 1800 A.D. the area south of Bank's Peninsula did not carry more than about 2,500-3,000 people. In 1844 Shortland's census gives a population of 1927 31 which is very small and may have been affected by introduced diseases. The population after about 1550 A.D. appears to have been drastically reduced either by lack of food or possibly by emigration to more favourable areas.
The basic shore fishing economy of the intermediate period was associated with seasonal exploitation of inland areas. As the forest with-drew to the western side of the island the hunters probably followed it inland. For the Mackenzie country, Ambrose 32 notes that forest and the moa ceased to exist after the incursion of fire and man in the 12th and 13th centuries A.D. Lack of forest-provided foods and the severe climate of the central Otago region was probably against any permanent habitation of the interior. In the 13th or 14th century, sites were established on the lakes and exploitation of Routeburn greenstone became important. In the 15-16th century the establishment of a “village” of twenty houses with paving and paved pathways is evidence of an extended use of the interior; the attractions of the area were eels and birds of the lake regions, trees for canoes, greenstone from the Routeburn and later, bowenite from Anita Bay. Judging by the artifact types, exploitation of the abundant birds, fish and sea mammals of Fiordland occurred throughout the area by the 17th century. There is some evidence for 16th century penetration but this is not sufficient to suggest exploitation. At any time the Fiordland area probably entered into the total economy only as a seasonal station, not necessarily annual, which was used by groups from Southland by way of the southern lakes and coast, from Otago by way of the lakes and Haast Pass and from Westland along the coast.
The type of basic economy characterised by shore fishing continued in Murihiku until the late 18th century when it was replaced by potato agriculture imported from the northern South Island. The importation of the Classic Maori form of material culture and the introduction of many distinctive forms of artifact such as pa, which occurs at this time, included the introduction of potatoes although this preceded actual - 13 European contact in the South. 33 The distribution of some of the Classic Maori settlements away from easily accessible landings suggests that any associated potato agriculture was purely for subsistence as part of a total economy which included seasonal stations. Exploitation of cabbage tree on the coast and along the routes to the inland lakes and greenstone probably repeated that of the earlier period. Villages established on the central lakes included one on Hawea, two on Wanaka, two on Wakatipu (one in the form of a pa), and others on Te Anau, Manapouri and Monowai. These settlements, which lasted until 1844, were all year round base villages made possible by potato agriculture. Parties still came in from the coast but only in the summer months. People at the palisaded village of Murdering Beach at the Heads of Otago Harbour were certainly manufacturing Routeburn greenstone into hei-tiki for trade to the north. The presence of a few European objects in this site of 40 houses, which was probably burnt down in 1817, is significant as they included a Cook medal given at Queen Charlotte Sound in 1772 and a Brazilian coin dated 1806. 34 The whaler, Kelly, who is said to have destroyed the settlement, landed to trade for potatoes. There is little to indicate any direct European contact with the village before this date. The lateness of Classic Maori penetration in the South is indicated by the fact that Murihiku settlements included only five coastal pa and one inland pa at the head of Wakatipu near the greenstone source and route. Access to greenstone must have been one of the attractions of Murihiku once potatoes made agriculture possible and the presence of Europeans eager to trade for greenstone made easy access to a supply an important feature of such a trade. The rapid introduction of a potato crop to the area is evident from the report of Stewart who in 1809 saw many acres growing at Bluff. 35 Although early Classic Maori sites in Otago and Southland show no evidence of European contact, it cannot be doubted that indirect contact was producing marked effects on the economy of the southern region at this time. Sporadic trading occurred from 1817 onwards though it was not until 1825 that the lives of Europeans were not endangered by landing. 36 By 1831 there was a regular trade in potatoes to Sydney which required 200-500 baskets three times a year. From 1831 the sale of potatoes and the beginnings of a European money type trading economy based on supplying ships or settlements in Australia is evident. In 1844 the paramount chief of the area, Tuhawaiki, was buying stores and reselling them at enhanced prices to ships and Maoris. 37 Maori and European population concentrated around the two main harbours, Otago and central Foveaux Strait. No sheltered harbours for sailing ships existed north of Otago Harbour though there were anchorages at Waikouaiti and Moeraki which were sites of Maori villages and whaling stations. By 1845 whaling needs were becoming less but the - 14 arrival of settlers at Waikouaiti from 1838 and Dunedin from 1848 provided a good market for Maori farm produce and fishing.
In the early coastal hunting period, the economic development of the northern part of the South Island was similar to that of Murihiku. Regional distribution of moa species is not well documented but it is possible that in some areas the full range of moa was, even at this time, no longer present. Certainly there are few sites recorded with the range of moa found in Otago. This could also suggest that earlier sites exist but have not been found. The Middle Period economy was one of forest hunting with a restricted range of moa and a variety of other forest birds. As in Murihiku, seashore resources were turned to when forest withdrew. The shore fishing economy did not last as long in the northern South Island as it did in the southern area. Some time in the 17th or possibly late 16th century there was a movement into the area of people from the southern and eastern North Island who brought with them an early form of Classic Maori culture together with kumara agriculture. The limit of kumara growth was Taumutu just south of Bank's Peninsula. Pits, presumably for kumara storage, are recorded from Golden Bay and northern Marlborough with further groups at Waipapa, Kaikoura, Pariwhakatau and at Pa Bay, Bank's Peninsula. 38 What are identified as gravel pits at Temuka are unconvincing in the absence of pits or any made kumara soils in the area. Attempts to grow kumara at Arowhenua village, just north of Temuka about 20 years ago failed by not producing a consistent crop. 39
Wellman in his survey of occupation layers on D'Urville Island 40 recorded the characteristic content of five different types of layer which he organised into two series based on the younger series containing abundant barracouta bones, few argillite flakes and no moa, while the older series contained broken moa bones, many flakes of baked argillite and rare obsidian. A layer of pebbles, which Wellman considers transported, on three sites on Greville Harbour sandbar provide possible evidence for kumara agriculture associated with the older series. 41 No other definite characteristics of kumara soils seem to be present in the older series or just beneath them though the pebbles usually rest on a sandy soil rather than sand. Though this is not brought out by Wellman, the younger series of layers contain more definite evidence of agriculture in the form of dark gravelly soil, or gravel and charcoal, 42 which while not easily correlated with Wellman's definition of the younger series, occur at or near pa sites and appear to be quite typical of later occupation. The characteristic change in economy typified by the presence of much - 15 bone or much shellfish is evident from Wellman's descriptions and in fact the main characteristics of the younger series are abundant shellfish and fish rather than just barracouta. As he remarks for the Opotiki Bay section “The upward increase in proportion of shellfish to bones indicates a significant deterioration in diet during the early occupation period”. 43 The possible presence of agriculture with the earliest occupation needs further investigation but could indicate that D'Urville Island was important, as Wellman suggests, 44 as a secondary centre for kumara agriculture.
The western part of the D'Urville-Nelson area has been surveyed by Wilkes, 45 and a site on the west coast at the mouth of the Heaphy River has been excavated. 46 Wilkes and Scarlett summarise the site survey by saying, “it seems that along the western coast of Nelson there may have been a relatively large population in earlier times when moa was available, but there was very little, if any, permanent settlement in the years prior to and during pakeha exploration”. 47
The Heaphy River mouth site and another site in the district are typically sealed beneath mudflow deposits on which there are or have been stands of nikau palms and broadleaf trees. 48 The general area vegetation is lowland broadleaf forest with a rata-kamahi element along the coast. The occupation layer at the Heaphy site was a 10-24 in. layer of charcoal blackened sand with some lenses of mussel shell. 49 Shell was practically absent in the site except in the shell dumps, some of which contained fairly abundant small bird and fish bone. Seal and moa bone was concentrated beneath the dumps. The main shell content was black mussel (Mytilus planulatus canaliculus) and some ribbed mussel (Aucolacomya maoriana). One very small dump largely contained pipi (Amphidesma australe). According to Wilkes and Scarlett, shellfish dumps usually seem to occur more in the upper part of the occupation layer whereas the bone material is concentrated in the two or three inches above the bottom where it is often trodden into the ground surface. Some bone was also found in the shell refuse from ovens. 50 Small Anomalopteryx moa and fur seal provided the most numerous bones. Smaller forest birds, penguins and sea-shore birds and dog provided the rest of the bone material. These food resources seem typical of the coastal hunting economy. Raw materials used are drawn from near and far indicating a wide area of exploitation or trading, stretching from Milford Sound and Lake Wakatipu in Murihiku, to Marlborough in the east and the northern coast of Nelson. Obsidian from the Bay of Plenty was almost certainly traded. Artifacts are early to middle period in type. A single radio carbon date from “one of the shell heaps” 51 gave a date - 16 of A.D. 1518 ± 70 which is much too late for the earlier part of the occupation. The general division with bone in the bottom, shell on top, would suggest that the occupation layer is not as homogeneous as it appears. From the description it is possible to isolate three main economic phases:
That the single occupation layer is the result of more than one phase of activity is strongly suggested by the sedimentary history given 52 in which three phases of occupation are mentioned. The first phase on top of a mudflow deposit, is separated by windblown sand from the second phase, which is in turn sealed in by a mudflow on top of which the third phase people lived. A broadleaf forest soil formed over this deposit in most areas. The three phases of occupation probably correlate with the three phases of economic activity but until further information is available little more can be said.
The early economy of South Taranaki is very similar to that described for the South Island. Sites are typically coastal with a range of moa and other extinct birds. Buist, for Kaupokonui, notes “the sudden change of diet from the upper to the middle level”. 53 Lower and middle levels at this site were similar in the range of birds hunted and “as at Ohawe, shellfish were not greatly exploited at the same time” (as moa-hunting). 54 The upper level by contrast contained “fish bone, a few seashells, dog bones and a large amount of dog droppings”. 55 As he remarks 56 “the moa and other extinct birds were no longer available for consumption and other food had to be sought”.
That the southern North Island underwent a somewhat similar restriction of forest, at least temporarily, is indicated by pollen profile evidence from the Haurangi Mountains 57 and lowland North Island. 58
In North Taranaki, Parker 59 has defined, at Kumara-Kaiamo near Urenui, a site belonging to an early agricultural technology which he equates with a similar site at Opito in Coromandel dated to the 14th century A.D. The early pit complexes are followed by others of Classic Maori type associated with a pa. The final occupation of the site is associated with European items and probably dates from between 1820 and 1830.- 17
In the Wellington area the lower layers of a site at Makara containing five species of moa are overlain by a large shell midden containing modified early type artifacts. 60 At a Foxton midden site on the Motuiti dune formation, 61 there were found to be a series of layers; a bottom layer predominantly of bones including a layer mainly of cockle shells (Chione stutchburyi or Amphibola crenata). Both shell layers also included “a considerable portion of bone including seal”. A layer of sterile sand separated this group from the upper group of a thick cockle layer, a tuatua layer and the upper layer of black material containing scattered shell and stones. The lower bone layers of the mound were associated with structural features. Associated artifacts were early types.
The few pit complexes in the Wellington area are probably all associated with historic potato cultivation. McFadgen 62 suggests that the Wellington Harbour region was not favourable for kumara and was not occupied by kumara agriculturists. This does not apply to the coastal regions further north.
Sites where moa were hunted are fairly rare on the East Coast, but are known to exist in Palliser Bay and along the Wairarapa Coast. Similarly, some sites containing moa are known to exist in Hawke's Bay at Ocean Beach, Ahuriri Lagoon and Tolaga Bay but the whole of this area has been little studied.
Wellman 63 has recorded a number of types of occupation layers along the East Coast where an upper series of occupation soils or shell midden is separated from lower layers by Kaharoa ash dated to about 1300 A.D. The lower series contain more obsidian, flint, rat and seal bones and some shell, and are often mixed with or resting upon Loisels pumice. In some sections charcoal and bones occur just below the Loisels pumice layer. 64 This has led Wellman to suggest a population peak building up to the time of the Loisels eruption (13th century) and continuing until about the time of the Kaharoa ash shower (14th century). The population then declined “due to possible over-hunting and to a possible inland migration prior to the growth of the latter population that was based essentially on kumara”. 65 Wellman does not record seeing any moa bone, but the reason may be, as Hartree 66 suggests, that there never was a very extensive moa population in this area. Ample birds and sea mammals would be available at this time.
Bay of Plenty
The Bay of Plenty has received some attention from Green 67, Wellman(68) and Pullar 68. Wellman's information is the most detailed. - 18 There is a broad correlation with other areas in the presence or absence of shell. In one case, occupation was at or before the time of the deposition of Loisels pumice, but the information is at present too meagre for any further inferences.
The Western Bay of Plenty from Tauranga to Coromandel seems to form a single area in that fairly early occupation is known from Waihi Beach to near the tip of Coromandel Peninsula. Archaeological evidence from this area has been summarised by Green. 69 The economic aspects have been used as one determinant of the cultural sequence of the Auckland Province. Green characterises the economy of his Settlement Phase (circa 900-1100 A.D.), as “primary dependence on the hunting of a full range of now extinct avifauna including most species of moa; an equally heavy use of sea mammals, fish and the rocky shore shellfish found in abundance and of large size”. 70
Green's Developmental Phase from about 1100-1300 A.D., includes “intensive exploitation of selected species of moa and remaining avifauna . . . heavy exploitation of the marine environment, especially sea mammals, fish and rocky-shore shellfish”. 71 The Experimental Phase, circa 1350-1450 A.D., depended on “few species of moa surviving or being hunted, except inland; more use of mud flat species of shellfish than formerly”. 72 The change in economy as noted is in the kinds of shellfish gathered, the deposition of compact and homogeneous middens and the reduction in artifactual materials suggesting specialised areas. The homogeneous middens produce some fish bone but little other bone, “suggesting that hunting now plays a minor role”. 73 Green infers that agriculture became important during this changeover. The following summarises his conclusions. 74
A probable 15th century pa site at Kauri Point, Tauranga, can be compared with the Skippers Ridge site at Coromandel containing buttress pits, presumably for storing kumara, which can be dated to the 14th century, and the Opito site on the beach below also dated to the 14th century. The latter sites both contained archaic adzes and other material of a diagnostic character. A site at Sarah's Gully has a similar economic sequence. Three species of moa and much other bird bone occur in the lower layers and Green notes a change from rocky shore to mudflat species of shellfish from lower to upper layers though shellfish are not very important in the lower layers. The changes are similar to those at Skippers Ridge. In the lower layers at Skippers Ridge and in the middle layers of Sarah's Gully storage pits appear and parallel the economic change. Similar types of pit underlie the Kauri Point, Great Mercury Island and Harataonga sequences and, as Parker has shown, 75 can be - 19 equated with similar early storage structures underlying more recognisable Classic Maori structures in Taranaki.
The evidence so far available suggests a spread of kumara agriculture to either side of the North Island south of Auckland by the 15th century A.D. 76 Green, postulating a deteriorating climate which led to the need for storage, calls this phase Experimental. Any climate deterioration was minimal, not exceeding a one degree average lowering of temperature. 77 Even before this “deterioration” kumara agriculture without storage would have been nearly if not quite impossible. Local area adaptation would seem very unlikely.
The Central Plain
Very little is known archaeologically of the area though field recording suggests that the most intensive use was after the introduction of white potato which enabled a more permanent exploitation. 78 This pattern is in some respects similar to that of inland Otago or Wellington. Moa hunting was carried on at a site in Tokoroa 79 where their remains have been found associated with archaic artifacts. At Whakamoenga Cave, Taupo, small moa were hunted by a people coming from the coast 80 possibly, as at Tokoroa, just before a change in forest cover which affected a large part of the inland region and was probably associated with Maori occupation of the area some 400-500 years ago. 81
The coastal region from Taranaki to Auckland has been examined for exposed Holocene sections by Wellman. 82 Artifact finds indicate that coastal hunting could have taken place at river and harbour mouths but as yet only Wellman's coastal surveys are available. At the mouth of the Onaero Beach, Wellman notes occupation underlying Burrel ash erupted from Egmont about 1650. 83 Newall ash from the same source lies above a black beach sand with abundant charcoal and burnt stones. Druce suggests 84 that before approximately 1500 A.D. there must have been a period of quiescence on Mount Egmont lasting from several years to many hundreds of years. In the Waikato the sections described by Wellman 85 are well formed but inconclusive, ranging from layers with shells, charcoal and obsidian, to sandy soil with charcoal, presumably from cultivation or clearing. Specialised shell middens such as those described by Wellman seem to be comparatively late. Earlier occupation is known to have existed at the mouth of Raglan Harbour and probably also at Aotea and Kawhia, but typically it is situated on the back dunes and has not yet been investigated archaeologically. Similar sites on the - 20 South Manukau Head yielded much bone of dog and Dinornis species moa. 86
The offshore islands of the Hauraki Gulf, Great Barrier, Motutapu, Ponui, and the inshore areas of Devonport, Buckland's Beach and Howick are known to have been occupied at an early date. Harataonga site on Great Barrier Island was excavated by Spring-Rice. 87 This beach site contained moa, fur seal, dog, a wide variety of sea and forest birds, and rocky shore shellfish. A site from under the Rangitoto ash of 1200 A.D. on Motutapu is known but details are not yet available. A seasonal site at the entrance of a lagoon at Pig Bay, Motutapu, is dated between the Rangitoto ash and A.D. 1670 ± 40. 88 The various occupation layers are sandwiched between alluvial beds from the lagoon. The occupation in the lower levels is a “black greasy layer” likened to those in the moahunting sites of South Otago. 89 Bird bone including some moa, sea mammal and fish bones characterise the lower layers while the uppermost layer is a “few thin middens of fish bone and shell”. 90 The lower layer contained evidence of hunting and fishing and some gathering, with dog probably important as a food. The upper layers have a much more noticeable shell content intermingled with fish bones and scales. A distinct upper layer of shell continues along the present beach front. A site on Ponui Island 91 consisted of two occupation layers: an upper agricultural type gravel soil mixed with charcoal, sand and shell, a lower layer containing moa (Dinornis), bird bone, whale or seal bone, and fish bone associated with early artifacts, and a pit presumably for storage of a crop. Artifact distribution suggests that similar sites existed on the mainland areas mentioned previously but the sites have either not been excavated or have been destroyed.
Occupation on the east coast, north of Auckland, was reported for the Pataua River area by Thorne in 1875. 92 The descriptions make it fairly certain that moa and man were associated with a range of early artifacts. Thorne distinguishes between deposits containing moa and shell middens. Wellman's sections for the same coast 93 and further north indicate that agricultural activity can possibly be recognised in the form of abundant charcoal in formed soils below Loisels pumice of the 14th century, a situation repeated at Onewhero, Cable Bay, Whatuwhiwhi, Matai Bay, Te Pua, Te Hapua and Twilight where well defined occupation layers were located under Loisels pumice. 94 An interesting feature is that - 21 described for Te Hapua with occupation following an old soil with kauri stumps (Agathis australis) succesded by an occupation layer with Loisels pumice at the base, itself overlain by a layer containing abundant Loisels pumice. If the interpretation of Wellman's sections is correct, agricultural soils appear much earlier in the far north than elsewhere. Robinson's reconnaissance 95 of Doubtless Bay points to the pattern in some parts of the beach, of a back dune series with midden of moa bones and shell and a forward dune series with heavy shell midden but no moa. Sections on display in Auckland Museum takem from the excavation at Mount Camel, Houhora, indicate that important economic information on early costal hunting in the area will be forthcoming from this site, as moa, dolphin, seal and fish are recorded in the midden 96 together with a wide range of very early artifacts. There is also some indication of a change in economy which is, superficially at least, similar to that noted for other areas. 97 The western shores of Northland are little known archaeologically. Earluy sites are reported at the Heads of the Hokianga, Kaipara and Manukau Harbours. North Hokianga does not seem to have attractive as a settlement area until about the sixteenth century as the rugged interior contains no evidence of settlement expet along the Mangamuka River.
In terms of the distribution pattern of early East Polynesian artifact types, the utilisation of baked argillite as a dominant raw mnaterial for stone tools and the great similarities present in both artifact types and site types, it is logical to suggest small, very mobil sea-borne bands of early settlers. A particular type of baked argillite which may have been from sources near D'Urville Island in the Cook Strait region is found made into almost identical adze forms from the southern tip of the South Island to the northern tip of the North Island. Distribution of the particular adze types in argillite tends to be at or near good river or estuary mouths or on sheltered beaches; in fact where a canoe would be expected to land. The few exceptions are on navigable rivers leading inland. It is thus necessary to keep in mind that all the early sites may only be seasonal aspect of one total culture. However, until much firmer and more complete evidence is available for definitely associating sites and documenting movement of this kind, little more than the possibility of such an association can be stated. Alternative hypotheses may prove to be equally valid. Initial settlement may have centred on Cool Strait with a quick dispersal, ort D'Urille-Nelson may have have been a secondary centre. The economy of the settlement and early period peoples throughout New zealand was coastal hunting with use of forest, sea and other resources. Whale, dolphin, seal, fish, moa and other forest and shore - 22 birds, together with rocky shore shellfish and fish were utilised as well as the imported dog and rat.
In both the South and North Islands there is good evidence for the development of regional aspects of New Zealand East Polynesian during the middle period. The economy was still coastal hunting but the best reported food source, moa, was present in decreasing numbers and restricted genera and species. Typically, the character of the layers changes from black with bones to layers containing both bones and shell. In the South Island the change of vegetation cover and consequent loss of the coastal forest explains the change in economy. In the North Island a similar change, from coastal hunting to shore fishing, occurred in all areas. In the extreme north, where agricultural soils appear much earlier than elsewhere, and in the Auckland Isthmus, Coromandel and Taranaki areas this may have resulted from a change to agriculture but the restriction on the species of avifauna in Coromandel for instance, occurs before any inferred effect of an agricultural economy. Agriculture in the North Island mitigated the long term effect of turning to a shore fishing economy, which in the South Island probably led to a drastic loss of population. In the North Island there was no widespread succession of forest by grassland but there is evidence from archaeological sites that a change of some sort took place in the ecology, which affected all the lowland areas and reduced the available food supply obtained from the forest. The spread of agriculture and associated pit storage enabled the change in ecology to be bridged so that the North Island population not only remained stable but grew. For this reason, sites of the late period are much more numerous than those of any earlier period. The reverse is true for the Murihiku area where early sites are commoner than late.
There are thus two main problems to be considered in more detail, the vegetation history of the North Island, and the spread of agriculture.
It is clear from the available archaeological information for the North Island, that a change in general economic orientation took place there between the 14th and 16th centuries and had an effect similar to the vast ecological change resulting from the change in forest cover in the South Island. In the lower half of the North Island this does not seem to have resulted from the introduction of agriculture. One of the criteria for late middens of the agricultural period appears to be specialisation in exploitation of the available resources. Earlier middens can be distinguished by the range of genera utilised. However, there is still the distinction between sites or layers containing a range of birds, including moa, shellfish and fish, and similar sites which do not include the range of birds but instead have rather more fish and shellfish. This has led excavators to refer to “bone layers” and “shell layers”. In the South Island the withdrawal of the forest from the eastern side correlates with a change in economy. No withdrawal of the forest can be invoked to explain the data for the North Island. If the vegetation cover remained the same and there was no catastrophic change in climate, then it is - 23 difficult to find any reason for moa, and other birds, forest snails, or tuatara becoming either extinct or restricted in range. Fleming remarks that “we find the record disturbed by the effect of man's arrival and spread just about the time of the beginning of the last warm period in Europe. The drastic consequences of his coming are now certain and the cherished principle of economy of hypotheses prevents us from seeking climatic causes”. 98 Archaeological evidence suggests that many birds became scarce in the lowland areas even though lowland forest was still present and conditions were apparently favourable. The explanation for this can almost certainly be found in the replacement of former podocarp forest by broadleaf forest.
An important contribution to vegetation history studies was by Leonard Cockayne in 1928. He notes that “the most important principle underlying forest succession in New Zealand is the relation of the different species to light”. 99 He goes on to categorise “two main types of associations which are important in the life history of New Zealand forest, those composed respectively of podocarps and broad leaved dicotylous trees”. 100 Podocarps are light demanding and so enter a forest earlier than other species. 101 Dicotylous trees are more tolerant of shade and so can grow under podocarps but produce a dense shade in which podocarps cannot germinate. 102 Thus podocarp forest is gradually turned into dicotylous forest. Cockayne quotes 103 the work of Wilson who found that “the kauri and podocarps were being replaced by broad-leaved dicotylous trees”. Distribution of the podocarp forest is now largely montane whereas dicotylous forest occupies the lowland and some of the montane regions. 104 More recently, Robbins 105 has argued strongly that “in cooler geological times generally, podocarp forest dominated in New Zealand. . . . Especially did they follow over large peri-glacial areas on the retreat of the glaciers. Shade-tolerating dicotylous trees grew up under the podocarps and these, not a changing climate, have progressively stifled the natural regeneration of the podocarps”. 106 The coastal areas (Cockayne's lowland region), are today occupied by various types of forest; these are kauri forests in the north of the North Island, beech forest in the southern South Island and mixed podocarp-broadleaf forest and broadleaf forest on lowland North Island areas. Robbins' explanation of the mixed forest of the present day is that it is a fusion of two forest types which in themselves are quite distinct. 107 He suggests, as Cockayne did, that there is a slow and gradual rise to dominance of one forest type over another. This process is largely complete on the coastal sector of New Zealand but much less advanced in the - 24 montane areas. 108 That podocarps once dominated many areas of forest is evident from pollen studies. Moar has demonstrated this change for the western Ruahines. 109 In five peat samples at the time of the Waimihia pumice shower at about 2000 B.C., podocarp pollen dominated. By the time of the Taupo ash shower of about 150 A.D. podocarps were still dominant but the broadleaf element was much more important. At the end of the peat deposition, broadleaf pollen was dominant. Harris notes 110 that an advance of beech (broadleaf) forest seems to have begun at least 2,000 years ago. He discusses pollen profiles from the Hauraki Plains, Hamilton, Wellington, Lower Hutt and Lake Monk and shows a steady rate of change from a dominant podocarp forest to mixed podocarp-broadleaf as in Hamilton, or to broadleaf forest in other areas. 111 Elder also comments 112 on the change, “Recently podocarps have failed to maintain themselves in the surviving lowland forest except under special conditions where they may regenerate vigorously as in kahikatea semi-swamp forest”.
That this change is part of a much wider succession is shown by the presence of podocarp pollen in the profiles from the Chatham Islands 113 and from the Auckland Islands. 114 These are both areas whose forests are today composed of broadleaf only, 115 though in other respects their fauna and flora are largely similar in genera and species to those of the mainland of New Zealand. This means that if podocarps were present in the Chatham Islands and the Auckland Islands they have now been replaced by broadleaf. Wind drifting could be an explanation for the presence of pollen but that such an explanation can be applied to all occurrences is unlikely because of the presence of buried podocarp species timber in the Chatham Island swamps 116 and also of soils closely parallel in structure to those formed under podocarp on the mainland of New Zealand. 117
The hand of man in inducing changes in vegetation cover has often been emphasized but it is necessary to bear in mind, first that the hand must be deliberate if local conditions have not produced an unstable forest, since it is impossible to burn standing virgin subtropical rainforest, 118 second that the total effect of Polynesian man on the New Zealand forest was slight 119 and finally that cleared areas would revert to the original plant covering 120 if no more than artificial clearing were involved.
Evidence for the association of moa and many other forest birds with forest (and especially podocarp dominated forest) in the South Island, - 25 was presented in a recent paper 121 and will not be repeated here. It seems very likely that many of the extinct ground birds fed on podocarp berries. Only 13 per cent of the trees in a broadleaf forest bear berries or drupes which are attractive to birds. 122 Certainly the range of shrubs and undergrowth on the more open podocarp forest floor is much greater than in a broadleaf forest. It is significant that the plant species listed 123 as being contained in the crops of moa from Pyramid Valley in the South Island are mostly found in podocarp or podocarp-broadleaf association. 124
There is no information on the feeding habits of moa, or any of the other extinct birds in the North Island, but the restriction of range of many species evident from sub-fossil or archaeological records suggests a restriction of food resources, which may correlate with forest succession of the type described above. In this context it is interesting to note that many of the birds of highland New Guinea live on podocarp seeds and become rare or extinct when the natives clear an existing podocarp stand for agriculture. Podocarp forest in New Guinea is described as “tender” as it does not regenerate easily after clearing, and is replaced by other more vigorous species. 125 In Fiji as long ago as 1903-9 Guppy 126 suggested that the present podocarps there are relics, having been over-whelmed by tropical lowland broadleaf forest. Pollen profiles from Chile suggest a broadly similar succession involving Weinmannia, Nothofagus and Podocarpus species. 127 If an association can be assumed between podocarp forest and bird life, the information from vegetation studies correlates reasonably well with archaeological evidence. From the foregoing discussion and the co-incidence of the time factor involved there is some justification for doing this.
The Spread of Agriculture
The effect of a changing ecology can be seen in the sites from the southern part of the North Island. In the South Island a drop in the human population following loss of the eastern forest can be deduced from site size. From quantitative artifact distributions we can say that a similar though not as marked a drop occurred in the southern North Island but there seems to have been little effect on the northern half of the North Island where artifacts occur in increasing numbers. The explanation seems to lie in the introduction of agriculture, which allowed a certain measure of ecological control. The little archaeological evidence for the beginnings of agriculture outside of the north is from Opito in Coromandel and Kumara-Kaiamo in Taranaki. This material has been summarised by Green 128 and Groube 129, and was discussed earlier in - 26 this paper. Structures analogous to those of Opito and Kumara-Kaiamo were uncovered in the lowest layer at Kauri Point Pa, 130 where they are associated with a defensive earthwork. From this time on, the pa as the main political unit, was the focal point of exploitation possibly “founded upon simple geographical bases, such as the valley system”. 131 Kauri Point is important, however, because its early pit complexes are plausibly associated with the beginnings of a swamp sequence in which a series of wooden combs were found. 132 Wooden combs are essentially Classic Maori items but the decoration of the stratigraphically earlier combs is much more geometric and more “Polynesian” than that of combs from the end of the series which are almost identical to those collected in 1769-1773 by Cook. 133
In the early layers of the main Kauri Point site the evidence for fortification is followed by a series of defensive works culminating in a ring ditch pa which, according to traditional information, was probably built around 1800-1820. Excavations on fortified volcanic cone pa in Auckland have been rescue digs but have provided important evidence for the building of these immense forts. Mount Wellington 134 yielded a date of 1430 A.D. and a bewildering series of store pits. Mount Roskill pa was not dated but established “a fine sequence of intercutting pits of various sizes and posthole arrangements”. 135 As yet, sites with “early” type store pits such as at Opito and Kumara-Kaiamo are few but the next group which seem to follow these are numerous and often associated with defences. This in turn suggests an expanding population exerting pressure on a resource, in this case probably either suitable agricultural land or the stored kumara after harvesting.
Most available information on the economy of the agricultural groups was recorded at the time of European contact. Specialised shell middens have already been mentioned as typical. Taking the observations of some of the early visitors and adding those of missionaries it is possible to produce a chart of which foods were being actively sought, planted or harvested at various times in the northern area. (See Figure 2.) These indicate that seasonal occupations were very much a feature of Maori life at the period, giving a settlement pattern which Groube has defined as “temporary seasonal camps for economic pursuits, plus, perhaps, some sort of semi-permanent fortification or village . . . the pattern suggested . . . is of a flexible settlement pattern, centred around a pa or possibly a number of pas, which are the hub of a mobile population”. 136
Agriculture was of the slash and burn type with a plot being utilized for two or possibly three years before abandonment. 137 The only record of the cycle of return is the figure of fifteen years mentioned by Felton- 27
FIGURE 2- 28
SEASONAL HARVESTING OF FOOD RESOURCES 138
Mathew in 1840 for the Auckland area. 139 K. Shawcross 140 has recently argued that fern root agriculture was more important in the total economy of the Bay of Islands than any other practice and suggests that land was cleared for fern rather than kumara. That fern root was an important semi-crop is undoubted but the weight of information recorded from Maoris is against the elevation of fern as the dominant crop. Maori reports place fern root growing as a follow-on of bush clearing and cropping of kumara. Apart from any other considerations, previously dug land produces much larger fern rhizomes.
Post European Agriculture
After European contact the introduction of new crops and domestic animals led to a further series of economic changes. Kumara agriculture was the subsistence type and this was still true when potatoes were first introduced. Potatoes grew and matured before kumara and so replaced plant foods normally utilised during November through to March. Taro were dug in October to bridge the gap between the last of the stored kumara and potatoes; fern root was still an important semi-crop but became less so in later years. 141 The first change then was to kumara-potato subsistence agriculture. With increasing sealing and whaling activity on the coasts, potatoes and pigs were produced for trade purposes. Most coastal areas were soon growing far more potatoes than kumara 142 and pigs, which were allowed to roam in the bush, were rounded up for sale to European ships. This can be called the pig and potato trade economy and was supplemented by whaling activities. Missionaries arrived in the Bay of Islands in 1814 and started introducing mixed crops and the idea of hamlet farming centred around the church. Wheat, barley, maize, fruit, sheep and later cattle and horses were introduced along with many other items. Hamlet farming was extended to village farming and became the main source of produce for the European settlements. The disastrous effect of the Maori wars and the increasing independence of the European led to a concentration of Maori activity in rural areas, where farming of small subsistence units became the general practice. This pattern was not much altered until the second world war when increasing population and the need for labour in the cities drew increasing numbers of Maori to the urban areas.
The general hypothesis suggested is that there were a number of changes in the basic economy of prehistoric New Zealand, and that these were determined by man's ability to adapt to or to control his ecological situation. If agriculture was introduced by the earliest settlers then its effect was not significant outside of Northland until the technology of storage was developed.- 29
The active extension of kumara agriculture is one of the main dynamics of New Zealand prehistory. 143 Other dynamics are to be found in the ecological change which resulted in forest resources becoming unavailable in much of the South Island and much restricted in the North Island. The succession of plant cover in the South Island was from podocarp to grassland and in the lowland North Island was from podocarp to broadleaf forest. The effect in both cases was very much the same; the avifauna, which was an important element in the total economy, became restricted. Shore fishing or gathering became the basic economic pursuit in most areas. In much of the North Island the extension of basic kumara agriculture reduced the long term effects of this change. The introduction of kumara to coastal and later to inland areas of the North Island and the northern South Island is likely to have advanced fairly quickly. The need for defensive positions in the form of pa suggests an actual movement of people resulting from an expanding population, rather than a mere diffusion of agricultural techniques. 144
A general problem emerging from the ecological change is that of the total effect of man. The principle suggested here is that man may cause vast ecological change if the existing biosphere is predisposed to such changes. In other words, the effect of man's clearing activities in the form of uncontrolled fires and slash and burn agriculture, is not irreversible, unless the tendency to non-reversion is already present. In New Zealand it seems that natural forest succession would eventually have replaced podocarp forest by broadleaf or grassland. Man probably accelerated this process but did not himself cause it.
The recognition of these two main economic dynamics, the change in ecological orientation and the ecological control possible with kumara agriculture, makes possible a much clearer understanding of other facets of New Zealand prehistory. The chronological periods outlined earlier and used in the review of archaeological evidence were based on changes in the proportional distribution of artefacts in type, time and area. These can now be broadly correlated with the economic changes which took place.
Settlement period adzes occur mainly in two areas, Northland and D'Urville-Nelson with a possible smaller group in Coromandel. 145
Early period adzes which appear to have developed under the impetus of the discovery of D'Urville-Nelson argillite are found around all coasts except Fiordland. In both periods the economy was coastal hunting by mobile canoe-borne groups except in Northland where some tuber agriculture was probably being practised. The importance of Northland - 30 in this respect lies in the possibility of survival of some kumara plants in sheltered areas even though experiments in growing methods and storage could have proved disastrous for the main crop. Any extensive cropping of kumara would have required storage even in the far north of Northland
Middle period regional developments were probably caused mainly by increasing population but this in turn meant a pressure on what was an inherently unstable ecology, already displaying a reduced range of avifauna and in the North Island, sea-mammals. During the era of Early Period coastal forest hunting the areas exploited were extensive so that man's effect on them was not as great as it became in the Middle Period when increased local hunting also correlated with the long term culmination of forest succession. Towards the end of the Middle Period the kumara agriculturalists of Northland were probably starting to extend their area southwards to Auckland Isthmus, Coromandel and Taranaki.
It was in the Intermediate period that the climax of ecological change forced a reorientation of economic activity. The podocarp forest of the Eastern South Island was finally destroyed by fire but was probably already too dry to support the moa and many of the forest birds which became totally extinct or very reduced in range. In the North Island and western areas of the South Island much of the lowland podocarp had been replaced by broadleaf which had a similar effect on the birds. In the lower half of the North Island and the South Island the economy was reorientated to sea-shore resources and a consequent shift in artefact concentration occurs. It was probably during this period that increasing population in the nuclear agricultural area forced a series of movements to less favourable areas further south, taking with them proto-classic Maori culture, which is evident from this period but which may have been developing in Northland for some time previously. The movement of agriculturalists with fortifications indicates that (a) there had been a need to develop fortifications in the home area suggesting pressure on resources; (b) the 18th century Classic Maori tribes who ascribed their origin to migratory canoes from Hawaiki during this period were kumara agriculturalists moving out of Northland. That this is so is suggested by the review of traditions given earlier; (c) the proto-classic Maori absorbed the local descendants of Intermediate period hunting groups.
The Late period saw the eventual extension of agriculture and Classic Maori culture to D'Urville-Nelson and Canterbury-Kaikoura areas in the South Island to the possible limit of kumara at Taumutu in South Canterbury.
In the Initial-contact period the introduction of white potatoes by Europeans enabled Classic Maori culture to extend its range to the extreme south of Murihiku and so complete the movement begun in Northland when kumara were first introduced some thousand years earlier.
Thanks are due to Dr H. D. Skinner and Dr M. McLean for discussion and criticism.- 31
1 Fleming 1962:105.
2 Green 1963:25.
3 Groube 1965:93-96.
4 Cook 1784. Ellis 1831. Suggs 1961. Green et al. 1967.
5 Groube 1965:103.
6 Shortland 1842 ms. Journal Coromandel-Kawhia and copy of Puckey-Martin ms. 1854.
7 Baucke 1928:359
8 Wright 1959, and information collated by Rhys Richards.
9 Cranwell and Von Post 1938:303.
10 Shortland 1844 ms.
11 Colenso 1880:28.
12 Baucke 1928:359.
13 Shortland 1844:ms.
14 Shortland 1844:ms.
15 Skinner 1921.
16 Te Rangi Hiroa 1949:47.
17 Groube 1965:Table III.
18 Roberton 1962.
19 Roberton 1961:304.
20 Shortland ms. Hocken Library.
21 This will be examined in greater detail in a future paper.
22 Smith 1896:43, 44, 57.
24 Ibid:59, 90.
25 Groube 1968:144.
26 Groube 1967:10.
27 See Groube 1967:26, 27, divisions based on type, area, time, proportional distribution of artifacts ref. Simmons 1967, 1969 and following.
28 See Simmons 1967 and 1969 for a detailed review of artifact and economic changes in the South Island.
29 See Simmons 1968:116, for a review of this question.
30 Simmons 1969 in press.
31 Shortland 1844 ms.
32 Ambrose 1969:591-2.
33 Simmons 1967:37.
34 Skinner 1959.
35 Stewart 1816:58-92. According to McNab 1909:205, Williams reported 100 acres in 1813.
36 McNab 1909:262.
37 Shortland 1851:81.
38 G. Laws:Personal communication from unpublished ms.
39 E. Murchie:Personal communication.
40 Wellman 1962a:55-73.
42 Ibid:63, 69.
45 Wilkes 1961.
46 Scarlett and Wilkes 1967.
53 Buist 1963:178.
54 Buist 1963:183.
55 Buist 1963:177.
56 Buist 1963:178.
57 Yaldwyn 1958:132.
58 Wardle 1963:313; Elder 1965:44, 46.
59 Parker 1962.
60 McFadgen:personal communication and Davis, 1962.
63 Wellman, 1962b:38-50.
66 Hartree 1960.
67 Green 1963:72ff.
68 Pullar 1961.
69 Green 1959, 1963a, 1963b; Green and Green 1963.
70 Green 1963b:102.
74 Ibid:55-57, 62-72.
75 Parker 1962.
76 Green 1963b:70.
77 Fleming 1963:51.
78 Ward 1965.
79 Cook and Green 1962.
80 Hoskings 1962.
81 Elder 1965:47.
82 Wellman 1962:64ff.
83 Druce 1966:26.
85 Wellman 1962b:64-68.
86 Ambrose 1961.
87 Spring-Rice 1963.
88 Golson and Brothers 1959:6.
89 Golson 1959, Plate II.
90 Golson and Brothers 1959:6.
91 Nichols 1963.
92 Thorne 1875.
93 Wellman 1962b:56ff.
94 Ibid:72. See also Groube 1968, who dates a garden soil in the Bay of Islands to A.D. 800.
95 Robinson 1963.
96 Shawcross and Roe 1966.
97 D. Witter pers. comm.
98 Fleming 1963:51. The presence of juvenile and adult seals in some very early Northland sites such as Houhora could suggest settlement during a colder climatic period.
99 Cockayne 1928:150.
104 Ibid:163, 167.
105 Robbins 1961.
109 Moar 1961.
110 Harris 1963.
112 Elder 1963:45.
113 Cockayne 1928:418.
114 Moar 1958.
115 Cockayne 1928.
116 Wright 1959:23.
118 Cockayne 1928:352 (footnote 2).
121 Simmons 1968.
122 Cockayne 1928:43.
123 See Simmons 1968:119, also Kikkawa 1966:267.
124 Jean Goulding pers. comm. notes that they mostly belong to swampy fringes of a mixed podocarp-broadleaf forest, i.e. around the bush edge of the former Pyramid Valley lake.
125 R. N. H. Bulmer pers. comm.
126 Quoted in Robbins 1961:71.
127 Heusser 1964:107ff.
128 Green 1963:51, 55-59.
129 Groube 1965:25-28.
130 Ambrose 1962.
131 Groube 1965:34.
132 Shawcross 1962.
133 Shawcross 1962, 1964.
134 Golson 1960.
135 Groube 1965:26.
136 Groube 1965:55.
137 Shortland Puckey-Martin ms. 1854 and Coromandel-Kawhia Journal 1842.
138 See Banks 1962, Colenso 1880, Cook and King 1784, Ellis 1784, Hutton and Drummond 1923, Nicholas 1817. Applies to Northland area only; similar cycles were probably in evidence in other areas depending on local environments. Thanks to Dr R. Cooper and B. Stevenson for supplying information.
139 Felton Mathew ms. fieldbook.
140 K. Shawcross 1967.
141 See Hargreaves 1963, for a fuller treatment of this important topic.
142 Shortland, Journal Coromandel-Kawhia, Maketu census.
143 Thanks are due to Les Groube for many stimulating discussions on this and related topics. See also Groube 1968:142-144.
144 See Groube 1968:144.
145 E.g. adzes; archaeological and surface finds which would appear to be early are especially abundant around and north of Kaitaia in the extreme north and in Nelson. In both areas there is a high proportion of high reversed quadrangular adzes with or without grip, most of them being very similar in plan to the “hoofed” adzes of Polynesia though rarely plano-convex in section. Accompanying these in the north are shallower sectioned reversed quadrangular, or in both areas, triangular apex up, half of which have no grip. Front gripped quadrangular are also present but are not dominant.