New Zealand's tiniest native birds could be the missing piece of a puzzle facing scientists around the world.
A new study aims to answer whether New Zealand wrens - the petite group of birds now only represented by our smallest bird, the rifleman, and the mountain-dwelling New Zealand rock wren - hold a special trait that allows them to learn new sounds.
But the implications of the research, supported with a $300,000 Marsden Fund grant, could reach beyond our shores.
Although most animals communicate with innate sounds, a select group - among them whales, dolphins, bats, elephants, some birds, and us - can develop their own.
What's called "vocal learning" is often associated with songbirds and parrots, whose famed vocal abilities and learning process are so similar to ours, they've become important research subjects for understanding of how speech evolved in humans.
Much of this research is based on assumptions about when, where and why vocal learning evolved - but recent research has drastically reordered the bird family tree.
"For decades, we thought that parrots and songbirds were only distantly related, so we assumed that each group had evolved vocal learning independently," said Dr Kristal Cain, an integrative biologist at the University of Auckland.
"However, new research has shown that they are actually very close to each other on the bird family tree, and scientists now think their distant ancestor may have evolved vocal learning.
"If so, vocal learning is much, much older than previously thought, and probably evolved for very different reasons than we previously thought."
One special group of birds are on the branch between parrots and songbirds - New Zealand wrens - but no one yet knew if they were vocal learners.
"If they are, this would really change how we view vocal learning, and alter the direction of future research."
The New Zealand wrens, once six known species before humans arrived here, are considered unusual and ancient, having evolved before all of the songbirds found in our wilderness today.
"They are very important for understanding the evolution of vocal learning and how the brain evolved to allow vocal learning, but we know almost nothing about their vocal patterns," Cain said.
"Because they make very simple noises, it has long been assumed that they are not learners, but recent research suggests it is possible they actually learn those simple vocalisations."
Her team will focus on rifleman birds living in the wild, tracking them from birth to adulthood and recording their sounds.
They'll then compare those sounds with those of adults around them, and to other populations, to find whether they'd been learned or not.
The study came with some obvious challenges: the birds were tiny, moved quickly, and tended to group together, which made it hard to pick individual sounds.
"Further, their vocalisations are quite simple, so detecting differences among family groups may be difficult and require lots of recordings," Cain said.
"We're countering these challenges by working with folks that have done research on this species in the past."
They'll also draw upon new technology allowing automated recordings at the nest to be paired to video, so large amounts of data can be gathered without disturbing the birds.
"This will solve one problem, but it also creates a new one: how to efficiently deal with stacks of recordings.
"To deal with that challenge we're using special software that will allow us to train the computer to find the noises we're interested in.
"The best part of this project, is that whatever the answer - whether they are or not learners - the results will be very exciting and important for future research."
New Zealand was prized by biologists as a natural laboratory for studying the evolutionary history of birds, as it was once part of the supercontinent Gondwana yet still boasted very old species of songbirds and parrots.
"Songbirds are the most widespread and diverse group of birds, and parrots are widely regarded as some of the most intelligent and vocally flexible.
"Consequently, New Zealand could become one of the world leaders driving our understanding of how and why these bird groups evolved the way they did."
Further, understanding whether the New Zealand wrens were vocal learners, and what role those vocalisations played in the social behaviour of this species, would also be important for conservation, she said.
"For example, who we translocate, and at what age, will depend on whether they learn and whether vocalisations are used in finding mates or coordinating parental behaviours.
"Finally, to understand when, why and how vocal learning evolved in birds we must have a better idea of vocalisations in the New Zealand wrens.
"They really are a missing piece to this puzzle."