Australasian Science: Australia's authority on science since 1938

A Solid Grip on the Moa Extinction


A bird bone to reckon with: Dr Morten Allentoft contemplates the shin bone of a giant female moa.

By Richard Holdaway

Were humans responsible for the extinction of New Zealand’s moa, or were they already in decline?

Nine species of moa became extinct in New Zealand within a century or so of Polynesians settling there in the 14th century. Since Sir Richard Owen recognised the existence of these large, wingless, ratite birds in the mid-19th century, moa have been as much part of the mythology of extinction as the dodo itself, with ongoing controversy over the role humans might have had in their demise.

Our research has closed the door to environmental factors by showing that, despite changing climate, moa populations were large and stable in the millennia before the arrival of Polynesians.

The fates of nine species of moa might not seem important when humans have now placed a large chunk of global biodiversity under threat. But knowing how and why moa met their end might be our best chance to understand what happens when people possessed of even the most basic technologies reach a new land, or, what amounts to the same thing, expand into previously unoccupied territory to accommodate their growing populations. If people with only fire and stone tools were around when that many species of large vertebrate died out in two or three generations, then we need to know whether they were on the way out or whether people eliminated thriving populations. If it was the second scenario, then we should be concerned.

The effects of our ancestors on pristine environments are blurred by geologic time, as dating becomes progressively less accurate and samples become rarer, more widely separated in time, and without a clear association with the extinction event itself. In Australia, for example, uncertainties in dating the arrival and spread of native Australians and the extinctions through time and space of the giant marsupials, mihirung birds and giant goannas, can be measured in millennia. In New Zealand, however, the megafauna met its end nearly 50,000 years later, and these uncertainties in dating are reduced to decades. The sharp focus provided by the recent, well-preserved, and precisely dated natural and archaeological records allows us to follow events in almost real time. It then becomes crucial to understand the state of the megafauna populations in the millennia before and up to Polynesian settlement.

The events between 600 and 700 years ago in New Zealand provide perhaps our best window into what happens when people and ecosystems collide, and – more specifically for the bigger animals that are at risk today, like elephants, rhinos and tigers – the processes behind megafaunal extinction.

We know with some certainty that people settled in New Zealand about 650 years ago, and that moa were extinct within the following 150 years. But a key point of contention has been how the moa populations were doing in the centuries and millennia before Polynesian settlement. If they were in free-fall decline from natural causes, as some researchers have suggested, then people would have, at the most, delivered the coup de grace. If they were doing just fine then the focus moves firmly to what the new settlers were doing.

The size of populations of extinct species is usually inferred from how many fossils there are in each layer of the geological stratigraphy. However, population sizes are usually only hinted at by the number of animals preserved in the successive layers.

Estimates are therefore hampered by all the uncertainties and vagaries of deposition, discovery and recovery. Animals may be present but just not incorporated in the fossil record (false absence), or their bones can be exposed and redeposited (false presence), so their abundance in the death assemblage is often only the roughest indication of how common they were in the environment at that time.

Absence of evidence is not evidence of absence, and mere presence tells us little about species abundance. Most notably there is still controversy as to whether the dinosaurs were on the way out in the few millions of years before their extinction and that the Chicxulub asteroid just finished them off, or even that the asteroid struck after the main extinction process had taken place over millions of years.

Thus New Zealand’s moa allow us to study the effect of a “human impact” in immeasurably finer chronological detail.

Taking advantage of the large number of splendidly preserved moa in five sites in a small part of North Canterbury, in the northeastern South Island of New Zealand, we set out to genetically identify, sex and radiocarbon date as many moa from the four major local species as possible. The vast majority of the bones sampled had been in museum collections for many decades.

Morten Allentoft, at the time a University of Canterbury PhD student under the supervision of Mike Bunce (now of Curtin University) and Marie Hale (University of Canterbury), isolated and characterised six highly variable nuclear DNA sequences from 217 carbon-dated individuals representing the four major species found in North Canterbury. The 217 moa that were genetically characterised and dated constitute one of the largest such data sets on megafaunal assemblages anywhere.

Morten then modelled the size of the local moa populations over thousands of years, and integrated these results with the archaeological record in collaboration with Chris Jacomb of the University of Otago, Dunedin.

Time-stamped genetic information such as this has never been available before in such abundance, so new methods had to be developed and existing methods adapted before the data could be fully analysed. Morten worked on this with international researchers, and concluded that all four species – including the South Island giant moa Dinornis robustus, whose population had previously been estimated genetically to have been in the millions – were all as abundant immediately before Polynesian settlement as they had ever been.

Dinornis robustus was analysed in more detail because it had yielded the most data and because previous work had suggested that its population had crashed before Polynesian settlement. This and other population simulations failed to show any decline in moa numbers prior to Polynesian settlement.

So, rather than collapsing, it seems that the moa populations were doing just fine. If anything they were still recovering their numbers after the end of the most recent glaciation, known in New Zealand as the Otiran Glacial, which peaked about 20,000 years ago and ended about 10,000 years ago.

Our study indicates that when Polynesians arrived they confronted large, healthy moa populations, occupying dynamic ecosystems from the rainforests of the North Island and the north and west of the South Island to the dry and highly productive eastern forests and the southern alpine and coastal shrublands. One-hundred-and-fifty years later they had all gone, along with most of the best habitat and dozens of other species.

At a time of stable climatic conditions, only one conclusion can be drawn. The New Zealand megafaunal extinction was caused by humans burning vast areas of vegetation and assiduously hunting moa wherever and whenever they could. The moa would have been as naïve to the dangers of humans as Australia’s now-extinct mega-marsupials. Indeed, archaeological sites from the first decades of Polynesian occupation contain about 150,000 moa. The extinction occurred so soon after Polynesian settlement that the total human population involved was likely to have been very small indeed, certainly less than 5000.

Resolving events in New Zealand 600 years ago to an accuracy of decades provides a window into the extinction of the mammoths and other megafauna in the Americas 10,000 to 15,000 years ago and Australia’s megafauna nearly 50,000 years ago. In those wider perspectives, understanding these past extinctions gives us a better handle on what is happening now.

Richard Holdaway is a fixed-term Professor at the University of Canterbury’s Department of Geological Sciences.