60 SHORT NOTES NOTORNIS 39

I thank the late G.W. Johnstone (Antarctic Division, Australia) for

confirming the identification of the penguin from photographs, Robert Jones
and Jeremy Smith for helping to measure and weigh the penguin and
commenting on the manuscript, and Peter Dann for commenting on the
manuscript.
LITERATURE CITED
MARCHANT, S.; HIGGINS, P. J. 1990. Handbook of Australian, New Zealand and Antarctic Birds.
Vol.l(A). South Melbourne: Oxford University Press.
NAPIER, R.B. 1968. Erect-crested and Rockhopper Penguins interbreeding in the Falkland Islands.
Br. Antarct. Surv. Bull. 16.71-72.
REILLY, P.N.; CULLEN J.M. 1982. The Little Penguin Eudyptula minor in Victoria. 111. Dispersal
of chicks and survival after banding. Emu 82:137-142.
TUCK, G. S.; HEINZEL, H. 1980. A Field Guide to the Seabirds of Australia and the World. London:
Collins.
WATSON, G.E. 1975. Birds of the Antarctic and Subantarctic. Washington D.C.: Amer. Geophys.
Union.
CLAIRE SPEEDIE c/o Max Downes, Wildlife Management Consultancy,
Bridge Street, Foster, Victoria 3960, Australia
Present address: Penguin Reserve Committee of Management, Phillip Island
Penguin Reserve, P O Box 97, Cowes, Victoria 3922, Australia

Weka predation on eggs and chicks of

Fiordland Crested Penguins
Between July and October 1989, we studied the breeding biology of
Fiordland Crested Penguins (Eudyptespachyrhymhus) on Taumaka, the larger
of the Open Bay Islands (43O 52' S, 168O 53' E), 5 km west of Haast. The
Fiordland Crested Penguin is the rarest of the six crested penguin species, .
breeding only on the rugged coasts of South Westland and Fiordland, and
on Solander and Stewart Islands and offshore islands. Like its congeners,
it lays a clutch of two different-sized eggs and only one of these, usually
the larger, results in a fledged chick (Warham 1975). Displacement and
desertion account for most of the natural egg mortiality, whereas chick
mortality most often results from starvation or exposure to bad weather.
These penguins appear to have no native, naturally occurring land predators
(Warham 1974), but introduced rats, cats and stoats may take eggs and chicks
in some areas.
The Open Bay Islands have no introduced mammalian predators and
so are an important breeding place for several species of seabird. Although
no formal census has been made, the breeding population of Fiordland
Crested Penguins may be in the order of 300 - 400 birds on the island of
Taumaka. Among the nests watched in our study, predation by the
introduced Weka (Gallirallus australis) was the greatest single cause of egg
mortality and contributed significantly to chick loss.
We monitored the contents of 61 nests daily and determined the cause
of egg loss by the nature and location of shell remains and by the presence
1992 SHORT NOTES 61

or absence of incubating adults. If adults were absent when we began daily

checks of a nest, we assumed the nest had been deserted, regardless of the
presence or condition of shell remains. In contrast, adults that had displaced
eggs or lost eggs to predation typically remained at the nest site for several
days. Weka predation was indicated by transverse openings in the shell
remains.
Of 115 eggs laid, 57 failed to hatch. Twenty-two (38%) of these were
taken by Wekas, 13 (23%) were abandoned, 4 (7%) were displaced, and
5 (9%) were addled or infertile. The remaining 13 (23%) were missing during
the daily checks, no shell remains were found, and so their fate was unknown.
Early in incubation, when penguins were least tenacious at incubation, our
presence in the colonies sometimes made penguins leave their nests
temporarily, thereby creating opportunities for Weka predation. In addition,
it is possible that some of the egg loss we attributed to Weka predation was
of eggs displaced by incubating penguins and then eaten by Wekas.
We once saw a Weka attack an egg while the parents were at the nest.
It pierced the egg while the breeding pair stood dozing beside it. In general,
crested penguin species stand over but do not incubate the first egg until
the second is laid (Warham 1975). At the sound of the shell breaking the
pair awoke and drove the Weka from the nest. The penguins did not recover
the damaged egg from where it had rolled during the brouhaha.
A small but substantial proportion of chick mortality also was due to
Wekas. Of the 55 eggs that hatched at nests we visited daily, 11 chicks (20%)
were taken by Wekas. Another 7 (13%) were abandoned or displaced, 13
(24%) starved, and 3 (5%) were missing. Only 21 (38%) of the chicks were
alive by the end of the study, 6-8 weeks before fledging. Nine of the 11 chicks
taken by Wekas were taken after they had entered the creche stage, the period
during which chicks are left unguarded by day while both parents forage.
Chicks entered the creche stage as early as 14 days after hatching in 1989,
younger than has been reported previously (Warham 1974; S. Phillipson,
pers. cornrn.) and this may have made them more vulnerable to Weka
predation. Several of the chicks taken appeared to be large and healthy; the
heaviest was 620 g and 17 days old. It appeared to have been killed by
repeated beak blows to the head in a manner similar to the territorial fights
we observed between breeding adult Wekas. Flesh was torn from the back,
neck and abdomen of the chick, although it was still alive when we reached it.
Chick predation did not appear to be associated with our disturbance;
most chicks were taken while we were away or watching from a distance.
On two separate occasions, I. McLean saw a Weka take a live chick from
.within a few centimetres of its brooding parent. Apart from direct
observation, we distinguished Weka predation by chick remains and by the
behaviour of parents. Those birds that had lost a chick usually returned to
the nest site for long periods, apparently waiting for the chick to return.
Breeding birds with live chicks tended to return only during the evening
or night for short feeding visits and were seldom seen by day.
Most penguins were indifferent to the movements of Wekas, which could
easily move between nests. However, some pairs, mainly those from which
a Weka had previously attempted to take an egg or chick, were clearly hostile
62 SHORT NOTES NOTORNIS 39

and actively drove Wekas from the vicinity of their nests. Of the nests we
watched regularly, there was not a single nest location that Wekas did not
visit at least occasionally.
Wekas were introduced to Taumaka early in this century, probably
between 1905 and 1912 (Stirling & Johns 1969). A smaller, probably self-
introduced, population is on the island of Popotai, about 100 metres
southwest of Taumaka. They were introduced to Taumaka and other offshore
islands apparently to be food for shipwreck survivors. Douglas was prophetic
when he wrote, circa 1899, that there were
...numbers of uninhabited islands on which shipwrecked mariners often die of
starvation, or at least live miserably. Why not put Maorie [sic] hens [Wekas] on
those islands? They would have no enemies in such places, they are easy to catch,
good healthy food, and will live and thrive anywhere eating and digesting anything....
(Pascoe 1957)
From a small initial population, Wekas have increased on Taumaka to
an unusually high density. They occupy every conceivable niche from the
intertidal zone to seemingly inaccessible cliff ledges. They seemed highly
territorial throughout our study, although territories were much smaller than
reported for a mainland population (Beauchamp 1987). We could often see
up to five birds patrolling a small area and squabbles were common.
Wekas appear to have greatly altered the characteristicsof the vegetated
areas. Ground cover was continually uprooted by Wekas foraging for insects
and other prey. Burrows (1972) reported that extensive areas of vegetation,
particularly rock plants, were severely harmed by Weka activity.
The effects of Wekas on native fauna area also severe. We found the
remains of at least seven Fairy Prions (Pachyptila turtur) in and around Weka
lairs. Wekas routinely entered the small burrows and holes where prions,
Sooty Shearwaters (Puffinus griseus) and Blue Penguins (Eudyptula minor)
were known to nest or roost. The New Zealand land leech (Hirudobdella
antipodum), known only from Open Bay and Snares Islands, may also be
eaten by Wekas (Stirling & Johns 1969). .
Wekas were capable of climbing to heights of 2 m among the kiekie
(Freycinetia banksiz] that covers most of the island. Kiekie flowers appeared
to be the only food sought, but C. Miskelly (pers. comrn.) has seen Wekas
take tree wetas (Hemideina thoracica) from the leaf bases of kiekie on
Taumaka. Tree-nesting birds and a gecko species (probably Hoplodactylus
grcmulatus) may also be vulnerable to predation as a result of this behaviour.
Although Wekas have clearly affected the populations of native species
on Taumaka, the severity of this impact has not been assessed. Stirling &
Johns (1969) believed that Wekas were m o w i n g the natural fauna so greatly
that they should be removed. Most likely, Weka numbers and damage have
increased since the 1960s.
Removing Wekas from Open Bay Islands, as has been done on Codfish
and other islands, could restore native populations. However, Wekas have
disappeared from much of their natural range and are elsewhere declining.
Recent attempts to re-establish the North Island Weka (Gallirallus australis
greyz] in the Waitakere "
1992 SHORT NOTES 63

Therefore, removal of Wekas from Open Bay Islands could mean the loss
of a potentially important reservoir.
Nonetheless, Weka predation on Taurnaka may be nearing the threshold
at which other species cease to co-exist. We believe that a detailed study
of the effects of Wekas on the flora and fauna of the Open Bay Islands is
urgently needed. These effects must then be balanced against the value of
the islands' otherwise pristine nature, the importance of native populations
that will require protection if they are to remain, and the value of a thriving
Weka population. Removal of Wekas, possibly only to the adjacent mainland
where they are reported to have declined in recent years (B. Glubb, pers.
comm.) may be an ideal and relatively inexpensive solution.
ACKNOWLEDGEMENTS
Financial and logistical support for research on Open Bay Islands was
provided by the Department of Zoology, University of Canterbury; the New
Zealand Lottery Grants Board; the Ornithological Society of New Zealand;
the Stocker Scholarship of the Royal Forest and Bird Protection Society;
and the World Wide Fund for Nature (WWF) - New Zealand. We thank
I. McLean and J. Schieck for help in the field, B. and P. Glubb for providmg
radio contact during the study, and B. Heather, I. McLean, C. Miskelly,
H. Robertson and C. L. St. Clair for improving the manuscript.
LITERATURE CITED
BEAUCHAMP, A. J. 1987. The Social structure of the Weka (Gallirallus australis) at Double Cove,
Marlborough Sounds. Notornis 34: 317-325.
BURROWS, C. J. 1972. The flora and vegetation of Open Bay Islands. J. Roy. Soc. NZ 2: 15-42.
MacMILLAN, B.W.H. 1990. Attempts to re-establish Wekas, Brown Kiwis and Red-crowned
Parakeets in the Waitakere Ranges. Notornis 37: 45-51.
PASCOE, J. 1957. Mr. Explorer Douglas. Wellington: A.H. and A.W. Reed.
STIRLING, I.; JOHNS, P.M. 1969. Notes on the bud fauna of Open Bay Islands. Notornis 16:
121-125.
WARHAM, 1974. The Fiordland Crested Penguin. Ibis, 116: 1-27.
WARHAM, J. 1975. The crested penguins. Pages 189-269 171 The B~ologyof Penguins. Stonehouse,
B., ed. London: MacMillan.
COLLEEN CASSADY ST. CLAIR'72and ROBERT C. ST. CLAIR2
'Department of Zoology, University of Canterbuy, Christchurch 1,
New Zealand
'Present address: Department of Zoology, University of Oklahoma, Norman
Oklahoma, U. S.A., 73019

Unique, dark olive-green moa eggshell from Redcliffe Hill,

Rakaia Gorge, Canterbury
Moa eggshell is known from many, widely distributed localities throughout
New Zealand, having been most frequently found in dunesand and loess
deposits and beneath rock overhangs which have been used as nesting sites.
It also occurs on archaeological sites and in swamps and caves. Most is buff
or creamy white, but pale green moa eggshell has also been collected, notably
from Otago (see Hutton 1876, White 1886). This pale green shell is similar