The mystery of the giant hoho weevils of Rangatira Island

The mystery of the giant hoho weevils of Rangatira Island

Te Papa curator Colin Miskelly has recently returned from volunteering for the Department of Conservation on the Chatham Islands. His inordinate fondness for weevils led to some new discoveries about one of Rangatira Island’s more cryptic yet spectacular inhabitants.

The coxella weevil Hadramphus spinipennis was one of the 50 or so Chatham Island insects discovered by Thomas Hall, an itinerant shepherd and rabbiter who sacrificed his life on the battlefields of western Europe during the first world war. For a few years around 1906-08, Hall worked for Richard Paynter on Pitt Island in the Chatham Islands, where he collected many insects then unknown to science.

His collection ended up in the hands of the entomologist Major Thomas Broun, a prolific describer of new beetle species. Among Hall’s haul was a large flightless weevil (about 20 mm long) that he found only on a large feathery-looking herb, then known as Coxella (also known as soft speargrass, Aciphylla dieffenbachii).

Mating coxella weevils, Rangatira Island, March 2018. Photograph by Colin Miskelly

Coxella is highly vulnerable to browsing from farmstock, and as a result became confined to inaccessible cliffs and outlying islets of the Chatham Islands. And on the two main islands of the group, predation by introduced rats and mice soon extirpated the coxella weevil, resulting in its current highly restricted range.

Coxella speargrasses in the Department of Conservation nursery at Te One, Chatham Island. Photograph by Colin Miskelly

Working on weevils

By the time naturalists started paying attention to the distribution and ecology of the coxella weevil, it was found only on Mangere and Rangatira Islands, both now protected nature reserves off the coast of Pitt Island. The weevil was found to be entirely dependent on coxella, with the larvae feeding on the large tap root, and the adults emerging at night to browse on the stems, leaves, and flower heads.

The plants and weevils were easily found and studied on Mangere Island, where large patches of big healthy coxella plants (up to a metre high) are common, including near the Department of Conservation hut. But on Rangatira the few coxella plants are short and stunted, and are mainly found on the opposite side of the island to the DOC hut. Mangere Island therefor became the main study site for coxella weevil ecology, including a thesis study by Katrin Schöps in the 1990s.

Coxella weevils (Hadramphus spinipennis) feeding on coxella (Aciphylla dieffenbachii) at night, Rangatira Island, March 2018. Photograph by Colin Miskelly

Mangere and Rangatira Islands hold the entire world population of black robins, a critically endangered bird that was reduced to only five individuals in 1979. An intensive recovery programme has built up their numbers to a current population of about 280 adults. The black robin recovery programme plus the concurrent Chatham petrel recovery programme have meant that scores of naturalists have spent long periods of time on Rangatira.

From 1970 on, field workers started reporting seeing occasional large Hadramphus weevils on the trunks and stems of hoho trees and saplings on Rangatira, many hundreds of metres from known patches of coxella speargrass.

Black robin, Rangatira Island, March 2018. Photograph by Colin Miskelly

Barking up the wrong tree

Hoho is the Chatham Island lancewood (Pseudopanax chathamicus), and grows into a canopy tree with a trunk up to 40 cm in diameter. It is a minor component of the forest in the interior of Rangatira Island, whereas the speargrass patches are on the coastal fringe. What were these large flightless weevils doing on the wrong species and type of plant in the wrong habitat so far from their known host plants?

Hoho (Pseudopanax chathamicus), Rangatira Island, March 2018. Photograph by Colin Miskelly

Initial suggestions were that adult weevils were searching for new patches of speargrass, and were making do with hoho as an alternative food source on their wanderings. It was assumed that they were climbing up into the forest canopy to browse on the hoho leaves. Katrin Schöps even ran a food-choice experiment, providing 24 captive weevils from Mangere Island with the foliage of both plant species, and seeing which species they preferred to eat.

Unsurprisingly, the weevils preferred the speargrass – which was the plant species they were collected from (the weevils are harder to find on hoho) – and the experiment design was also based on the assumption that the weevils on hoho were seeking the foliage to eat.

Hadramphus weevil on hoho trunk at night, Rangatira Island, March 2018. Photograph by Colin Miskelly

I found a few Hadramphus weevils on the stems of hoho saplings during a fortnight-long stay on Rangatira Island in 2001, noting one as browsing on hoho leaves, and another as feeding on the 1 cm thick stem of a hoho sapling.

For the last few years, DOC staff have been monitoring these weevils on a large hoho tree near the hut on Rangatira, assuming that they were on their way up to the canopy to feed on foliage. But while photographing one at night in late March, I noticed that it had gnawed a small dent in the bark of the tree, exposing the yellow-orange cambium underneath. This weevil was feeding on wood or sap, not on leaves.

Hadramphus weevil and its feeding scar on a hoho trunk at night, Rangatira Island, March 2018. Photograph by Colin Miskelly

On the trail of cryptic weevils

A quick search of the trunk and two adjacent smaller trees revealed dozens of these little yellow scars. The weevils had been nibbling the bark all along, but we hadn’t seen the wood for the trees.

Weevil feeding scars on hoho trunk, Kokopu Creek catchment, Rangatira Island, March 2018. Photograph by Colin Miskelly

Armed with this new information on what Hadramphus weevil feeding sign on hoho looks like, I looked for other trees with the same feeding sign. Our Chatham petrel work took us through forested areas up to 2 km distant from the hut, and I soon had a list of target hoho trees to revisit at night, to see if they had weevils gnawing on their trunks.

Department of Conservation ranger Gemma Green standing beside a hoho tree with weevil feeding sign, Kokopu Creek catchment, Rangatira Island, March 2018. Note the petrel boards worn to minimise damage to petrel burrows while working on this fragile island. Photograph by Colin Miskelly

We were running out of days on the island, and bad weather was forecast (the weevils are usually only active on fine nights), but on the night of 27 March I returned to the distant scarred trees noted earlier that day, and found Hadramphus weevils on two of them in addition to the tree near the hut.

Map of Rangatira Island showing locations where blog images were taken. S = weevils feeding on speargrass, H1 = hoho tree near the hut, H2 = weevil site on summit track, H3 = weevil site in Kokopu Creek catchment.

We now know that Hadramphus weevils are living on and feeding on hoho at sites scattered all over the island. It is almost certain that these animals are using the plant to complete their entire life-cycle (i.e. larval stages as well as adults). But we don’t yet know what these hoho weevils are. They could be a cryptic new species, which look like the coxella weevils but diverged from them eons ago. Or they could be a separate ecomorph of the same species, which have recently learned to live on a different host plant, but have not yet diverged enough to be considered a separate species.

Hadramphus weevil on a hoho tree in the Kokopu Creek catchment at night, Rangatira Island, March 2018. Photograph by Colin Miskelly

Recognition of the characteristic feeding scars left by weevils feeding on hoho provides a rapid survey method for these secretive beetles, and also a way to find enough animals to include in a study of their genetic and morphological diversity compared to coxella weevils.

Hadramphus weevil on a hoho tree in beside the summit track at night, Rangatira Island, March 2018. Photograph by Colin Miskelly

With many thanks to the Department of Conservation for the opportunity to assist with Chatham threatened bird recovery programmes and to visit Rangatira Island Nature Reserve.

Related blogs


  1. Hi Colin

    This study by Emily Fountain will be of interest.

    Slightly off-topic but of interest will be this study as well.

    I thought Emily was going to sample some of the weevils from the hoho trees during her study but it doesn’t seem to show up in her paper. There may be some unpublished data still available.

    During my Rangatira trips I have seen the weevils on hoho at night on Rons, Swamp and upper Fran and Rua tracks so well inside the forest interior. Good observations on how they use the hoho as a food source.

    Also worth noting that Aciphylla (coxella) is in the Apiaceae family and hoho (Pseudopanax) is in the Araliaceae. These two families are the sole members of the order Apiales and represent a tropical/temperate pairing in that order. This suggests there may be some biochemical linkages in these plants that the weevils are exploiting.

    Really interesting blog.



  2. Wonderful Blog’s, such a great read. I look forward to seeing next season’s instalments!

    1. Author

      Thanks very much for your feedback Sarah
      Ngā mihi

  3. Great detective work Colin – another piece of the puzzle found!

    1. Author

      Thanks very much Dave. It will be fascinating to put the next piece in place, to figure out whether the speargrass-eating weevils and the hoho-eating weevils are the same thing or not – and the implications this information will have for habitat restoration and management on Rangatira and Mangere Islands.
      Ngā mihi

  4. Remarkable observations Colin.

  5. Amazing stuff, it will be interesting to see the final results of this work.

  6. Hi Colin. Agreed that I was not directly addressing the host range, although understanding the origins of the species on the Chatham’s does provide a spatiotemporal context for interpreting host plant relationships (especially if one that has a history extending back to Mesozoic time).

    In regard to the question of why these weevils are found on two such ‘completely different’ host plants on the same islands one could say why not? Further, if the host relationships are driven by biochemical, rather than taxonomic, cues, the two hosts may not be all that different as far as the weevils may be concerned.

    It would be interesting to know if the two feeding patterns reflect a taxonomic difference in the weevils (as you intimate), but the host range itself does not seem to me to be particularly remarkable (admittedly my familiarity with these beasties is limited so I might be missing something here).

    The dynamic historical sequence of island evolution and succession in archipelago situations such as the Chatham’s (and Hawaii, Galapagos, Canaries etc) may provide the kind of spatiotemporal reorganization of metapopulations that leads to changes in host plant relationships. At least that’s one theory that comes to my mind. Good luck with future investigations on this matter! And look forward to more of your interesting posts.

    1. Author

      Thanks for these comments John
      There is a long history of these insects being named and thought of as coxella weevils, but maybe they were hoho weevils all along, and Dieffenbach’s speargrass (i.e. coxella) is the secondary host plant.
      Ngā mihi

  7. Very interesting, thanks.

  8. Very nice to see these great photos of Hadramphus which represents a classic biogeographic connection between the Chatham Islands and southern New Zealand. This genus was included in the landmark study “Continuing the synthesis between panbiogeography, phylogenetic systematics and geology as illustrated by empirical studies on the biogeography of New Zealand and the Chatham Islands” by Robin Craw (1988) where it was shown that the Chatham Islands represented an overlap between two distribution patterns that were otherwise allopatric in the mainland. The weevil represented one of the southern distributions. This overlap between these different New Zealand distributions on the islands can be understood as a remnant Gondwana pattern where some of the original biota was more ‘northern’ (relative to the modern day ma) while others were more southern. In the Chathams the boundary of these two distributions comes very close together as a result of these Gondwana organisms surviving on a series of ephemeral islands in the region (Michael Heads’ metapopulation model). These organisms may seem trivial in size and geographic location, but they are outstanding features of New Zealand evolutionary history.

    1. Author

      Thanks very much for these comments John. However, the question of how and when these weevils reached the Chatham Islands is quite different from the question raised in the blog, namely why are these weevils found on two such completely different host plants on the same island?
      Ngā mihi

Leave a Reply

Your email address will not be published. Required fields are marked *