Some of the coprolite samples reached 15cm in length
An analysis of the dried faeces of the giant moa, an extinct bird from New Zealand, has overturned ideas about what the flightless giants ate.
At three metres in height, it was thought moa grazed on trees and bushes, but the faeces turned up evidence only of tiny herbs.
The study also showed that moa diets were significantly different to those of the species introduced later.
The research appears in the current issue of Quaternary Science Reviews.
Alan Cooper, director of the Australian Centre for Ancient DNA and senior author on the study, has been collating DNA from a range of different species of moa in recent years.
Prior efforts investigating ancient DNA in the region have shown how varied species of birds diverged with the breakup of the islands of the South Pacific, providing the researchers with a catalogue of DNA samples to which they can refer.
In the new research, Jamie Wood of the University of Otago and colleagues studied the leaf fragments, plant seeds, and DNA in more than 1,500 faecal fossils - known as coprolites - that have survived only because of the moa's behaviour.
"When animals shelter in caves and rock shelters, they leave faeces which can survive for thousands of years if dried out," Professor Cooper says.
By analysing the DNA in the coprolites, the researchers were able to discern the diets of the varying species of moa that were scattered around New Zealand.
"Over half the plants we detected in the faeces were under 30cm in height," says Dr Wood. "This suggests that some moa grazed on tiny herbs, in contrast to the current view of them as mainly shrub and tree browsers.
"We also found many plant species that are currently threatened or rare, suggesting that the extinction of the moa has impacted their ability to reproduce or disperse."
British zoologist Richard Owen first described the moa
The moa's evident diet is in sharp contrast to that of the herbivores that were introduced later to the islands, for whom the moa's diet would have been toxic - suggesting that the moa co-evolved with the native plants they ate.
"It's a fairly elegant approach that most people wouldn't have thought of doing," says Ian Barnes, a bio-archaeologist at Royal Holloway, University of London.
"The ability to look at a species that went extinct several hundred years ago and get some insight into the real ecology of that animal - what it ate, whether they were in competition with each other, whether different species ate different stuff, how they divided the landscape up - is quite remarkable."
What is more, the approach could be applied in Australia to conduct similar studies - but the evidence has so far been lacking.
"Given the arid conditions, Australia should probably have similar deposits from the extinct giant marsupials," says Professor Cooper.
"A key question for us is 'where has all the Australian poo gone?'"