Clues left by the first Polynesian arrivals to Northland have led scientists to uncover one of New Zealand's earliest human settlements.
He began hearing the stories about Mangahawea almost as soon as he took the job.
It was 2006, and Andrew Blanshard was a newly-appointed heritage ranger for the Department of Conservation in the Bay of Islands. One of his first tasks was to put a track in at Moturua Island, a scenic reserve in the Rawhiti Inlet.
While working there, several local hapu members mentioned an archaeological site, at Mangahawea Bay on the island's western side. Not a pā or a terrace, but something else, older.
"That will be a good one to read up about," Blanshard thought.
But when he got back to the office at Kerikeri he found nothing written down. Unable to shake the feeling that he was missing something, Blanshard spent some time combing through the department's archives.
There, among a bunch of files he found a single, scruffy field notebook from an excavation in 1981, carried out by archaeologists from Auckland. Among their findings had been moa bone, the remains of a now-extinct limpet, and a rare shell pendant.
Properly curious now, Blanshard called James Robinson, an archaeologist from Heritage New Zealand.
"What do you know?" he asked. Apparently, Robinson said, there were more boxes in Whangārei. So Blanshard collected those. And he kept asking friends and colleagues, eventually finding more boxes in Wellington, in Auckland, in Dunedin, in the vaults of government departments and universities, dormant, gathering dust.
It took nine years to get the full collection, which still wasn't much. Though the context and age of the finds was unclear, it was enough for a start.
"Even at that preliminary level you could tell that there was some real importance to this," Blanshard says.
"In that time I had also got the sense that Mangahawea was almost like an archaeology fable within New Zealand. It was this site that everyone knew about and everyone knew was important but we hadn't quite teased out why."
Together with Robinson, Blanshard approached the local hapu of Ngāti Kuta and Patukeha, then joined with scientists from Otago University to apply for funding to re-visit the site.
Unfortunately the leader of the first excavation had died in 2016, however, several other archaeologists from the 1981 project were available to help.
In early 2017 a team returned to Moturua Island, under the tikanga of the hapū. There they confirmed the work of the earlier excavation, and found some new artefacts, all which point to a significant finding - that Mangahawea Bay was one of the first places where the ancestors of modern Māori settled after their long journey from Polynesia, at the very beginning of Aotearoa's human history, 800 years ago.
An early site - but how early?
To get to Moturua in the modern age requires a much shorter trip. We drive 250km from Auckland to Paihia, and then take a half-hour boat trip to the island itself.
Blanshard stops at the wharf at Russell to pick up Ngati Kuta kaumātua Matu Clendon and fellow Ngāti Kuta member Robert Willoughby. The Department of Conservation's small alloy runabout vies for a space with the large tourist launches.
Rounding the headland to Rāwhiti Inlet, we see a small pod of dolphins, surrounded by boats. Luxury holiday homes perch on some of the islands, including at the southern end of Moturua, where a handful of sections were sold into private hands some years ago.
The island has four sheltered bays, one roughly at each compass point. White sand shines below regenerating bush, with rocky outcrops marking the end of each beach. Some of Moturua's history is already well known.
Captain Cook anchored offshore in 1769, and traded with Māori for food. Three years later, the French explorer Marion Du Fresne set up a hospital and a forge on the island. It ended badly, with an iwi attack on the French, and then the retaliatory sacking of a local pa, including the deaths of 250 Māori. In World War II, the navy operated a mine control station in Army Bay, with some remains still visible.
Mangahawea is on the western side, the bay you'd see first if you were coming from the harbour entrance. It is broad and sheltered, a creek running across the sand at each end.
Back from the beach is a flat, grassy meadow, and this is where the scientists have focused their attention. The 1981 team excavated three sites, all which were found and re-examined by the scientists over two weeks during January, 2017.
The work isn't glamorous. This isn't Indiana Jones. It is hot, and dusty, and the digging is a hard slog. Anything of interest must be meticulously bagged and labelled. Though some of the material was sieved on site, a large sample of midden was also shipped down to Dunedin, to be examined for much smaller artefacts.
It is painstaking and intricate, with each tiny detail - such as identifying species of fish - taking hours, as each jawbone is washed, sorted, compared and named.
For their troubles, the archaeologists at Mangahawea have been rewarded with a beautiful stratified site. The periods of human activity are layered like a cake in the soil, the oldest at the bottom. This archaic layer forms what Robinson calls a "cultural horizon", spreading across the field, and containing evidence about the lives of the people who lived there during that time.
The scientists have matched the old finds to the new site, and uncovered more artefacts as they went, including the teeth of kurī and bits of obsidian. They also discovered a huge hāngī at the crumbling edge of the stream bank, measuring 5 by 3 metres, full of moa and seal bones, lined with rocks.
During our visit, a year later, yet more of the hāngī pit has been eaten away by erosion, its blackened centre exposed.
Robinson, the lead researcher, crouches next to it and begins gesturing to the soil. "Here you can see the base of the hāngī, and that's the charcoal, and the stones that lined it."
He climbs up the bank and keeps talking. "This was probably the cooking area of the village, or one of them. Here is where the waka would have come up, so you'd expect to find shellfish there. And there, where we found kurī coprolites, is where they might have kept the dogs."
Listening to him, the site before us is transformed. You can imagine the thatched huts of the village, the smoking hāngī, the kūmara pits. The waka lined along the shore. A small, scattered group of Polynesian colonists, making Aotearoa their new home.
In their bid to prove the Mangahawea site was inhabited by that first generation of settlers, the archaeologists are relying on four strands of evidence.
The first strand is faunal - the remains of "naive" native species which only survived shortly after human arrival, meaning anyone eating them must have been among the first people here.
Moa, for example, are thought to have lasted only 200 years, due to predation. The large limpet Cellana Denticulata, also found at the site, probably died out even sooner.
Archaeologists also look at the style of artefacts - the fish hooks carved from a single piece of bone, for example - or worked pieces of pāua shell that also appear very Polynesian in style.
"From those things alone we know this is a very early site, but how early?" Robinson says.
To get a more specific answer, they've used radiocarbon dating, where scientists measure the amount of a radioactive form of carbon (carbon-14) retained in a once-living object. Because carbon-14 decays at a known rate, scientists can then determine its age.
This strand of the work was overseen by Professor Richard Walter, from the Department of Anthropology and Archaeology at the University of Otago. So far, Walter says, the testing suggests people were living on the island in the 1300s.
"That's about as early as you get in New Zealand," Walter says.
"There's a whole lot of sites about that age around the country, which shows how widespread the early settlement phase was."
Though the site is therefore interesting to archaeologists, particularly given the rarity of such finds in the north, it is unlikely to be significant to the general public.
Unless, of course, they can prove the origin of one specific artefact - what is believed to be a pearl shell pendant, probably brought all the way from across the Pacific.
The last place settled on Earth
Though it was once a subject of contention, it is now generally agreed Polynesia was first colonised around 1500 BC, by the Lapita people.
From Southeast Asia, they moved through Western Polynesia - to Fiji, Tonga and Samoa, spreading their language and distinctive pottery, building villages and homes.
After another 1000 years, they made it to Eastern Polynesia - Tahiti, the Marquesas, the Society Islands. In 600 AD, they settled Easter Island, or Rapanui. One hundred years later, they arrived in Hawaii.
Each time, they would take what scientists call the Polynesian "toolbox" - a starter kit for life in a new place, with domesticated animals like pigs, rats and dogs; plants like coconut, taro and kūmara; and their knowledge of the sea and wind and stars.
New Zealand - by then the last place to be settled on Earth - was their final push.
The Polynesian colonists reached the shores of Aotearoa around 1200 to 1300 AD, after an epic journey.
According to the seminal archaeologist Janet Davidson, the settlers would have looked like people from Hawaii, the Cook Islands, or Tahiti. They spoke a language ancestral to modern Māori, and brought or made artefacts similar to those found in the Society Islands and the Marquesas.
However, because the cultures on the islands were so similar at the time, it was almost impossible to tell exactly which place they came from. For Davidson, it was one of just many questions that would likely always go unanswered.
"We do not know when the first settlers arrived; we do not know where they landed; we do not know where they came from. Worse, we shall probably never know exactly," she wrote in a 1983 article for The Journal of the Polynesian Society.
Part of the struggle to place the Polynesians exactly is the lack of cultural artefacts tying them back to the Pacific. So far, amid all the research done in New Zealand, there are only two examples - a chisel found at the famous archaeological site at the Wairau Bar in Marlborough; and a fishing lure, discovered at an archaeological site in Tairua, in the Coromandel. Both are made from shells found only in the tropics.
During the 1981 excavation at Mangahawea, the archaeologists also found an item they thought might be made from pearl shell. An oval pendant, with two little tangs in each end, it was found in the same level as some moa bones, so it is assumed to be from the early period too.
"The design is very Polynesian in its artistry. But we don't know if it's a New Zealand originating shell or a Polynesian originating shell," Robinson says.
"If the shell pendant is made from a tropical species then the pendant actually came on one of those waka from the Pacific and it therefore becomes an heirloom of infinite significance."
The pendant has already been to Te Papa, where it was unable to be identified. The scientists are now hoping they will be able to use a more comprehensive shell reference system from Australia to identify it.
If it matches a Pacific shell, it will be only the third artefact of its type, and would potentially elevate the Mangahawea site to one of national interest.
However, knowing it's from Polynesia still won't be able to tell the scientists whether the shell is Tahitian or from the Marquesas, or the Cooks, given there is a similar lack of artefacts in those islands. Equally, it's possible the ancestors came from a variety of locations, on multiple voyages, over a number of years.
'Names just don't fall out of the sky'
During the 2017 excavation, Ngāti Kuta kaumātua Matu Clendon stayed on Moturua Island every night. He was born and raised there, and is now responsible for keeping visitors to the site safe.
"It's a powerful place," Clendon says.
"That's why we do the karakia, at the beginning, and at the end. And that's why I stayed, to keep the mauri in place."
Clendon's family used to own the island, and farmed there before the war. However, when he and his brother planned to buy their elders out, the council raised the valuation, dashing their hopes.
He knows all the stories of Mangahawea, and of Ngāti Kuta, and is entrusted with keeping them safe. Despite the bitter history, the hapū's story - the cultural science - is the fourth strand the scientists are using in their work, enabled by a partnership of mutual respect and a joint desire for knowledge.
The story of the arrival, as it was told to Clendon, goes that after Kupe discovered Aotearoa, the ancestors planned the voyage of the seven founding waka.
The colonists headed for Rakaumangamanga, or Cape Brett, which was one of the waypoints in the Polynesian triangle - the others being Rapanui, or Easter Island, and Taputapuatea, in Tahiti.
"They would have gone straight for it, following the winds, the stars, the tide. And when they got there, the waka all split off," Clendon says. Hence, the translation of Rakaumangamanga - "to branch".
"From there, you can see their footprints. Our ancestors, when they go to a new place, they leave a footprint via a name. Names don't just fall out of the sky."
Most of the names in the Bay of Islands seem to link back to Tahiti. In the bay, there's a "Poroporo" island - or Bora Bora. Mangahawea itself is an old name.
The most significant is the name given to the tiny island off the north of Mangahawea Bay - Rangiātea. Clendon says it's a transliteration of Tahitian island "Ra'iatea", the home of the sacred site Taputapuatea and the cultural centre of the Polynesian world.
"That's why there is such a strong spiritual presence here," Clendon says. "It's calling all the way back to Ra'iatea."
To Clendon, and Willoughby, the reason the new arrivals would have chosen Mangahawea for their new home is obvious.
"If you look at Moturua as you come through the Bay of Islands passage, it sticks out. In a line of sight, the waka would have circled and found this place and thought it was a good place to settle," Willoughby says.
"Also, it was an island. And the thing about the island is that they can control it. They wouldn't have known what was on the mainland, they would have taken their time to venture into the hinterland there."
The bay is sheltered, with water, and close access to the sea. Willoughby says the people would have built nīkau huts, fashioned on what they knew. They would have planted kūmara, and fished.
"The one thing with the old people is that they understood the seas. They were fisherpeople basically, they would have figured out where to go and what to do."
The hapū don't need the science to tell them about their past. But they are excited that the island could become an important part of New Zealand history, and that more people could learn about the nation's origins.
"If what we are saying can be proven that's a good thing," Clendon says. "If it means more information, and it will uphold the mana of Māori it's a good thing."
"Otherwise we're just digging holes"
The next step for researchers is to find out more about what the village looked like. It was likely small. The closest neighbours would have been a long way away.
The scientists presume, that like those found in the South Island, it was a classic nucleated village, surrounded by a fence. But, given this is the "horticultural" north, where it appears the settlers gardened more successfully than in the south, it may prove to be something different again.
However, so far, they've only excavated less than 1 per cent of the site, so there is much to be done.
"One of the problems with archaeology is that it's very labour intensive," Robinson says.
"Trying to understand an area that goes 100 to 150 metres back in these hills is going to require quite a different approach to digging a few holes."
To do it properly, they want to do a laser scan that will give a detailed contour of the site, and then a grid of auger holes to reconstruct the stratigraphy of the land.
At the back of the meadow, where the gardens and the houses were, Robinson says they hope to use ground-penetrating radar and a magnetometer - an instrument sensitive to magnetic fields - to find things like kumara pits or fireplaces. Only after that will the spades come out.
If their current funding bid is successful, work will start in earnest in early 2019.
While the great hope is that somehow they can prove Mangahawea was one of the very first places where people from the Pacific arrived, for most of the scientists, just finding a site that is so well-preserved in Northland - where continuous habitation has wrecked much of the other evidence - is special enough.
"We really don't know much about the early phase of settlement in the Bay of Islands at all," Walter says.
"It's a really unusual period in New Zealand history, and we are in a unique position, one of very few countries in the world where you can see the first sites, the first occupation of the landscape. It's very rare."
The excavation has also been a chance to lift a weight from their shoulders. Ever since discovering the notebook, Blanshard says he carried the burden of the unfinished work at the site, and felt a kind of "moral imperative" to see the research complete.
"As archaeologists, and we all know it, if you go and dig a hole you can't go re-dig that hole. That hole is dug, all the information is gone," he says.
"It's actually a destructive science, so as good custodians it's actually beholden on us to do the analysis once we've got the holes. Because otherwise we are just digging holes."
After working in partnership with iwi, that burden only grew heavier, in the realisation that the artefacts their predecessors had taken from the site were also precious taonga.
"One of the things that weighed on my mind was when we had huge floods in 2007 and I returned to the site and the river had ploughed it out, and I just thought, 'whoa'. Like if we don't do something now we may never get the chance," Blanshard says.
"There's a really good chance this material could have been lost almost forever, and Mangahawea would have just become another one of the myths of archaeology."
As we leave the site, it's after noon. We climb up the track, until we are high above the site, the hāngī pit just visible below.
Clendon says a karakia, thanking the ancestors for looking after us. We look down and see Mangahawea as his tīpuna must have, waves crashing gently on the shore, the protective arm of Rangiātea sheltering the sandy bay, empty but for the flaxes rustling in the breeze.