There are four of us in the Murchison Mountains, walking slowly uphill through a boulderfield smothered in alpine flowers and shrubs. I am an all-too-infrequent visitor to such places, and my eyes are feasting on the garden at my feet: the glossy green saucers of Mt Cook buttercups, cushion plants that seem to glue the rocks together, the feathery tops of wild carrot and zipper-like leaves of whipcord hebe, spaniard with its wicked sock-piercing spines, the snow-white loveliness of gentian.

My companions’ focus is elsewhere. They scan the valley slopes, straining their eyes against the harsh sunlight, trying to coax an extra pixel or two of acuity that will enable them to spot the flicker of movement that signals a rock wren, New Zealand’s avian mountaineer. Every now and then they stop to twitch a bird caller, a cunning device that sends out a squeak when you twist its metal core.

We reach an old rockfall, a favoured wren habitat, and, sure enough, we spot a pair of wrens bobbing and curtseying and making short darting flights from boulder to boulder. After some discussion about the best place to erect a mist net (which works best if fitted into a natural depression in the landscape, so that the birds can be funnelled into it) we set up three aluminium poles and string a fine black mesh between them, reaching about three metres off the ground.

The net is constructed as a series of billowing pockets, so that a trapped bird will fall into the belly of a fold rather than bounce off a taut wall of netting. A pair of loudspeakers is set up, and a staticky recording of high-pitched tweets hisses out into the silent landscape. The stage is set.

Then we set about herding wrens, as if it were a sheepdog trial. Megan Willans, the Department of Conservation’s point person on rock wrens in Fiordland, acts as the heading dog on one side, with me on the other. We are helped by her colleagues; Martin Genet, the huntaway, works the target down the slope from above, while Hannah Edmonds crouches below the net, ready to disentangle the captive, should we be so lucky.

We try to predict the wren’s net-avoidance strategy while clambering down the rocks, shepherding it towards the net. Suddenly the bird makes a long, low flight into the meshes.

Edmonds disentangles the wren and hands it to Willans for banding while Genet organises “lunch”—half a grape, a dish of meal worms and a square of moistened bread—and puts the food into a transfer box. As Willans bands the bird, she points out a long loop of dead skin around one of its legs, like some sort of ethnic ankle bracelet, or a stocking that has fallen down. It is one of the primitive features of New Zealand wrens that they slough skin like a reptile.

Once our wren’s gaudy identification bands are in place, it is popped into the box, where it scuffles about briefly, then settles. We give each other high-fives, as if we’d just won a point in beach volleyball. But of course this is no game. We’re here because rock wrens are in trouble. Even here in the mountains, far from the nodes of human settlement, the rising tide of introduced predators is nipping at their toes. Mice enter their nests and eat the eggs. Stoats, stealthy agents of death, take adults and young alike. The birds are already living above the treeline, often nesting in almost vertical bluffs, where they can’t go much higher. In response, DOC has decided to establish a satellite population on a predator-free island—as insurance, in case the little mountaineers perish.

[Chapter Break]

When the great New Zealand nature writer and birder Herbert Guthrie-Smith encountered rock wrens while walking the Milford Track in the late 1930s, he wrote, “Xenicus gilviventris, I am glad to think, is one of the species likely to survive changes that from the forester’s and field naturalist’s point of view have desolated New Zealand. The ravages wrought elsewhere by deer, rabbits, opossums, birds, and other imported vermin are unlikely to affect the welfare of the rock wren. Even weasels and rats—and I know they ascend to great heights—are hardly likely to draw sufficient recompense in prey from such unpeopled solitudes… With cover and food supplies unmodified, the rock wren may be considered relatively safe.”

It didn’t take long for Guthrie-Smith’s optimism to seem misplaced. By the 1980s, when Otago University student Sue Michelsen-Heath was beginning her MSc research on the breeding biology of rock wrens, hunters and trampers were already reporting fewer sightings of the bird. Michelsen-Heath spent months counting, banding and studying wrens in an area of the Murchison Mountains, west of Lake Te Anau—the same area where takahe were rediscovered in 1949.

She remembers finding a nest that had been torn apart by a stoat. The female returned repeatedly to the nest site and pecked among the remains, searching for her chicks. “I realised we had to do something about the predation,” she said. “The situation was only going to get worse.”

It wasn’t until 2006, however, that alarm bells really began to ring. The summer of 2005/06 had seen a prolific and simultaneous seeding of beech and tussock—a so-called mast seeding event. Such events produce an abundance of food for rodents and stoats and lead to population booms in the following year.

Willans, a biodiversity officer at DOC Te Anau who had been working on takahe, realised that 2006 would be an ideal year to assess the impact of predation on rock wrens, as mouse and stoat densities would be high. Over three months in the summer of 2006/07—an atrocious El Niño summer, when snowstorms lashed the Murchisons repeatedly—248 person-hours were spent searching for wrens near McKenzie Burn. Thirty-four pairs were found, of which 14 were observed to produce fledglings. Two nests were predated, one by mice and the other by stoats.

The pressure from predators is likely to get worse. The wrens inhabit areas above the altitudinal limit for ship rats, effectively removing one predator from the list. But this protective barrier could be undone by climate change. Already living in the alpine habitats, they can climb no higher, yet as the world warms under the malignant influence of climate change, predators will also rise to threaten them.

But the most troubling statistic of the 2006/07 search was that there appeared to have been a 44 per cent decline in numbers in the area since Michelsen-Heath’s previous census, in 1984/85. If rock wren populations were dropping fast in remote Fiordland, they could hardly be faring better further north. An analysis of records between 1912 and 2005 confirmed the trend: the rock wren’s range had decreased by a quarter since 1984.

Xenicus gilviventris, it seemed, might be going the way of its cousins.

[Chapter Break]

Once there were seven. Seven diminutive birds—four flightless and three aerially challenged—in a family so ancient they have no close affinity to any other living bird. And so different are they one from the other that they occupy five separate genera. Dubbed wrens, though genetically unrelated to their Northern hemisphere counterparts (see sidebar), this diverse cluster of species is now considered as characteristic of the fauna that evolved on Aotearoa as its supersized neighbours, the moa.

When flightlessness became a liability in this country, thanks to the introduction of rats and other ground-hunting predators, it was predictable which of the wrens would go to the wall first. The long-billed wren, with its scimitar-shaped beak for probing into deep crevices, and the North and South Island species of stout-legged wren, plumpest and heaviest of the family, were expunged in pre-European times, victims of kiore, the Pacific rat.

Remarkably, these three species were not identified until the 1980s, when researchers began systematically examining the wren bones that had been gathering dust in museum collections. The long-billed wren has the distinction of being New Zealand’s rarest fossil bird, its existence known from the remains of only six individuals.

Lyall’s wren, the fourth of the flightless quartet and the smallest flightless bird ever to have existed (weighing about 22g, a little heavier than a rock wren), found temporary sanctuary on kiore-free Stephens Island, in Cook Strait. Known initially as the Stephens Island wren (before its fossil remains were discovered at mainland sites as well) its demise has become famous, told and retold as the tale of the lighthouse keeper’s cat. The story goes that in 1894, shortly after the Stephens Island lighthouse was commissioned, the principal keeper’s cat started bringing carcasses of an unusual bird to the door of his house. The keeper, David Lyall, sent one to colonial bird expert Walter Buller, who immediately recognised it as a new species and set about describing it. Meanwhile, the cat kept delivering dead wrens, some of which were sold to a curio dealer named Travers, who saw the commercial value of the birds and tried to sell them for £50 apiece. After a few months, the cat stopped bringing wrens to the keeper’s door. The wretched animal had not only discovered but exterminated the only flightless passerine ever to have been seen by Europeans.

[sidebar-1]

In reality, many cats were involved, not just the apocryphal “Tibbles” of the story. By 1899, the keepers were shooting hundreds of feral cats to try to control their spread. But as Richard Holdaway and Trevor Worthy point out in their definitive work on the prehistoric life of New Zealand, The Lost World of the Moa, the almost total deforestation of Stephens Island to establish a farm for the lighthouse station was as big a factor in the wren’s extinction as out-of-control cats, not to mention the professional collectors who made a frenzied rush to the island to secure specimens for their overseas patrons.

Lyall’s wren was formally described in 1894 and extinct by 1895. Walter Buller’s comment that the scientific world heard “almost simultaneously of the bird’s discovery and of its disappearance” remains one of the saddest lines in New Zealand’s avian literature.

The bush wren, or matuhi, was the next to go. Again progressively eliminated from the mainland by rats, it clung to survival on a few muttonbird islands off the south coast of Stewart Island. Herbert Guthrie-Smith observed the wrens there in 1913, watching with pleasure as they moved “through the darkling underscrub like a forest gnome, like a woodland brownie” and reporting that their conversational calls sounded like “the noise of a small wrist-watch in process of winding“. Their powers of flight were weak, limited “to a flutter of a few yards”—a fact that has led some to describe the species as flightless.

Tragedy struck in 1964, when black rats came ashore on Big South Cape, the bush wrens’ final holdout. New Zealand Wildlife Service rangers—among them a new recruit named Don Merton—attempted an ambitious species rescue by shifting six wrens to a predator-free island. But it was too little, too late. The relocated birds failed to breed, and the last official sighting of a matuhi was in 1972.

Along with the bush wren, two other species which were surviving on Big South Cape—the Stewart Island snipe and the greater short-tailed bat—were exterminated during the rat invasion. Merton, seething with frustration over the loss of three endemic species in one swoop, vowed that never again would he acquiesce with official policy to let nature take its course, and went on to mastermind translocations that saved the Chatham Island black robin and kakapo.

Now just two wrens remain: titipounamu, the rifleman, and piwauwau, the rock wren. “Titipounamu” refers to the greenstone-coloured plumage of the rifleman, while “piwauwau” means “little complaining bird”—a touch unfortunate, given its merry tweet. Another name for rock wren is tuke, which can mean elbow, a possible but misleading reference to the rock wren’s prominent stripe above the eye (which is by no means elbow-angled), or twitch, a logical allusion to the wren’s distinctive bobbing up and down. The rifleman’s European name is thought to have arisen either from its habit of foraging around tree trunks in a spiral pattern (as per the rifling of a gun barrel) or from its similarity in colour to the uniforms of a colonial regiment.

An adult rifleman—New Zealand’s smallest bird—weighs a mere six to seven grams, the weight of a box of matches. Rock wrens weigh twice that much, while the extinct stout-legged wren may have reached 50 g—heavier than a bellbird.

Riflemen are found in forests and scrublands of both main islands, but rock wrens are now restricted to fellfields, rock jumbles and other alpine habitats along the mountainous spine of the South Island.

For a long time, people wondered how rock wrens survived the brutal alpine winters, when snow might lie on the ground for up to five months. (In the Murchisons, for instance, the average winter temperature is -0.5ºC.) Some theorised that the wrens went into some form of torpor. Others suggested they moved down into the bush.

In the 1970s, photographer Rod Morris, then a Wildlife Service ranger, was searching for kakapo in winter on the steep slopes of the Tutuko Valley, near Milford Sound. “There was a thick blanket of snow over the entire area, but suddenly, from under a boulder, out hopped a wren as lively as on any sunny day.” Morris thinks that rock wrens can survive quite happily in the alpine winter because of the diversity of their habitat. He has watched them go underground in rock jumbles for five minutes or more. “They’re obviously foraging for invertebrates,” he says.

Like riflemen, rock wrens are mainly insectivorous, although they also eat berries, seeds and even sip the nectar from flax flowers. In the open, they sometimes flutter up and catch insects on the wing, but primarily they peck in the substrate, grubbing for millipedes, beetles and so forth. Michelsen-Heath has watched them systematically gleaning through hebe and coprosma bushes, moving around the branches like mice.

They lay their eggs in late spring, with both sexes incubating the clutch (typically three cream-coloured eggs) through the last of the winter snows. Chicks hatch without a feather on their bodies. Their parents feed them soft-bodied insects such as caterpillars for the first week or so, then step up to more chitinous fare as the springtime flush of arthropods makes a wider range of prey available. By the time the chicks fledge, at around three weeks of age, the snow has melted and the alpine valleys and rockeries are pulsing with insect life. Even after fledging, the young continue to beg food from their parents. Michelsen-Heath has seen an adult stuff a two-centimetre weevil down the throat of a chick that was about six weeks old, the age when young birds are ejected from the family group and forced to fend for themselves.

The nests are cosy, roomy cavities excavated into crevices in the rocks, with walls that have been densely woven from tussock stalks, moss, lichen and twigs and insulated with a thick felting of feathers that can be up to eight centimetres thick. Guthrie-Smith, patient naturalist that he was, once counted 791 feathers in a single nest. Not only did he count them, he identified each one. More than 700 were weka feathers, and the rest were from kiwi, kakapo, kea and pigeons.

This nest-feathering habit had an unusual spin-off in the search for the last remaining Fiordland kakapo. Rod Morris remembers being told by his field supervisor, Marsh Small, to look in rock-wren nests to check which feathers they had used. If there were fresh kakapo feathers, there was a good chance that one of the big parrots was living in the vicinity.

The only snag was locating the nests, which have an entrance “no bigger than a mouse hole”, says Morris. “The trick to finding them was to carry a little bag of goose down that you’d pulled from the bottom of your sleeping bag. If you saw a wren, you would try to blow a bit of down in its direction. To a wren, a feather is like gold, and it would drop whatever it was doing, grab the feather and fly straight back to its nest.”

Of course, if humans can locate wren nests, how much more readily can a scent-following mouse or stoat sniff out the occupants? In the long term, rock-wren survival turns on Guthrie-Smith’s question: how much recompense in prey does a predator require “from such unpeopled solitudes” before it takes up residence there?

With five out of seven wren species already extinct, no one cares to wait around for the answer. But the reality is that DOC’s threatened-species budget stretches only so far, and were it not for private funding this wren work would not be happening.

At his home not far from the Karori Wildlife Sanctuary in Wellington, Barry Dent told me how he and his partner, Sue Freitag, came to be involved. Both have scientific backgrounds—Dent as an organic chemist, Freitag as a geochemist—and both have long been interested in conservation “in all its manifestations”.

In 1996, Dent formed BDG Synthesis to provide materials and services to pharmaceutical companies. As the business grew, he and Freitag began investing a portion of each year’s profits in conservation and social development projects.

When considering which projects to support, they took a pragmatic, one could say analytical, view. “Interventions on behalf of plants and marine life seemed to call for large-scale land purchases or legislative change to have much impact, whereas the predation threat to terrestrial vertebrates and invertebrates could be addressed by small, discrete projects.

“Many species, though vulnerable, are below the threshold at which urgent action is required. We realised we could provide a quantum of intervention for species not yet at the critical-threat stage, slotting in under the level at which government action kicks in.”

It’s a commando-style approach to conservation—short, focused sorties rather than a full army deployment. To Dent, it makes good fiscal sense. “If you wait too long to act, everything becomes a kakapo-type crisis.”

As well as funding the wren work, BDG has put money into such ventures as the translocation of mohua to Whenua Hou/Codfish Island and brown teal to Tuhua/Mayor Island, the building of a hut for cat-trappers on Stewart Island, and a researcher’s trip to look for snipe on Campbell Island. The couple don’t just donate money, but time. Freitag is part of Forest and Bird’s home-nursery programme, in which individuals grow endemic plants for use in land-restoration projects, and recently they both spent a month as conservation volunteers in the Chatham Islands, feeding endangered petrel chicks. A few days after we talked, Dent flew to Te Anau to help catch more wrens for the Secretary Island flock.

[Chapter Break]

A thud of rotor blades; the helicopter is here. We have caught five wrens in a little over three hours—a terrific result, the DOC team agree. Some days, if the birds are up on the bluffs, they don’t catch any. We have two probable pairs and one extra male. It can be difficult to tell: although rock wrens are sexually dimorphic, with males more vividly coloured than females, a freshly moulted female can look very similar to a male. Curiously, rock wrens in the Mt Cook area are much browner than the Fiordland birds, while those further north, in the mountains around Nelson, are closer to the Fiordland green colour. It is possible that the Fiordland wrens are a separate subspecies.

We load the gear and transfer boxes and climb in for the flight to Secretary Island, at the head of Doubtful Sound, due west of the capture site. We fly above a heartstopping world of blackwater tarns, forested gullies, tussock basins and ice-scoured crags. One nameless lake brims to the top of its enclosing basin like an infinity pool. A slot in the rim lets a white trail of water fall hundreds of metres to the valley below.

The pilot lands on a snowgrass ridge and Willans and Edmonds carry the transfer boxes to a promising bluff. First they let the pairs go, confirming sex by comparing birds in the hand on release, then open the lid to let the male out. A flutter of green and they’re off to make a new home by the sea.

This release brings the tally to 14 rock wrens on the island. Willans hopes to release a total of 40 birds there over the next couple of years. An earlier translocation of wrens to Anchor Island, in Dusky Sound, failed, but it was made earlier in the season, and she thinks it may have been too soon for the birds to regain condition after breeding. Our release is in March, well after the year’s juveniles have left the parental nest and found their own mates. (Rock wrens breed in their first year of life, and have a natural lifespan of around five years.) Moreover, Secretary Island (8140 ha) is five times larger than Anchor Island, and three times higher, providing a better habitat match. In fact, after the North and South Islands, Secretary is the tallest island in New Zealand. Its rodent-free and almost stoat-free status makes it the most suitable place to establish a rock-wren population.

The flight back to Te Anau has, if not a “mission accomplished” feel to it, at least a sense of mission under way. I’m glad to have been a tiny part of it. Even if rock wrens should prove ultimately resilient to predation on the mainland, the precautionary approach is the right one.

Guthrie-Smith, who saw the writing on the wall for many of New Zealand’s unique plants and animals, wrote, “Why should man and the rat possess the face of the habitable globe? Why should the sparrow be the only bird? I don’t say we shall quite come to that, but certain it is that with every species eliminated, by so much is the world robbed of light and colour.”

By establishing a new population on an island sanctuary, we gain the hope that the rock wren’s flash of green feathers, perky chirp and bobbing gait will remain part of the living world, and not just the historic record.

A chime of wrens

The world is rich in wrens. There are 80-odd species of “true wren” belonging to the family Troglodytidae (so named because Carl Linnaeus noticed that the European wren liked to fossick in dark crevices). In addition, there are two dozen unrelated Australasian wrens and two surviving New Zealand wrens, not to mention wren combos such as antwrens and wren-babblers. Basically, any bird that was especially small, had short wings and ate insects tended to be called a wren by explorers and naturalists trying to make sense of foreign faunas.

What all these birds have in common is that they belong to the singularly bountiful branch of the avian family tree known as the passerines, or perching birds, called that on account of the twig-gripping deployment of the toes: three forward and one back. There are around 5400 species of passerine, which together make up a little over half the world’s birds.

For years, ornithologists have puzzled over how New Zealand’s wrens, the Acanthisittidae, fit into the passerine lineage. First they were placed in the oscines sub-group (the virtuosi among songbirds), then in the suboscines subgroup (tweeters rather than singers). Now the consensus is that they don’t belong to either but are, as Jamie Wood, a palaeoecologist with Landcare Research, puts it, “an ancient radiation that is basal [that is, fundamental] to all other passerines”. If this interpretation is correct, then New Zealand’s wrens were hopping around long before the sparrow and its ilk arrived on the evolutionary scene. Furthermore, the archetypal songbird—the ur-passerine—probably looked quite similar to one of our wrens.

Wood, in collaboration with Trevor Worthy of the University of New South Wales and Alan Cooper of the University of Adelaide, are in the process of analysing the mitochondrial DNA of all New Zealand’s wrens, extinct and extant, to see how long ago they diverged from what would become the main passerine line. The work grew out of the recent discovery of 18-million-year-old acanthisittid wren bones in an ancient lake deposit in St Bathans, Central Otago. These bones, especially four tiny but diagnostic ankle bones, were found to be the same size as and closely related to those of a modern rifleman, suggesting that the rifleman may have already diverged from its closest relatives, the rock and bush wrens, by this period.

The idea is that if certain fast-mutating DNA segments can be located in the genomes of all seven wren species, it should be possible to use the New Zealand wren lineage as an evolutionary clock to find out when our wrens arrived (presumably from some part of the rupturing supercontinent) and even when the passerines began their ascendancy.

If the wren DNA sequencing is successful, the resulting phylogenetic tree will provide a snapshot of a critical period in the radiation of birds. New Zealand’s tiniest birds, long overshadowed in their importance by moa and the other flightless giants of our avifauna, could shed light on one of the great moments in evolutionary history.

Kennedy Warne

+ Read sidebar