The Rat's Tale
A historical "document" isn't always a writing sample
Enter the Rat
But with this remarkably well-documented record, why is it so difficult to date human arrival in New Zealand? Thereby hangs the tail (forgive the pun) of the kiore or Polynesian rat, Rattus exulans. Prior to human arrival, bats were the only indigenous mammals on New Zealand or the other islands of Remote Oceania. The kiore arrived in Polynesia with humans as a notorious and highly successful stowaway, like the Norwegian rat more familiar to Europeans and Americans.
In 1994, accelerator mass spectrometry (AMS) radiocarbon dating began at the Rafter Laboratory in New Zealand, making accurate dating of small bones possible. Richard Holdaway of Canterbury University and Atholl Anderson of Australian National University got the brilliant idea of using specimens of Polynesian rats as documents of human presence. Even if the founding colonies of humans were very small and slow to reproduce, the rats that came with humans were likely to have reproduced rapidly and spread quickly. Finding and dating rat bones should be much easier than finding human bones or other human traces. Their idea was to test the reliability of rat bone dates using archaeological specimens from sites of known radiocarbon age and then to extend the method to specimens from natural sites of unknown age.
Holdaway used rat bones, excavated from New Zealand sites that had been formed by laughing owls. These birds swallow their prey whole and, after digestion, sit on favorite roosts and cough up neat little packets of bones known as owl pellets. Based on gelatin extracted from these owl pellet bones, the rats were almost 2,200 years old, as Holdaway reported in 1996 in Nature. However, this meant that rats were present on New Zealand for at least 1,000 years before the earliest archaeological record of humans. Holdaway explained the discrepancy by hypothesizing that Maori had paid an early, transient visit to New Zealand—leaving rats—but did not settle there in numbers until about A.D. 1300.
The early rat dates implied that faunal and floral extinctions in New Zealand occurred in two stages. The first stage saw perhaps a drop in species abundance, but no extinctions: that was the "rats only" phase during which the rodents bred freely and ate plants, fruits, seeds, insects, lizards, snails, eggs and nestlings of ground-breeding birds. The second stage, about 1,000 years later, saw the founding of sizeable settlements of humans who hunted, fished and trapped with enthusiasm, triggering numerous extinctions. This interpretation is known as the traditional or "long chronology" of New Zealand.
Anderson addressed the dating issue by using material from an excavation he had led between 1988 and 1992 of the huge, stratified, moa-hunting middens at a site called Shag River Mouth. The remains of at least 6,000 moas show how skillful and systematic the moa hunters were. Anderson found a pattern of deliberate butchery, with moa heads and necks discarded in one pile, pelvises in another, complete ribcages in a third. The major bone of the thigh, the femur, carried big muscles and was usually intact but the big bone of the lower leg, the tibiotarsus, was smashed, for marrow or to obtain thick bone pieces for manufacturing tools. Sometimes a heap of bones was formed of only one species of moa, suggesting that Maori may have taken down an entire flock. Moas were roasted across small depressions or ovens, many of which were preserved intact.
In addition to moa remains and ovens, there were at least 43 hearths, 36 huts, many adzes and hafted blades, fish hooks, manufacturing tools, various ornaments and two intact moa eggs, one near a human burial. At the time of excavation, Anderson took nearly 50 samples for radiocarbon dating—samples of charcoal from hearths and moa ovens, bird bone, marine shell and moa eggshell—and the dates grouped tightly around A.D. 1350 to 1400.
In the face of this hard evidence that Maori had a very well-developed and effective moa-hunting adaptation between A.D. 1300 and 1400, Anderson was troubled by the early dates Holdaway had obtained for Polynesian rats.
"It just did not make sense to me," Anderson explains, "that people could have been in New Zealand for 1,000 years before they started killing moas and seals, making shell middens, flaking stone and so on. What were they doing instead? Holdaway's default position, that people had arrived and then left almost immediately, also looked improbable to me."
"To resolve the problem, I took six samples of rat bone from the same provenances as the other 50 dating samples from Shag River Mouth. The Rafter laboratory in New Zealand (which had dated Holdaway's rat bones) dated four of them between 180 B.C. and A.D. 700, but two samples that we sent to the lab in Oxford were dated to A.D. 1000–1200. Clearly something was wrong," he says. Anderson's paper suggesting that there were technical problems with dating rat bones was published in Archaeology in Oceania in 1996, at the same time as Holdaway's Nature paper reporting the early dates on rat bones.
This led Anderson and Wilmshurst to argue for the "short" chronology, which favors a more-recent peopling of eastern Polynesia that began about A.D. 900 and reached New Zealand around A.D. 1300.
As the debate proceeded, Anderson noticed that Holdaway's dates clustered according to the time at which the Rafter Laboratory had processed them. Thirty of the first 32 dates calculated by that laboratory fell between 200 B.C. and A.D. 800. Of the next 28 dates run by the same lab, 25 were from A.D. 900 or more recent. Anderson suggested that perhaps the Rafter laboratory had failed to remove contaminating, older carbon from the rat samples in pre-treatment or the rats had incorporated older carbon from their diet into their bones.
Like Anderson, Wilmshurst had found discrepancies between Holdaway's rat dates and the evidence from the sedimentary profiles she examined in the 1990s and early 2000s. In core after core—she has studied more than 150 now—she found a major change in the flora and in the frequency of charcoal-bearing levels dating to about A.D. 1300. The signature of human arrival seemed clear and matched the archaeological evidence.
In the face of skepticism, Holdaway continued to test rat bones, seeking further refinement of the long chronology hypothesis. None of the bones in his second round of testing were as old as those in the first.
In recent years, the short chronology has been strengthened by additional new dates indicating that all of East Polynesia except New Zealand—Hawaii, the Marquesas, Easter Island, the Society Islands—was colonized about A.D. 900, much later than the rat dates in New Zealand. These findings raised a new paradox. If New Zealand was colonized at 200 B.C. and the rest of East Polynesia was not settled until A.D. 900, then the Maori colonizers of New Zealand could not have brought East Polynesian culture to New Zealand when they settled there. However, Maori culture is indisputably part of East Polynesian culture.
Coming from separate perspectives, Wilmshurst and Anderson found themselves agreeing that if the date of Wairau Bar was an accurate document of a very early colonization of New Zealand by humans, then the early rat dates could not be correct. A new approach to testing the dating was needed.