Rotokura: An Archaeological Site in Tasman Bay

Rotokura, located in Cable Bay on the eastern side of Tasman Bay (see Map 1), is an archaeological site which was occupied periodically at least from the beginning of the fourteenth century. Cable Bay has a reputation of excellent weather and consistently light seas, as is true for Tasman Bay generally. Although the hills are now virtually all cleared of native bush due to the expansion of farmland, it is thought that they were once covered with a variant of the North Island lowland forest, pockets of which are still extant in the area.

A wide range of exploitable resources were available to the prehistoric Maori. The open sea, sheltered bay, tidal flats and streams all provide abundant supplies of fish. Although the complement of species varies from season to season, there would always have been an adequate supply of fish at Cable Bay. The forest would have provided birds, fruits and berries. These resources are also seasonal. No season, however, is marked by a complete absence of resources. The sea also provides sea mammals (seals, sea lions, etc.), the exploitation of which was restricted by the social and breeding habits of each species involved. The Maori also exploited the Polynesian dog for food. In addition to these naturally occurring resources, it is highly probable that the Maori of Cable Bay had access to kumara, since this plant can be grown in the northern part of the South Island. There is, however, no specific evidence available to the author which would associate horticultural activities with Rotokura or Cable Bay. However, more intensive site surveying may locate 'made soils', pits, terraces, stone rows or alignments, or other surface features associated with horticulture in the Cable Bay area. Such evidence is known for D'Urville Island.

Tasman Bay in European History

Two of the most famous explorers of the South Island, Tasman and Cook, did not investigate the eastern coastline of Tasman Bay. This task was, however, superficially undertaken by D'Urville. Before D'Urville's visits sealers had plied the western coast of the South Island, but had spent little time actually on the eastern coast of Tasman Bay. Records left by these illiterate men, vying in a highly competitive industry, are understandably virtually non-existent. D'Urville visited New Zealand in the Coquelle in 1824 as an officer. He returned in command of the same ship, renamed Astrolabe, in 1827. After staying several days on the western side of Tasman Bay, he ran across to Croisilles Bay, missing any close inspection of Cable Bay. However, the descriptions of the Tasman Bay Maori are extremely valuable, since they record the situation before 1829, when the Te Atiawa, Ngati Rarua and Ngati Tama inflicted a crushing page 6defeat on the Tasman Bay Maori. From this time on there seems to have been a steady decline in the Maori population of this area, the victors tending not to settle permanently in their newly acquired territories.

Thus in 1839 when Colonel William Wakefield arrived to buy land for the New Zealand Company, the low density of Maori population in the area persuaded him that it would be suitable for settlement. Whaling also became important in Cook Strait during this period. The Nelson settlement was established in 1841 and European settlement spread rapidly throughout the region.

The period from 1642 through until at least the 1830s in Tasman Bay can legitimately be called the Protohistoric. With careful archaeological research in the Tasman Bay area, we may be able to gain valuable evidence to document this period, about which we can learn so little through the traditional documentary sources of the historian.

Archaeology at Rotokura

When Mr D. Millar was excavating at Tahunanui in 1966, he was informed by Mr R. Tobin of a site at Cable Bay which had yielded artefacts and midden. This site, now known as Rotokura, was given immediate attention by the Archaeological Group of the Nelson Historical Society. There were two important reasons why Millar thought the site should be excavated. The first was that there had already been fossicking at the site and thus what was left of the site should be salvaged before further destruction occurred. The second reason is related to the overall concept of New Zealand prehistory and the artefacts which are known to have come from the site. New Zealand prehistory has been divided into Archaic and Classic Periods by various authors. Although more recently the idea of continuum, rather than of distinct periodization of New Zealand prehistory has gained favour, in academic circles at least, these terms are still used as description markers (for artefacts especially). The Archaic Maori has been termed the 'Moa Hunter'. Although the Maori did hunt the moa, it appears that it was not a dominant factor in the economic strategy of the early Maori in New Zealand. The Classic Maori is the Maori as Tasman, Cook and D'Urville observed the natives of New Zealand before extensive contact with the European.

One of the early problems which occupied New Zealand archaeologists was comprehending the transition from the Archaic to the Classic Maori. In one fossick hole at Rotokura, both Archaic and Classic Maori artefacts were found. It was therefore hoped that this site might enable prehistorians to document the transition period

This site is situated behind a boulder beach, next to a small lake (pond), after which it is named. A reference to the lake in the Nelson Evening Mail of 24/11/1949 reads:

"On the western side of what we call Schroders Mistake Cable Bay is a small lake, the waters of which from some peculiarity in the soil have a dull reddish hue, and for this reason the locality is known to the Maoris as Roto-kura or the Red Lake…"

Mr Jim Eyles has suggested to the author that a red moss (known locally as 'frog porridge'), which appears seasonally, may account for the name. Peart in his book Old Tasman Bay notes that not only was the whole of Cable Bay given the name Rotokura, but that it was also the name of a pa 'that formerly was placed where the boulder bank joins Pepin Is.'. His definition of the name was—'A red glow on the water—viz., a sunset' (p.125). The site discussed in this paper is not that marked on Peart's map (between pp.128 and 129).

Millar has described the site as follows:

"The surrounding topography—lagoon, boulder beach, and hillside—has had the effect of restricting the Maori inhabitants to a semi-flat area measuring approximately 35 yds long by 20 yds wide. This has reduced considerably the scatter of occupational residue and consequently the concentration of artefacts, midden refuse and for extremely difficult excavation in some areas of the structures has been considerable. To date approximately 550 sq. ft. of the site has been excavated and an impressive list of artefacts has been catalogued. The natural stratum consists mainly of beach boulders and this makes site."

(Millar, 1967:10)

Squares six feet by six feet were excavated by Millar, Eyles and six-teen other members of the Archaeological Group, including some school children. All bird, fish, dog, sea mammal and human bone was retained (though only after being put through large sieves). Shell samples were also taken

Before the work done by myself on the faunal material from Rotokura (Butts, n.d.), there had only been one publication relating to Rotokura. This was an article written by Millar and published in the Journal of the Nelson Historical Society 1967, entitled 'Recent Excavations in the Northern Part of the South Island'. In the article Millar outlined his initial impressions of the site. From the initial Archaic layer three genera of moa, fish, shellfish, birds, the Polynesian dog and the Polynesian rat were recovered. Bone and stone fishing lure shanks were well represented, as well as argil lite and greywacke adzes. Millar states that portions of finished adzes suggest it was more than an adze manufacturing centre. Because of the restricted nature of the site, not allowing for any great size in population, he suggests that "the site was occupied only seasonally, and that the main settlement existed somewhere else……" (1967:11). As the Archaic phase progresses, moa and seal bone diminish and subfossil^* moa bone is utilized more. The decline in moa bone is cited as a cause of certain changes in fish hook design page 8from one piece to composite fish hooks. The site did not, however, help to document the intermediate stage between Archaic and Classic.

The Classic phase is assigned to a period of invasions from the North Island. Adzes, generally argillite, are smaller (2B in Duff's classification). Whereas shellfish seemed to be of little importance in the Archaic phase, Millar perceived a greater importance for shellfish in the Classic phase, with a wide range of mudflat, sandy beach and rockshore species present. Birds also appeared to be present in greater numbers. European contact is presented in the top of the occupation by a clay pipe, nails, glass and some rusty pieces of an old iron 'go ashore' pot, of a type commonly used by sailing crews.

Recent Research

During 1977 the faunal material from the Rotokura excavations was shifted from the Nelson Provincial Museum to the Archaeological Laboratory of the Otago University Anthrology Department. The material has been stored at the Nelson Provincial Museum for some years. Only some of it had been sorted and this had been sent to Mr R. Scarlett of the Canterbury Museum. The bird bone identifications made by Scarlett are used here, but the sea mammal identifications have been done by Ian Smith, the dog bone by Jane Teal and Dr P. Houghton has studied the human remains. The fish bone was identified by Paul Wernham, supervised by Dr B. F. Leach, using the comparative collections at the Archaeological Laboratory, Otago University. Identifications of the fish species were made using five of the cranial bones and some other special bones such as the spine of the leather jacket (Navodon convexirostris).

Minimum numbers for each species of fish were calculated on the basis of the identifications. The maximum minimum number for each species is simply the largest minimum number achieved for any anatomical element, left or right. It is possible that using only cranial bones for identification there is a risk of generating biased data as a result of "various cultural practices related to preparation of fish by prehistoric people such as filleting, and also by techniques of preservation to overcome winter shortages" (B. F. Leach n.d.).

Stratigraphy and Periodization

The stratigraphy (a history of deposition) at Rotokura has been the subject of some discussion (Butts, n.d.). Millar's three period interpretation of the site has been followed. Layer 2, the most recent cultural deposition was divided into layers 2A and 2B. Though these may have been consistently applied in the recovery of artefactual material, this was not the case in the recovery of midden material. From artefactual recovery, layers 6, 4 and 2B are Archaic, and 2A is Classic. The prehistoric economics of the Maori are now viewed as a continuum. Millar has suggested that layer 6 (which is page 9represented by little faunal material) may be an internal feature of layer 4. Thus the analysis of the faunal material presented below is concerned largely with comparing and contrasting two periods of occupation represented by layers 2 and 4. There is no attempt to compare Archaic and Classic material as such. A single date of 625±70 BP (AD 1335±70) is available from a charcoal sample recovered by archaeological excavation from layer 4, while layer 2 has both prehistoric and European contact material thus suggesting it was rather late. It should be noted that layers 1, 3 and 5 are sterile (that is, are not the result of cultural activity), probably representing periods when the site was abandoned.

The remainder of this paper is devoted to a comparison of the faunal materials identified from layers 4 and 2, with some comment on layer 6. This will involve discussion of the wetfish species, bird and sea mammal material. The dog and human remains will be briefly discussed at the end of the paper.

Layer Four (Early Occupation) Fish

See Figure 1 for maximum minimum numbers of species identified. The two most frequently occurring species are snapper^* (77% of total number of indivduals) and Pseudolabrus spp. (7.4%). Other species occurring (though individually constituting less than 3% of total) are barracouta, red cod, leather jacket, tarakihi, moki, blue cod, ling, gurnard, sea perch and eagle ray. Two elasmobranch vertebrae were also identified, though these could not be speciated.

Figure 1

From studies undertaken by the author on the seasonal ability of the various species found at Rotokura, it is suggested that ca. 9% of the number of individuals identified are most likely to have been caught in late autumn, winter and early spring. This is very weak evidence for winter exploitation. Summer exploitation is more strongly indicated and the presence of barracouta, tarakihi and blue cod suggest activity at this site may have extended well back into spring or on into autumn.

Birds

See Figure 2 for maximum minimum numbers of species present. By far the most dominant species in this assemblage is the spotted shag^* (45% of total number of individuals). Of the remaining twenty-three species identified in this layer, the Fiordland crested penguin (6.5%) is the most frequently occurring. The spotted shag and Fiordland crested penguin suggest spring through summer exploitation. Evidence for probable winter exploitation is found in the presence of the New Zealand pigeon, tui and king shag. These birds are thought to be in their best condition during winter. However, winter exploitation indicators are minimal. If layer 6 is part of layer 4 (see pp. 8 and 9 above) then the addition of the South Island kaka and Australian gannet strengthen the case for winter exploitation. The latter species is almost certainly exclusive to the months June through September in the Cable Bay area.

Figure 2

Also present: Dinornis torosus (Hutton)
Anomalopteryx didiformis (Owen)
species larger than Anomalopteryx didiformis (Owen)

As is generally the case in the seasonal analysis of New Zealand archaeological assemblages of bird bone, the spring and summer exploitation is the most strongly represented. Thus in the final analysis, mid-spring to summer and early winter exploitation are suggested.

Sea Mammals

In layer 4, Smith (n.d.) has identified fur seals, sea lions, leopard seal and possibly an elephant seal. There are eleven fur seals present. Their age and sex characteristics provide possible indications of both summer and winter occupation. Though the evidence strongly suggests a summer exploitation, there is some reason to suspect hunting of this species may have occurred at more than one time of the year.

Summary of Layer Four

The evidence for summer exploitation of fish, birds and sea mammals is strong, but there is also a body of evidence to suggest at least some activity on the site through into late autumn/early winter. It is a failing of the probability methods used in the study that species most probably caught in winter cannot be definitely identified as such unless their distribution in the area is exclusive to the winter period.

Layer Two (Later Occupation) Fish

See Figure 1 for maximum minimum numbers of species identified. Of the twenty wetfish species identified in Layer 2, snapper (32.1% of total number of individuals) and barracouta (21.6%) are the most common. The former is a spring-summer indicator of exploitation, the latter a late autumn, winter and early spring indicator. Of the total assemblage 45.6% of total number of individuals are thought to indicate spring and summer exploitation, while 41.3% are thought to indicate late autumn, winter and early spring. The remainder can be caught with equal probability throughout the year. It is suggested that the fish bone from layer 2 represents an exploitation of wetfish throughout the year. Autumn and winter indicators are more in evidence with barracouta, red cod, tarakihi and blue cod constituting a large proportion of the total.

Birds

Of the twenty-four species identified in this layer, 50% are maritime birds, 29.2% forest birds and 20.8% are wetland birds. Spring and summer exploitation is indicated by the penguins, shags, prion, petrel, sooty shearwater, juvenile heron, ducks, New Zealand falcon and South Island kaka. Autumn and winter exploitation is indicated by the weka, mollyhawk, New Zealand pigeon and yellow-crowned page 13parakeet. The little shag is available and equally exploitable all year round.

Evidence is present for exploitation at all seasons if it is assumed that species are taken at the time when they are most easily exploited or when they are in their best condition. Best (1942:156) suggests that there was a clearly defined secondary bird catching period when birds which had reached their best condition during autumn were taken. In this category he especially notes pigeon, kaka and tui.

Sea Mammals

Smith (n.d.:3) has identified a sea lion, one other large sea mammal (though he is unable to identify it to species level) and two fur seals. This small number of individuals makes any conclusions very tentative using presently available methods of analysis. Smith (ibid: 15) has suggested winter exploitation of the fur seals and sea lion as the most probable interpretation. With the sea lion, however, there is the possibility that prehistoric distribution may have been different to that at present.

Summary of Layer Two

From the wetfish, birds and sea mammals present, this assemblage has given a much stronger evidence for winter occupation than the layer 4 assemblage. Summer occupation is also strongly evidenced by wetfish species and birds. It is therefore suggested that there is sufficient evidence from the faunal component of layer 2 to document activity at Rotokura during all seasons of the year. This does not mean nor infer that the site was permanently occupied throughout the year. A full excavation report is necessary before such an assessment could be attempted since without this the extent of structures for permanent occupation cannot be assessed. Also a knowledge of sites associated with Rotokura would be a crucial element in any final analysis of the site.

Discussion

The exploitation of available food sources by the prehistoric Maori in Tasman Bay is shown to be wide ranging by the analysis of the faunal material at Rotokura. If we were to add the organic materials which do not survive (e.g. berries) the range is even more extensive. The fish species present are evidence of both shore and canoe fishing, with the possibility of nets and line fishing. The sea also yielded a wide range of shellfish. Though these have not been analysed, initial observations suggest that both rocky and sandy shore species are present in both periods. Trends relating to shellfish species cannot be discussed without further work on the material

The bird material, represented by forty species (if layers 6, 4 and 2 are taken together) also demonstrates a wide ranging exploitation. In addition to these there are three species of moa present. Several other extinct species are identified, including a duck and a page 14swan. While the birds were probably most abundant in the summer months, research has shown that some species would have provided excellent food in the winter. In the later period the maritime species appear to be the mainstay, with forest and wetland birds forming a supportive role. However, this may be a biased view, with some species being taken, cooked or eaten at other more specialized sites or in other areas of the site than those excavated.

Smith (n.d.:10) suggests that the butchering pattern of the sea lions in the early periods indicates the presence of a colony close to the site. He also notes (ibid.:11) that the fur seals were not butchered on the site and that it is likely to have been some distance from Rotokura. A possibility is the rocky north coast of Pepin Island.

The Polynesian dog also provided a food source at Rotokura. Layer 4 has a maximum minimum number of eight, while layer 2 has sixteen. The dogs range in age from five to six weeks to older than eighteen months. Ages are based on tooth eruption and bone growth. The dog bone at Rotokura is thought to result from the exploitation of the species for food and artefact manufacture. It is suggested that dogs at Rotokura could have provided a valuable food supply if seasonal food supplies were interrupted by unusual climatic occurences, or in periods of transition from one season to another.

The human bone from Rotokura was analysed by Dr P. Houghton, Department of Anatomy, Medical School, University of Otago. Two adult persons are thought to be represented: one, a robust male about thirty years of age; the other, a female who appears to fall within the same parameters—Polynesian, aged about thirty years. The cranial material has been deliberately cut, presumably for the purpose of making artefacts. Millar (pers. comm., 1977) has suggested that the scattered appearance of the material during the excavation suggested the result of violent death. Dr Houghton was not able to confirm that the deaths of the individuals were the result of violence.

Conclusion

The Rotokura midden represents an exloitation pattern which involved wide spectrum hunting and fishing in the Cable Bay area. A trend towards a year round activity at the site has been documented (Butts, n.d.). The early period lacks conclusive evidence for winter activity, while the later period has more substantial evidence that winter activity was undertaken. Fishing, fowling, sealing and the consumption of Polynesian dog all formed essential (if proportionally changing) elements in the subsistence economics of the Rotokura inhabitants. The variety of species present in the midden at Rotokura suggests a far more varied diet than New Zealanders eat today. It is also worth noting that as a consequence of forest destruction, seal over-kill, and the suspected damaging effect of commercial fishing on fish supplies, it would be impossible to live the same life style within the same area at the present time.

Appendix

page 16

Scientific Names of Bird Species

South Island brown kiwi	Apteryx a. australis
Southern blue penquin	Eudyptula m. minor
Blue penguin	Eudyptula minor subsp.
Fiordland crested penguin	Eudyptes p. pachyrhynchus
Erect crested penguin	Eudyptes pachyrhynchus sclateri
Crested penguin	Eudyptes pachyrhynchus sp.
Mollymawk	Diomedea cauta subspp.
Prion	Pachyptila sp.
Australian gannet	Sula bassana serrator
Sooty shearwater	Puffinus griseus
Fluttering shearwater	Puffinus g. gavia
Huttons shearwater	Puffinus huttoni
Black shag	Phalacrocorax carbo novaehollandiae
Pied shag	Phalacrocorax v. varius
King shag	Leucocarbo c. carunculatus
Little shag	Phalacrocorax melanoleucos brevirostris
Spotted shag	Stictocarbo p. punctatus
Duck	Anas sp.
Swan	Anser sp.
Extinct swan	Cnemiornis calcitrans
New Zealand falcon	Falco novaeseelandiae
Western weka	Gallirallus a. australis
Weka	Gallirallus australis sp.
Rail	Gallirallus minor
Takahe	Notornis mantelli
Southern black backed gull	Larus dominicanus
New Zealand pigeon	Hemiphaga n. novaeseelandiae
South Island kaka	Nestor m. meridionalis
Yellow crowned parakeet	Cyanoramphus a. auriceps
Tui	Prosthemadora nu novaeseelandiae
South Island kokako	Callaeas c. cinerea
Extinct goose	Cygnus sumnerensis
Extinct duck	Euryanas finschi
	References

Best, E.	1942:	Forest Lore of the Maori.
		Dominion Museum Bulletin 14.
Butts, D. J.	n.d.:	Seasonality at Rotokura, Tasman Bay: A Case Study in the Use of Faunal Identifications to Establish the Season of Occupation for an Archaeological Site.
		Unpublished B.A.(Hons.) Dissertation, Anthropology Department, University of Otago.
Leach, B. F.	n.d.:	Fish and Crayfish from the Washpool Midden Site, New Zealand: Their use in determining season of occupation and Prehistoric fishing methods.
		Unpublished paper. Anthropology Department, University of Otago.
Millar, D. G. L.	1967:	Recent Archaeological Excavations in the Northern Part of the South Island.page 17
		Journal of the Nelson Historical Society 11(2):5–12.
Millar, D. G. L.	1977:	Pers comm.
		27 Holyrood Street, Taradale, Napier.
Peart, J. D.	1937:	Old Tasman Bay.
		R. Lucas and Son. Ltd., Nelson, New Zealand.
Smith, I. W. G.	n.d.:	Sea Mammals from Rotokura (S14/1).
		Unpublished paper, Anthropology Department, University of Otago.