1343_complete

© Copyright Australian Museum, 2001
Avifauna from the Emily Bay Settlement Site,
Norfolk Island: A Preliminary Account
RICHARD  N. H OLDAWAY 1 AND  ATHOLL  ANDERSON 2
1 Palaecol Research, PO Box 16569, Christchurch, New Zealand
[EMAIL]
2 Department of Archaeology & Natural History, Research School of Pacific and Asian Studies,
Australian National University, Canberra ACT 0200, Australia
[EMAIL]
ABSTRACT. The avifauna of the Emily Bay settlement site, Norfolk Island, southwest Pacific, is described.
Most of the remains, which consisted of nearly 10,000 identifiable bones (mostly fragmentary) and
several thousand unidentifiable elements and fragments, were of several species of petrel and shearwater
(Procellariiformes) and boobies (Sulidae), but some land birds were also represented in small numbers.
Two species of migratory wading bird (Charadriiformes) were identified in the deposits, but no terns,
which are dominant members of the present avifauna. The taphonomy of the remains indicates intensive
use of birds as food, but some material of other than cultural origin was also present. Remains were not
distributed evenly throughout the excavated parts of the site, and were concentrated in areas where
other evidence such as post holes and fires scoops indicated points of occupation. Some species that are
present on the island and palatable were not represented in the collections: possible reasons for their
absence are canvassed. An estimate of the biomass is presented, with the proviso that the variation in
density of deposition made extrapolation to the remainder of the site problematic. The size of the sample,
the preservation of elements such as vertebrae of small petrels, and the good condition of material of
apparent natural (non-cultural) origin indicate that the collection represents a good sample of the avifauna
used as food by the Polynesian inhabitants of Emily Bay.
HOLDAWAY, RICHARD N., AND ATHOLL ANDERSON, 2001. Avifauna from the Emily Bay settlement site, Norfolk
Island: a preliminary account. In The Prehistoric Archaeology of Norfolk Island, Southwest Pacific , ed. Atholl
Anderson and Peter White, pp. 85–100. Records of the Australian Museum, Supplement  27. Sydney: Australian
Museum.
Records of the Australian Museum, Supplement 27 (2001): 85–100. ISBN  0 7347 2305 9
Norfolk Island is one of the “mystery islands” of the South
Pacific that has evidence of former human occupation, but
which had been abandoned by its Polynesian inhabitants
before European discovery (Kirch, 1988). Even the most
apparently pristine of these islands usually has evidence of
the use of natural resources such as birds, and often of the
extinction of a range of bird species (Steadman and Olson,
1985). The ability of such islands to sustain a human
population in the long term has been questioned (Kirch,
1988; Anderson, in press), because of the probable scarcity
of food resources. Faunal remains in archaeological sites
on the islands can reveal much about the interaction of
people and pristine environments.
Norfolk Island is at the southern edge of the sub-tropics.
The nearest substantial land masses are Lord Howe Island
(900 km to the southwest), New Caledonia (700 km to the
north), New Zealand (800 km to the southeast), and the
Kermadecs (1300 km to the east): Australia is 1300 km to
the west. The Norfolk Island group is so placed that it has
an avifauna with links to all the surrounding faunal regions,

86       Records of the Australian Museum, Supplement 27 (2001)
the Australian, New Zealand, and Pacific faunas (Holdaway
et al., 2001). In particular, the seabirds breeding on the group
are a mixture of species from the sub-tropical and tropical
Pacific and from the cooler waters around the northern part
of New Zealand. The Norfolk Island group is the only
available nesting ground for seabirds over a substantial area
of ocean.
Seabirds (petrels, sulids, and terns) are the most
conspicuous component of the present-day avifauna. There
are, however, no gulls or cormorants and most littoral birds
are migratory visitors or rare vagrants. There was little open
land and no large permanent areas of fresh water before
Europeans dammed two small streams at Kingston and The
Cascades. A large area of swamp behind the beach at
Slaughter Bay was recorded as being open land by the first
Europeans to land on that side of the island in 1788.
Temporary pools of water and wet grassland on Kingston
Common provide the largest area of habitat for wading birds
on the island today (Wakelin, 1968; Schodde et al., 1983).
In contrast to the marine species, most of which bred
over large areas of the South Pacific and beyond, the few
Norfolk Island land birds are mostly endemic (Holdaway
et al., 2001). Norfolk Island is remote enough from adjacent
sources of terrestrial birds to have received only occasional
successful colonizations of terrestrial birds. As the islands
are about 3 million years old, there has been sufficient time
for distinctive species to evolve in several groups, including
rails, pigeons, parrots, and passerines.
The present fauna is relict, following the extinction of
some taxa (such as Pterodroma pycrofti) after Polynesian
settlement, and further losses (such as Nestor productus and
Lalage leucopyga) in the nineteenth and twentieth centuries
after the establishment of a series of European settlements.
Fossil deposits of late Quaternary age containing bird
remains are known from both natural and archaeological
contexts in the Kingston area, on the southern side of
Norfolk Island itself, and on Nepean Island about 1 km off
the southern coast (Meredith, 1985, 1991; Anderson, 1996).
Bird remains have been found in archaeological deposits
of two ages: a Polynesian site at Emily Bay (Anderson, 1996),
and in a First (European) Settlement site dating from the late
1790s. The First Settlement site contained many petrel bones
(Meredith, 1985). Some of those which were retained in the
Commissariat Museum at Kingston, Norfolk Island, were
examined in December 1995. During a famine arising from
the wreck of the supply ship HMS Sirius at the island in March
1790, the convicts, military, and free settler population relied
on the birds for sustenance until supplies could be obtained
from the host colony at Sydney (Hoare, 1987).
The present analysis is based on bird fossils collected
during archaeological investigations of the Polynesian site
at Emily Bay. This was discovered in December 1995
(Anderson, 1996), and further, more extensive, excavations
took place in the same area in 1996 and 1997, resulting in
the retrieval of the large amount of mainly fragmentary avian
material that is discussed here. The analysis deals with
collections from Trenches EB96:10, EB96:11, EB97:21,
EB97:22, EB97:23, and EB97:24 in Emily Bay. Distribution
of material is presented for the largest collection, that from
Trench EB97:23. The composition of the archaeological
avifauna is reported and discussed both in the contexts of the
past and present avifauna of the island and of the resource that
the birds represented to the Polynesian population.
People are part of an island ecosystem but they also
interact with the ecosystem in different ways than other
species, especially on islands with no history of human
occupation. A point of major consequence for faunal
analysis of archaeological deposits is that people interact
with living birds and whole avifaunas, not with the bones
that remain as evidence of the interaction. Hence, this
preliminary analysis of the avifaunal remains associated
with the archaeological deposit at Emily Bay, Norfolk Island
goes some way beyond simple lists of bones and species.
The differences in composition between an archaeological
fauna and that of the total avifauna of the island can provide
information such as the fowling strategy of the people, their
food preferences (Worthy, 1998), and the time of year that
the site was occupied. To that end, some attention is paid
here to the breeding season of the species that are best
represented in the deposits, and of the likely season of
passage or residence of migrants.
Before the archaeological avifauna can be compared with
the original avifauna, the composition of that original fauna
must be known. For Norfolk Island, determining which
species bred there at the time of Polynesian contact is made
more difficult by extinctions that occurred both before and
after the first written records were made in the late eighteenth
century. To provide a basis for the inclusion or exclusion of
species from the avifaunal list, problems of identification of
fossil material and background information on the
identification of species known from early European records
are dealt with in some detail in the Appendix to this paper.
The Appendix also discusses the present distributions of
species, which can also provide clues as to the identification
of species. Breeding distributions that enclose but do not
include Norfolk Island suggest that some species might have
been part of the Norfolk Island fauna, whereas others whose
distributions are marginal might be less likely to have bred
there in the past.
Extinctions are important because they indicate the
possible effects of human impact on the Norfolk Island
environment, and also have a major bearing on the
identification of the species in fossil deposits. Extinctions
are part of the process of human interactions with the
environment throughout the history of the Pacific
(Steadman, 1997) and so the loss of species that people
used for food, and those that went extinct in the same period
but for which there is no archaeological record of human
exploitation, are also discussed.
Methods
The excavation and collection techniques are described in
Anderson (1996), and Anderson, Smith and White (this
vol.). Material was identified to species using morphological
characters if preserved, otherwise the material was taken to
the lowest taxonomic rank that could be supported (e.g.,
Pterodroma petrel, or “other petrel” which included petrels
such as P . neglecta and P . solandri which are larger than P.
pycrofti as well as Puffinus assimilis —see below), or left
as unidentified. Minimum numbers of individuals (MNI)
were determined for each taxon in each spit of each square,
using the maximum number of ipsilateral examples of the
commonest element. Crude estimates of biomass were
developed from the total numbers of individuals of each
species or taxonomic group, to give an order of magnitude

Holdaway and Anderson: Avifauna       87
assessment of the resource represented in each excavated
area. Note that this calculation gives the greatest MNI and
biomass; quantification by layer or trench would reduce
the numbers considerably (cf. Table 1; Fig. 2, above).
Comparative material . Identifications were made with
reference to voucher specimens in the Museum of New
Zealand Te Papa Tongarewa (MNZ: Wellington, New
Zealand) and the Canterbury Museum (CM: Christchurch,
New Zealand). There is little skeletal material for some of
the most significant species of petrel expected in the Norfolk
Island deposits. In particular, only three specimens of
Pterodroma solandri (MNZ S23504; and two unregistered
individuals in MNZ, and three of Pterodroma neglecta
(MNZ S23720, CM Av 5201 and Av 27263 were available.
The paucity of material contributed to the difficulty of
identification of petrel remains. A full study of natural
material from the island in conjunction with a satisfactory
comparative collection will be necessary before the present
group taxon “other petrels” can be re-analysed with
confidence. Meredith (1985, 1991) has provided a useful
baseline but present knowledge of the South Pacific petrel
fauna is insufficient to justify confidence in identification
of the medium-sized petrels of the group.
Material consulted included representatives of all
relevant species held in MNZ, and the following specimens
from CM: Puffinus bulleri  (Av36803), Puffinus gavia
(Av12158), Gallirallus philippensis (Av36805), Gallirallus
australis  (Av5187), Porphyrio porphyrio  (Av22392),
Coenocorypha huegeli  (Av5200), Limosa lapponica
(Av36583), Cyanoramphus novaezelandiae  (Av5163),
Cyanoramphus auriceps  (Av5194), Nestor meridionalis
(Av9956), Ninox novaeseelandiae  (Av22387), and
Eudynamys taitensis  (Av14854). The archaeological
material from Emily Bay will be accessed to the collections
of the Norfolk Island Museum.
Biomass . Body masses cited are from Heather and
Robertson (1996) or Marchant and Higgins (1990). For large
petrels, a mean body mass of 500 g was assumed, the normal
mass of the three largest species. As most of the animal
appears to have been consumed, including much of the bone
(and presumably the viscera), the total meat mass was taken
to be the normal body mass. The estimate is therefore high
by an unknown amount. The total amount of avian biomass
represented in the excavations was calculated and
extrapolated to the whole Emily Bay site.
“Natural” versus “archaeological” deposits . The
distinction is made between “natural” and “archaeological”
deposits, as a convenient shorthand to describe the
taphonomic processes involved in the development of the
deposit. While there is clearly a case for considering humans
to be a natural part of the environment, the distinction is
useful not least because non-archaeological (“natural”)
deposits can be formed by several different processes.
Not all of these are immediately obvious, and each
involves its own biases. “Natural” deposits are formed
regardless of human presence. For example, much of the
“natural” fossil material on Norfolk Island appears to have
been accumulated beneath feeding stations of the one or
more species of Accipiter  that once lived there (RNH,
unpubl. data). It is therefore biased by the time of hunting,
maximum and minimum prey size, agility of prey, and
probably other factors including preferences of individual
predators. Other bones were preserved as a result of
mortality unrelated to predation, such as burrow collapse
or pathology. In contrast to the variety of possible origins of
the “natural” fossil avifauna, the “archaeological” material was
assembled as a result of by a single process, the collection of
birds by people for food or other resources. That material yields
information on the choices of prey by people, and other factors,
such as seasonality, which are useful in reconstructing human
associations with the environment.
The term “petrel” is used for brevity in the broad sense
of all Procellariiformes and includes shearwaters, species
of Pterodroma, and other genera that may have been present
in the past. As well as the unknown range of species that
may be involved and the paucity of comparative material
for many, and the fragmentary nature of almost all the petrel
(and other) material from the archaeological deposit, the
catch-all headings serve the purpose of archaeological
analysis while not conveying a false sense of exactitude in
identification. Only P . pycrofti was determined to species,
on the basis of size for small fragments. The term “other
petrel” indicates all the other species of petrel that bred on
Norfolk Island, which were known to be present in the
archaeological fauna, but which could not be positively
identified regularly enough for quantitative analysis. Sufficient
elements of these other petrels could be identified to species
with greater or lesser confidence, to support their inclusion
in the list of species counted as present in the deposit.
Results
General. A total of 8699 bones and bone fragments was
identified to taxon (20 fragments of Sula dactylatra were
not included in the determination of MNI and are not
included in tabulated values); at least that number again
of unidentified fragments and small elements such as
vertebrae was also examined. For example, in Square
E12 in Trench EB97:23, there were 155 identifiable
elements, representing a minimum of 27 individuals, but
there were 145 elements such as quadrates, vertebrae, and
fragments that could not be certainly assigned to taxon
within the constraints of the present study. The bird bone
collection also contained material of the Pacific rat Rattus
exulans, and fish, which had been hidden among the bulk
of the fragmentary bird bones.
Preservation/taphonomy. The presence of large numbers
of small, fragile elements, and the occasional cranium or
other fragile elements demonstrated that the preservation
conditions were optimal for small bird bones. Most of the
damage to material that caused difficulties of identification
resulted from modification by human activities such as
heating, burning, breakage during dismemberment of the
carcass, and chewing. Immature bones were more affected
than adult bones. Some avian material, particularly in
Trench EB97:24, was clearly of natural origin and showed
no sign of human handling. Some of the “natural” material
was recorded as having been deposited in posthole features
or burrows and appears to post-date Polynesian occupation
of the site. Some elements showed signs of damage by an
avian predator (either the owl Ninox novaeseelandiae  or
goshawk Accipiter cf. A. fasciatus ), on the basis that the
damage was similar to that characteristic of butchering by
accipitrid hawks, as observed in material from natural sites

88       Records of the Australian Museum, Supplement 27 (2001)
in New Zealand whose accumulation has been attributed to
the extinct harrier Circus eylesi (RNH, unpubl. data).
Complete bones, particularly long bones, were rare. Most
individuals were represented by proximal humeri,
fragmentary coracoids, anterior sterna, furcular symphyses,
proximal scapulae, or distal tibiotarsi or tarsometatarsi. Most
MNI were based on the proximal humeri and coracoids for
Pterodroma pycrofti  and on coracoids and fragments of
coracoids for the other petrels. The waders were best
represented by coracoids ( Pluvialis fulva ) and distal
tarsometatarsi ( Limosa lapponica ). Of the two largest
common terrestrial species, the most common distinguishing
elements were the coracoid and distal tibiotarsus for the
large parrot Nestor productus and the coracoid for the pigeon
Hemiphaga spadicea . Two taxa that differed from the
pattern were Sula dactylatra and Accipiter cf. A. fasciatus:
the first was best represented by fragments and vertebrae,
and most hawk elements were pedal phalanges, fibulae, and
distal tibiotarsi. These remains made estimates of numbers
of individuals suspect as the smaller, more numerous bones
in the body may have been widely spread through the site.
Emily Bay bird sample . The number of identifiable
elements and minimum numbers of individuals (MNI) of
taxa represented in the archaeological collections from
Trenches EB96:10, EB96:11, EB97:21, EB97:22, EB97:23,
Table 1. Minimum number of individuals ( above) and number of identified bones ( below) for each taxon or group of taxa in each
excavation in Emily Bay. Totals in parentheses are MNI calculated on representation in all trenches combined (every element potentially
from same bird regardless of position in excavated area).
trench
EB96:10 EB96:11 EB97:21 EB97:22 EB97:23 EB97:24 totals
Pterodroma pycrofti 83331 0 8 2 5 150 (141)
79 19 9 17 [PHONE]
other petrels and shearwaters 47 6 1 9 663 89 815 (807)
356 52 2 42 5[PHONE]
Limosa lapponica 2 2 — — 12 8 24 (23)
13 8 — — 112 59 192
Pluvialis fulva 1211851 8  (13)
18234 1 4 7 1 0 2
Nestor productus 2 2 ——2 1 7  ( 3 )
3 3 ——9 6 2 1
Hemiphaga spadicea 5 2 — 1 28 12 48 (46)
19 4 — 2 136 40 201
Sula dactylatra 2 2 — 1 16 5 26 (25)
47— 28 2 1 8 1 1 3
Gallirallus Norfolk — — — 1 3 — 4 (3)
———1 6 —7
Cyanoramphus novaezelandiae — 11— 428  ( 5 )
— 31— 1 2 42 0
passerines 1 1 — — 2 2 6 (3)
1 1 — — 9 10 21
Gallirallus cf. G. philippensis 1 ———2 1 4  ( 3 )
2 ———6 1 9
Accipiter cf. A. fasciatus — — — 1 1 — 2 (1)
———1 8 —9
Gallicolumba cf. G. norfolciensis ————1 —1  ( 1 )
————2 —2
Porphyrio s p . ————2 —2  ( 2 )
————7 —7
total 69 21 6 17 [PHONE]
478 105 14 68 7009 1005 8679
and EB97:24 at Emily Bay, Norfolk Island are shown in
Table 1. The 14 taxa or group taxa (e.g., “other petrels”)
differed in their proportions in the samples (Table 2). “Other
petrels”, Pterodroma pycrofti and the Norfolk Island pigeon
(Hemiphaga spadicea) contributed most individuals (Table
1) and elements (Table 3). Their representation in the
smallest sample (Tre nch EB97:21) was skewed in
comparison to those in the larger samples and reflects
the heterogeneity of the distribution of material
throughout the site. The heterogeneity within a large part
of the area (Trench EB97:23) was marked, both in terms
of elements (Fig. 1) and individuals (Fig. 2). Most squares
in Trench EB97:23 had fewer than 100 elements (Fig. 3,
above) and fewer than 20 individuals (Fig. 3, below).
Squares D10, D13, and E10 in Trench EB97:23 contained
the greatest concentration of bones (Fig. 1, below).
The concentrations of material appear to coincide with the
location of post-holes and a fire scoop (Anderson, Smith and
White, this vol.). Six of the 42 squares contained most of the
elements, and three, most of the individuals. The heterogeneity
of the distribution within Trench EB97:23 shows that estimates
of concentration over the whole Emily Bay area must take
into account the clumping of dense deposits amid a much lower
density of material over most of the site. Squares D10, D13,
and E10 in Trench EB97:23 contained the greatest numbers
of identifiable individual birds (Fig. 1, above).

Holdaway and Anderson: Avifauna       89
Table 2. Percentage of individuals in each excavation represented by each taxon or group of taxa.
trench
EB96:10 EB96:11 EB97:21 EB97:22 EB97:23 EB97:24 total
Pterodroma pycrofti 11.6 14.3 50.0 17.6 12.7 16.7 13.5
other petrels 68.1 28.6 16.7 52.9 77.8 59.3 73.1
Limosa lapponica 2.9 9.5 0 0 1.4 5.3 2.2
Pluvialis fulva 1.4 9.5 16.7 5.9 0.9 3.3 1.6
Nestor productus 2.9 9.5 0 0 0.2 0.7 0.6
Hemiphaga spadicea 7.2 9.5 0 5.9 3.3 8.0 4.3
Sula dactylatra 7.2 9.5 0 5.9 1.9 3.3 2.6
Gallirallus Norfolk 0 0 0 5.9 0.4 0 0.3
Cyanoramphus novaezelandiae 0 4.8 16.7 0 0.5 1.3 0.7
passerines 1.4 4.8 0 0 0.2 1.3 0.5
Gallirallus cf. G. philippensis 1.4 0 0 0 0.2 0.7 0.4
Accipiter cf. A. fasciatus 0 0 0 5.9 0.5 0 0.2
Gallicolumba cf. G. norfolciensis 0 0 0 0 0.1 0 0.01
Porphyrio sp. 0 0 0 0 0.2 0 0.02
Table 3. Percentage of elements in each excavation represented by each taxon or group of taxa.
trench
EB96:10 EB96:11 EB97:21 EB97:22 EB97:23 EB97:24 total
Pterodroma pycrofti 16.5 18.1 64.3 25.0 14.0 27.5 15.9
other petrels 74.5 49.5 14.3 61.8 79.9 54.1 76.0
Limosa lapponica 2.7 7.6 0 0 1.6 5.9 2.2
Pluvialis fulva 0.2 7.6 14.3 4.4 0.6 4.7 1.2
Nestor productus 0.6 2.9 0 0 0.2 0.6 0.2
Hemiphaga spadicea 4.0 3.8 0 2.9 1.9 4.0 2.3
Sula dactylatra 1.9 6.7 0 2.9 1.3 2.2 1.3
Gallirallus Norfolk 0 0 0 1.5 0.1 0 0.1
Cyanoramphus novaezelandiae 0 2.9 7.1 0 0.2 0.4 0.2
passerines 0.2 1.0 0 0 0.1 1.0 0.2
Gallirallus cf. G. philippensis 0.4 0 0 0 0.1 0.1 0.1
Accipiter cf. A. fasciatus 0 0 0 1.5 0.1 0 0.1
Gallicolumba cf. G. norfolciensis 0000 <0.1 0 <0.1
Porphyrio s p .0000 0 . 10 0 . 1
Patterns of representation of taxa within the site .
Pterodroma pycrofti was found in 76 (87.4%) of the squares;
other species of petrel were found in 78 squares (89.7%).
Least common were the Norfolk Island Ground Dove
(Gallicolumba  cf. G. norfolciensis ) and Swamphen
(Porphyrio species), elements of which were found in three
squares only.
Squares contained up to 11 taxa, with the mode at seven
(Fig. 4). The bimodality of the distribution results from the
greater abundance of material in squares in Trench EB97:23,
and the greater chance of finding more species in a larger
sample; the lower tail of the distribution reflects the small
number of species in the poorer parts of the deposit.
Biomass. Calculations of body mass represented by the
individuals of the different taxa are given in Table 4,
percentages of total biomass by trench in Table 5. By far
the greatest contribution to biomass was by the petrels, over
80% when Pterodroma pycrofti  are pooled with “other
petrels”. The Booby Sula dactylatra (Sulidae) contributed
significantly to the total biomass because of its greater
individual mass, although it was difficult to assess the total
numbers concerned because many examples were juvenile
and hence more poorly preserved.
All other taxa combined contributed less than 10% of
the biomass represented in the site. Of the terrestrial species,
only the Norfolk Island Pigeon Hemiphaga spadicea  was
important in the diet. Apparently vulnerable and palatable
species such as the flightless Norfolk Island Rail and the
Norfolk Island Kaka contributed less than the more common
of the two migrant wading birds ( Limosa lapponica). The
smaller species contributed negligible amounts to the total
biomass and were not favoured prey. The rarity or absence
of some palatable species (such as snipe and rails) and of
other species that might have been used for ornament, if
not for food, is noteworthy. In view of the common presence
of strongly flying species such as the two migrant waders,
the absence of snipe, for example, begs the question of
availability during the period of occupation.
The absence of terns from the sample might be explained
by the very small meat content of the species present, and
the fact that all but the Sooty Tern ( Sterna fuscata) nest in
trees. Sooty Terns breed on beaches or open ground, and
lack of open areas in Polynesian times except for the beaches
in the Kingston area (which were subject to disturbance by
people), suggests that Sooty Terns might not have bred in

90       Records of the Australian Museum, Supplement 27 (2001)
Figure 1 . Plan of excavated squares in Trench EB97:23, Emily Bay, Norfolk Island, showing number of bird
elements identifiable to taxon or taxon group, all spits in each square pooled ( above). Isometric representation of
number of identifiable bird elements in Trench EB97:23, Emily Bay, Norfolk Island, showing concentration of
material in a few squares centred on D10–13 to F10–13 ( below).
numbers on the main island in the past. In any event, larger
species that were easier to catch were abundant in and near
the occupation site. The archaeological sample is a subset
of the breeding bird fauna: many species, particularly terns
and tropicbirds, are not represented at all (Table 6).
Seasonality. Some impression of the time of year that the site
was inhabited can be obtained from the composition of the
faunal remains, most of which represent seabirds whose
abundance on the island fluctuates greatly with the seasons.
Many, including the most important taxa, were absent for at
least half the year while on non-breeding migration to the
Northern Hemisphere or elsewhere in the Pacific. Different
taxa can be defined as summer- or winter-breeding. Two of
the terrestrial species were also migratory, being present in the
southern summer. The present status and usual breeding season
of Norfolk Island birds are given in Table 6.
Systematic list of species represented in the
Emily Bay settlement site
The following descriptions provide the mean individual
body weight of live birds, the proportional distribution of
remains in the excavations and pertinent zoological and
behavioural information.

Holdaway and Anderson: Avifauna       91
Figure 2. Plan of excavated squares in Trench EB97:23, Emily Bay, Norfolk Island, showing the minimum number
of individuals represented by identifiable bones, all spits in each square pooled (above). Isometric representation of
minimum number of individuals represented by identifiable bones in Trench EB97:23, Emily Bay, Norfolk Island,
all spits in each square pooled. Individuals were also concentrated in one section of the excavation ( below).
PROCELLARIIDAE— PETRELS AND SHEARW ATERS
(89.7% OF SQUARES)
Pterodroma pycrofti  Pycroft’s Petrel 160 g (87.4% of
squares). The smallest petrel breeding in significant numbers
on Norfolk Island. The history of its discovery is discussed
in the Appendix.
Pterodroma solandri  Providence Petrel or Solander’s
Petrel 500 g. These and the other petrels cannot be assigned
a representation by square, because their records were
necessarily pooled (see Methods and Appendix). At the time
of European settlement, P . solandri seems to have been
largely, if not entirely, confined to the forest on Mt Pitt and
Mt Bates. Petrel bones of the size range of P . solandri are
abundant in the archaeological collections from Emily Bay.
Either the species had a wider geographical range on the
island before Europeans arrived, or the Polynesians
collected birds from farther afield on that island than the
environs of the Kingston flat. Pterodroma petrels can be
attracted to the ground from flight by making various loud
sounds (Tennyson and Taylor, 1990) so it is impossible to
tell from the presence of P . solandri in the Emily Bay site,
just where the birds were nesting and being hunted in
Polynesian times.
Pterodroma neglecta  Kermadec Petrel 500 g. The
Kermadec Petrel has not been recognized from Norfolk
Island fossil deposits before, although it has now been
recorded breeding on Philip Island (Moore, 1999). A

92       Records of the Australian Museum, Supplement 27 (2001)
Figure 3. Pattern of distribution of identifiable elements per square in Trench EB97:23, Emily Bay, Norfolk Island
(above). Pattern of distribution of individuals represented of identifiable elements per square in Trench EB97:23,
Emily Bay, Norfolk Island (below).
summary of the reasons for including P . neglecta in the
avifauna of Norfolk Island is given in the Appendix. The
absence of reports of ground-nesting petrels in the notes
and diaries of the first European settlers may result from
the rarity of the species as a result of predation by Pacific
rats, or because P . neglecta on Norfolk Island used different
nest sites to those in other populations. Kermadec Petrels
nest on the surface at the present colonies (Heather and
Robertson, 1996), but at none of these sites are the birds
hunted by diurnal birds of prey. On Norfolk Island at least
one large raptor was capable of killing P . neglecta; a petrel
population there would have had many thousands of years
to adopt a burrowing habit.
Puffinus pacificus Wedge-tailed Shearwater 450 g. This
is the common summer-breeding large petrel surviving on
Norfolk Island. It still attempts to breed on headlands on
the main island, but cats take many birds, and the population
is probably declining. Although they could not be quantified,
both adult and juvenile bones of P . pacificus were obviously
abundant in the remains, supporting the view established

Holdaway and Anderson: Avifauna       93
Table 4. Biomass (kg) represented by minimum number of individuals in each excavation at Emily Bay, based on normal body masses
given in text.
trench
EB96:10 EB96:11 EB97:21 EB97:22 EB97:23 EB97:24 total
Pterodroma pycrofti 1.28 0.48 0.48 0.48 17.28 4.0 24.0
other petrels 21.15 2.70 0.45 4.05 298.35 40.05 36.75
Limosa lapponica 0.6 0.6 0 0 3.6 2.4 7.2
Pluvialis fulva 0.13 0.26 0.13 0.13 1.04 0.65 2.34
Nestor productus 1.0 1.0 0 0 1.0 0.5 3.5
Hemiphaga spadicea 2.5 1.0 0 0.5 14.0 6.0 24.0
Sula dactylatra 3.4 3.4 0 1.7 27.2 8.5 44.2
Gallirallus Norfolk 0 0 0 0.25 0.75 0 1.0
Cyanoramphus novaezelandiae 0 0.075 0.075 0 0.3 0.15 0.6
passerines 0.09 0.09 0 0 0.18 0.18 0.54
Gallirallus cf. G. philippensis 0.17 0 0 0 0.34 0.17 0.68
Accipiter cf. A. fasciatus 0 0 0 0.5 0.5 0 1.0
Gallicolumba cf. G. norfolciensis 0000 0 . 20 0 . 2
Porphyrio s p .0000 1 . 60 1 . 6
total 30.32 9.61 1.14 7.61 366.34 62.6 477.61
biomass m-2 2.76 3.2 0.19 7.61 8.76 2.72 5.55
Table 5. Percentage of biomass contributed by each taxon or group of taxa in the excavations at Emily Bay.
trench
EB96:10 EB96:11 EB97:21 EB97:22 EB97:23 EB97:24 total
Pterodroma pycrofti 4.22 5.00 42.29 6.31 4.72 6.39 5.03
other petrels 69.76 28.11 39.65 53.22 81.44 63.98 76.79
Limosa lapponica 1.98 6.25 0 0 0.98 3.83 1.51
Pluvialis fulva 0.43 2.71 11.45 1.71 0.28 1.04 0.49
Nestor productus 3.30 10.41 0 0 0.27 0.80 0.73
Hemiphaga spadicea 8.25 10.41 0 6.57 3.82 9.58 5.03
Sula dactylatra 11.21 35.40 0 22.34 7.42 13.58 9.25
Gallirallus Norfolk 0 0 0 3.29 0.20 0 0.21
Cyanoramphus novaezelandiae 0 0.78 6.61 0 0.08 0.24 0.13
passerines 0.30 0.94 0 0 0.05 0.29 0.11
Gallirallus cf. G. philippensis 0.56 0 0 0 0.09 0.27 0.14
Accipiter cf. A. fasciatus 0 0 0 6.57 0.14 0 0.21
Gallicolumba cf. G. norfolciensis 0000 0.05 0 0.04
Porphyrio s p .0000 0.44 0 0.34
by the presence of migratory wading birds that part of the
deposit was laid down in summer. Puffinus pacificus  is
migratory in the opposite sense to the Bar-tailed Godwit and
Golden Plover, breeding on the islands and migrating to the
North Pacific in the southern winter. They are large and
aggressive petrels and are not known to be attracted to strange
noises so must have been collected at the breeding colonies.
Puffinus assimilis Norfolk Island Little Shearwater 200
g. Holdaway et al. (2001) recognize the Norfolk Island form
of little shearwater as a separate species from others at the
Kermadecs, northern New Zealand, and the New Zealand
subantarctic. On this view, P . assimilis  is a rare and
endangered species, as most breeding attempts on the main
island are thwarted by cats and rats. Only the population
breeding on Philip Island has both the space and freedom
from predation to be sure of survival in the medium to long
term. Although this species is not abundant in the
archaeological samples, P . assimilis was certainly part of
the diet of the Emily Bay people. Both P . assimilis and the
similar-sized P . auricularis newelli may have bred on the
island formerly (Appendix). Their size makes both species
vulnerable to predation by Pacific rats (Booth et al., 1996)
and hence they may have suffered more from rat predation
than from human exploitation when there were larger
species to concentrate on.
S
ULIDAE— GANNETS AND BOOBIES
Sula dactylatra Masked Booby 1700 g (73.6% of squares).
Boobies were the largest terrestrial prey available on the
Norfolk Island group. It is not surprising that they were
relatively common in the deposit (26 individuals; 2.6%
of total birds; 9.25% total body mass). At about four times
the mass of a large petrel, the boobies would have been
attractive and easy prey. At first contact, it is likely that S.
dactylatra nested on open, flat areas such as the tops of
stacks (to which they are largely confined today by human
persecution) and on the sandy beaches so would have been
extremely vulnerable. It is unlikely that beach colonies of
S. dactylatra could have survived the first year of human
occupation of Norfolk Island. Masked Boobies elsewhere

94       Records of the Australian Museum, Supplement 27 (2001)
Figure 4. Distribution of taxa in squares in Trenches EB96:10, EB96:11, EB97:21, EB97:22, EB97:23, and EB97:24,
Emily Bay, Norfolk Island.
nest mainly on low islands and amid dunes (Serventy et al.,
1971). A larger species described from the island is not
accepted here (Appendix).
P
HAETHONTIDAE— TROPICBIRDS
Only Phaethon rubricauda , the Red-tailed Tropicbird at
900–1,000 g approached the mass of a Masked Booby. At
twice the mass of a large petrel, and in relative abundance
around the islands, it is surprising that Red-tailed
Tropicbirds do not figure in the food remains. Their tails
are highly prized elsewhere in Polynesia (Steadman, 1997)
and it might have been expected that some birds would have
been taken for that purpose, but if so their remains were
not found at the Emily Bay site.
A
CCIPITRIDAE— HAWKS AND EAGLES
Accipiter cf. A. fasciatus Brown Goshawk 350–500 g (9.2%
of squares). Remains of the goshawk were rare in the site,
but most had clearly been cooked and eaten. Remains of
harriers ( Circus  species) are occasionally found in
Polynesian sites in New Zealand and appear to have been
eaten there as well.
R
ALLIDAE— RAILS
Gallirallus philippensis  Banded Rail 170 g (9.2% of
squares). The status of the Banded Rail on Norfolk Island
has been the subject of debate (e.g., Schodde et al., 1983),
but fossil remains reported by Meredith (1985, 1991) and
in this paper show that it was present in pre-European times
in sufficient numbers for it to be available as a resource for
the Polynesian inhabitants. It was therefore sympatric with
the undescribed endemic flightless species of Gallirallus
and was possibly a relatively recent arrival.
Gallirallus  new species ?250 g (5.7% of squares).
Meredith (1985, 1991) reported the presence of a rail larger
than Gallirallus philippensis  in fossil collections from
Norfolk Island, but did not describe or name it. He referred
it to Gallirallus and suggested that it was flightless. A rail
painted during the first European settlement of Norfolk
Island (no. 79 in the “Sydney” series, Hindwood, 1965)
has been identified as G. philippensis but could equally be
this species.
Porphyrio species Swamphen 800 g (3.4% of squares).
Races of Porphyrio porphyrio  and other species of
Porphyrio  are widespread in the Pacific (Ripley, 1977;
Steadman, 1988). Ramsay (1888) was the first to record
Swamphens ( Porphyrio porphyrio ) on Norfolk Island in
historic times. By 1978 the species was still regarded as an
“uncommon, restricted self-introduced resident” (Schodde
et al., 1983). Meredith (1985, 1991) did not record fossil
material of Porphyrio. A few bones of a small Porphyrio
found in Trench EB97:23 constitute the only evidence for
the former presence of Swamphens on the island.
Unfortunately, there is insufficient material for a statistical
analysis, but comparison with an individual of the New
Zealand population suggests that the Norfolk Island
Porphyrio were smaller than New Zealand birds and perhaps
more similar to the small races of the islands to the north
than to the Australian and New Zealand forms. The presence
of more than one individual of Porphyrio at the Emily Bay
site indicates the presence of a population and hence that
suitable habitat for the species was present during the
Polynesian occupation.
Norfolk Island had at least four species of rail, a high
diversity of rails for its area. Besides the three discussed
here, the Spotless Crake ( Porzana tabuensis ) was also
present (Meredith, 1991, and Appendix).
S
COLOPACIDAE— SNIPE AND GODWIT
Limosa lapponica  Bar-tailed Godwit 300 g (65.5% of
squares). Limosa lapponica  migrates each year from the
breeding area on the Siberian tundra to wintering grounds
that include the estuaries of New Zealand. They occur on
Norfolk Island from September to March (Schodde et al.,
1983). The number of individuals in the Emily Bay deposit

Holdaway and Anderson: Avifauna       95
Table 6. The late Holocene avifauna of Norfolk Island, with representative body mass (g), present status at Norfolk Island, representation
in the Emily Bay archaeological site, and breeding season. EB, present in the Emily Bay fauna; E, now extinct as a breeding species.
mass status EB season
Procellariidae (petrels and shearwaters)
Pterodroma neglecta Kermadec Petrel 500 recolonizing Y all year
Pterodroma solandri Providence Petrel 500 recolonizing Y Winter
Pterodroma pycrofti Pycroft’s Petrel 160 extinct Y Summer
?Pterodroma nigripennis Black-winged Petrel 175 ?recolonizing Summer
Puffinus pacificus Wedge-tailed Shearwater 450 breeding Y Summer
Puffinus assimilis Norfolk Island Little Shearwater 200 breeding Winter
Hydrobatidae (storm petrels)
Pelagodroma albiclunis Kermadec Storm Petrel 45 extinct Summer
?Fregetta grallaria White-bellied Storm Petrel 50 ?extinct late Summer
Sulidae (gannets and boobies)
Sula dactylatra Masked Booby 1700 breeding Y all year
Phaethontidae (tropicbirds)
Phaethon rubricauda Red-tailed Tropic Bird c. 900 breeding Summer
Accipitridae (hawks and eagles)
Accipiter cf. A. fasciatus ?Brown Goshawk 500 extinct Y ?
Rallidae (rails)
Gallirallus undescribed sp. Norfolk Island Rail 250? extinct Y ?
Gallirallus philippensis Banded Rail 170 vagrant Y Summer
Porzana tabuensis Spotless Crake 45 E, vagrant Summer
Porphyrio sp. Swamphen 800 extinct Y Spring–Summer
Scolopacidae (snipe and godwits)
Coenocorypha undescribed sp. Norfolk Island Snipe 100 extinct Summer
Limosa lapponica Bar-tailed Godwit 300 migrant Y Summer
Numenius phaeopus Whimbrel 450 migrant Summer
Charadriidae (dotterels and plovers)
Pluvialis fulva Pacific Golden Plover 130 migrant Y Summer
Laridae (gulls and terns)
Sterna fuscata Sooty Tern 210 breeding Spring–Summer
Anous stolidus Common Noddy 200 breeding Spring–Summer
Anous minutus Black Noddy 100 breeding Spring
Procelsterna cerulea Grey Ternlet 75 breeding Spring
Gygis alba White Tern 11 0 breeding Spring
Columbidae (pigeons and doves)
Hemiphaga spadicea Norfolk Island Pigeon 650 extinct Y all year?
Gallicolumba cf. G. norfolciensis Norfolk Island Ground Dove 200? extinct Y ?
Psittacidae (parrots and parakeets)
Cyanoramphus novaezelandiae cookii Norfolk Island Green Parrot 75 endangered Y Spring–Summer
Nestor productus Norfolk Island Kaka 400 extinct Y Summer
Cuculidae (cuckoos)
Eudynamys taitensis Long-tailed Cuckoo 125 migrant Summer
Strigidae (typical owls)
Ninox novaeseelandiae Southern Boobook 175 endangered Summer
Alcedinidae (kingfishers)
Halcyon sancta Sacred Kingfisher 65 breeding Summer
Songbirds
Campephagidae (trillers)
Lalage leucopyga Long-tailed Triller 50? extinct ?
Muscicapidae (northern flycatchers)
Turdus poliocephalus Grey-headed Blackbird 90? extinct Spring?
Pachycephalidae (thickheads)
Pachycephala pectoralis Golden Whistler 50? breeding Spring?
Acanthizidae (flyeaters)
Gerygone modesta Norfolk Island Gerygone 6.5 breeding Spring?
Monarchidae (monarch flycatchers)
Rhipidura fuliginosa Grey Fantail 8 breeding Spring?
Petroicidae (southern robins)
Petroica multicolor Pacific Robin 11 breeding Spring?
Zosteropidae (silvereyes)
Zosterops tenuirostris Slender-billed White-eye 15 breeding Spring?
Zosterops albogularis White-chested White-eye 16 breeding Spring?
Sturnidae (starlings)
Aplonis fusca Norfolk Island Starling 80? extinct Spring?

96       Records of the Australian Museum, Supplement 27 (2001)
suggests that L. lapponica was taken from migrating flocks,
as relatively few birds would have been resident on the
island when it was still mainly forested.
C
HARADRIIDAE— PLOVERS
Pluvialis fulva  Pacific Golden Plover 130 g (48.3% of
squares). Less common than the godwit in the samples,
Pluvialis fulva  was nevertheless more abundant than
resident land birds such as the parrot and parakeet. Because
this species is a strong-flying migrant, the degree of
representation is rather anomalous, even though the birds
are much tamer on smaller islands than they are on the New
Zealand mainland (Heather and Robertson, 1996) and might
therefore have been easier to catch. The identification is
based on the likelihood of occurrence of the two species of
golden plover in the Pacific. The species is listed in older
literature as Pluvialis dominica, but that is now regarded as
a separate species with allopatric non-breeding distributions
(Connors, 1983).
C
OLUMBIDAE
Hemiphaga spadicea Norfolk Island Pigeon 650 g (73.6%
of squares). The Norfolk Island Pigeon was very similar to
Hemiphaga novaeseelandiae , the New Zealand Pigeon, of
which it has been regarded as a subspecies by some authors
(Schodde et al., 1983). The New Zealand bird was a regular
item in the diet of Polynesians there (Oliver, 1955).
Hemiphaga spadicea  survived on Norfolk Island into the
early nineteenth century. It is the most abundant of the
terrestrial birds in the archaeological avifauna, in which
young birds as well as adults were represented.
Gallicolumba cf. G. norfolciensis Norfolk Island Ground
Dove 200 g (3.4% of squares). The status of the small doves
on Norfolk Island has yet to be resolved. Whether
Chalcophaps indica was resident before habitat destruction
became important in the European era is not known.
Although fossils have been found (Meredith, 1991), none
has been dated. The presence of a species of Gallicolumba
was confirmed by Meredith (1985, 1991). It is probably
this species which was mentioned in the diaries of the first
European settlers and is the subject of a painting done before
1800 (no. 89, “Sydney” series, Hindwood, 1965). Bones of
a dove-sized pigeon, apparently this species, were rare in
the archaeological fauna. The fragmentary material did not
allow complete certainty of the identification.
P
SITTACIDAE
Nestor productus Norfolk Island, or Long-billed, Kaka 400
g (18.4% of squares). Although Nestor productus  was
apparently common when Europeans reached the island and
therefore was likely to have been a prominent component
of the avifauna exploited by Polynesians, surprisingly few
were represented in the archaeological deposits. The New
Zealand species N. meridionalis was a favoured item in the
diet of Maori from settlement to historic times, and its
feathers were used in cloak manufacture (Oliver, 1955): N.
productus was even more colourful so Polynesians may have
caught it for its feathers as well as the meat.
Cyanoramphus novaezelandiae cooki  Norfolk Island
Green Parrot; Norfolk Island Red-crowned Parakeet 75 g
(21.8% of squares). Green Parrots are rare in the
archaeological record and constituted only a minor and
opportunistic food source.
P
ASSERINES (20.7% OF SQUARES)
The Grey-headed Blackbird Turdus poliocephalus (?90 g)
and other species such as the Norfolk Island Starling Aplonis
fuscus represented a tiny percentage of the total biomass in
the deposits. Their presence is important mainly as an
indication of the eclectic diet of the inhabitants, which again
parallels that observed in New Zealand sites where even
Petroicas (New Zealand Robins) were consumed.
Discussion
The faunal remains in the archaeological site at Emily Bay
are confined mainly to birds, fish, and introduced Pacific
rats. The dominance of seabirds is understandable, because
the species were very abundant, easy to catch, and mostly
large enough to constitute worthwhile additions to the diet.
It appears that the larger petrels were, indeed, a staple food
for the early Polynesian inhabitants of the island, as one
species was for Europeans for a short period in the early
part of their occupation. Other marine species, and most of
the land birds, appear to have been included in the diet only
as adjuncts, probably when they could be caught with little
or no effort. As most of the Pterodroma petrels can be called
from the sky during the breeding season, food gathering on
Norfolk Island—while the petrel populations survived—
would have been remarkably easy; but the diet might have
been rather monotonous.
Taphonomy, butchering, and consumption . Patterns of
survival of elements confirmed that most of the deposit
accumulated as debris from human occupation. The pattern
of bone breakage and survival was typical of a large predator
that could process and remove as much of the available
nutrients from each carcass as possible. For the major long
bones, such as tibiotarsus, humerus, and ulna, only one end
(often the proximal) is well represented, indicating that the
sections holding the most meat have been removed, and
probably consumed along with the meat. Interestingly, the
pattern of archaeological bone survival is quite similar to
that found in deposits accumulated by the large extinct
harrier Circus eylesi in New Zealand (RNH, unpubl. data)
in which the larger long bones and most bones from the
body are not present, and peripheral, even tiny, elements
are well-represented in the sample.
The taphonomic processes contributed to the difficulty
of assigning most petrel material to species, apart from the
very small P . pycrofti, exacerbating the already difficult task
of separating closely similar species with inadequate
reference material. However, the lumping of petrel taxa into
two groups should not have altered the main conclusion of
the study, that petrels were the main item of diet of the people
who lived at the site.
Limited avifauna as part of a limited natural food supply.
Polynesians living on Norfolk Island had a much smaller
choice of natural foods than they did on other island groups
elsewhere in the Pacific. In particular, most of the coastline
consists of steep slopes above a rocky shore or vertical cliffs.
There are few beaches (Anson Bay, Cemetery Bay, Emily
Bay, Slaughter Bay), and the littoral zone has very few
species of (small) mollusc and echinoderms. The small area
of reef at Slaughter Bay meant that only fishing in deep
water was likely to be productive enough to support a human
population. That in turn meant that the birds would have

Holdaway and Anderson: Avifauna       97
been proportionately more important than on most other
islands. Their importance is reflected in their relative
abundance in the food remains at Emily Bay. Although the
avifauna was limited in variety, some species, including two
or three petrels, were extremely abundant. Others, including
all the terns that are such a feature of the present avifauna,
were, for whatever reason, virtually ignored for food.
Choice versus availability of food. The biases in the avian
remains from the archaeological excavations could indicate
either that the people actively selected a limited range of
species from those available on the island, or took by default
those species that were most abundant and easy to catch. In
general, most species in the deposits could have been collected
within 500 m of the site. The two most abundant species were
both Pterodroma petrels that could be harvested from the
surface or from burrows in the neighbourhood of Emily Bay,
and which could also have been called down from the sky by
shouting or hand-clapping (Tennyson and Taylor, 1990).
Meredith (1985) reported that both P . pycrofti and his
Pterodroma new species were rare in the First Settlement
deposits that he examined. As the former, which is the
smaller of the two, is well within the size range for predation
on adults, eggs, and young by Pacific rats (Holdaway, 1999),
its rarity by the late eighteenth century is not surprising.
The apparently low abundance of the larger, Pterodroma
new species (probably P . neglecta) at European contact is
more difficult to explain. Pterodroma neglecta survived in
the presence of Pacific rats on Raoul Island from about 650
years ago (Anderson, 1980) into the early twentieth century,
so apparently can cope with some predation even though
its egg is just within the ability of Pacific rats to open
(Holdaway, 1999). High numbers of rats maintained by the
year-round availability of animal and vegetable food might
have created conditions that allowed rat predation to be more
severe than it might have been otherwise. Another possible
factor in the rarity of P . neglecta in the First Settlement
deposits was the degree to which it might have been taken
as a preferred food by the Polynesians. Unfortunately,
identification problems for the petrels made quantification
of their relative representation impossible.
Seasonality . It is apparent that food was harvested
throughout the year at Emily Bay: P . pycrofti, a summer
breeder (Heather and Robertson, 1996), is a major part of
the sample, as is P . solandri, which breeds in the southern
winter. The young of P . pycrofti would be available up to
the time of their fledging in late summer and autumn. Again,
although P . neglecta has been recorded nesting at most times
of the year in various parts of its huge breeding range, birds
in the remaining colonies usually lay their eggs from
October to March, so that species and Puffinus pacificus
would have supplemented the supply of P . pycrofti, which
as a smaller bird would provide less meat per animal.
A greater variety of birds would be available in summer,
with the presence of the migrant waders and several species
of petrel. In winter, the people would have had to depend
on P . solandri and Puffinus assimilis . That dependence, at
a time of greater frequency of storms and hence lower
availability of fish, may have limited the human population
that could be supported on the island over a period of years.
The population of P . solandri on the higher areas of the
island could not sustain predation by the few hundred people
of the First Settlement for more than a few years.
Species representation and extinction . There were
substantially fewer species represented in the Emily Bay
archaeological site than were available in the local
environment. To some extent, the comparison is unequal,
because the natural bone deposits from which the
composition of the late Holocene avifauna of Norfolk Island
has been established (Meredith, 1985, 1991) accumulated
over thousands of years, whereas the archaeological site
may have existed for less than 200 years and, in addition,
only about 3.5% of it was excavated. While it is unlikely
that the avifauna found by the first Polynesian settlers
differed greatly from that reported by the earliest European
inhabitants—plus the known extinct species—it is difficult
to tell whether the archaeological absence of the more than
half of the late Holocene avifauna (excluding songbirds) can
be explained by sampling biases during the prehistoric fowling
or during archaeological recovery, or by other factors.
The rarity of terrestrial species in the deposit is
noteworthy. Contrary to Meredith (1985), it is likely that
Rattus exulans was responsible for the extinction of several
of the smaller, terrestrial birds on Norfolk Island. No other
environmental factors are known which could have affected
small species, and terrestrial as well as oceanic species. In
fact, the extinctions of small vertebrates on Norfolk Island
parallels the far more extensive extinctions attributed to the
Pacific rat in New Zealand (Holdaway, 1999). Species
lacking from the archaeol ogical collection but known from
natural fossil deposits at Cemetery Bay and elsewhere include
a prion ( Pachyptila species), a storm petrel ( Pelagodroma
species, presumably P . albiclunis, the Kermadec Storm petrel),
and a southern snipe (Coenocorypha new species), all of which
were palatable to both humans and rats.
The absence of Norfolk Island Snipe (Coenocorypha new
species) from the archaeological deposits is particularly
noticeable as there are remains of two shorebirds (Pluvialis
fulva, Limosa lapponica ) that should have been harder to
catch. Coenocorypha new species and P . fulva had roughly
the same body mass (105 g vs 130 g). The former is unlikely
to have been missed in recovery of material, because
passerine bones were recovered and many elements smaller
than snipe bones were common in the collection. Given the
size of the sample, it is unlikely that Coenocorypha new
species was present when the archaeological deposits were
formed. Why this should be, when other wading birds were
eaten regularly, is unknown, although the Pacific rat and
snipe of the genus Coenocorypha  have been unable to
coexist elsewhere (Holdaway, 1999). If the rat population
rose to and was sustained at a high level by abundant petrels,
as it would have been on Norfolk Island, then extinctions
could have occurred extremely rapidly, as occurred when
Rattus rattus  reached the Big South Cape Islands off
southern New Zealand in the early 1960s (Bell, 1978).
Various scenarios are possible in relation to the absence
of snipe and other taxa in the archaeological assemblage.
The lack of terns, which are the most obvious seabirds on
Norfolk Island today, may be related to the cost and benefits
of harvesting. Some species, such as the Sooty Tern, may
have been more limited in numbers on the island in the past.
Tropicbirds are used by Pacific peoples for ornamentation as
well as for food, so the absence of this presently common
species from the archaeological avifauna is rather surprising.
It is unfortunate that the archaeology of Norfolk Island does
not, so far, offer later prehistoric settlement sites with bird
bone middens in which to test some of the propositions

98       Records of the Australian Museum, Supplement 27 (2001)
suggested here. The only comparisons possible at present are
with the records made by the first European settlers (Meredith,
1985) and the material in the deposits of that date.
Conclusions
In terms of body mass, and therefore available meat value,
the archaeological avifauna from Emily Bay discloses a
strong predominance of petrels and boobies, as might be
predicted from the relative body mass and probable
abundance of these taxa in the local environment. Larger
forest birds are relatively scarce, especially the rails, Norfolk
Island Kaka, and Norfolk Island Ground Dove. Some small
terrestrial taxa are absent and either were not sought or may
have become extinct so rapidly (probably as a result of
predation by the Pacific rat, Rattus exulans) that they were
not incorporated in the Emily Bay deposits. The archae-
ological fauna of Norfolk Island includes species such as
Pycroft’s Petrel ( Pterodroma pycrofti ) that were either
locally extinct, or at least very rare, when Europeans reached
the island. It is apparent, therefore, that Norfolk Island fits
the pattern of other Pacific Islands, where early contact by
Polynesian settlers resulted in the extinction of the more
vulnerable of the resident bird species (Steadman, 1997).
ACKNOWLEDGMENTS . RNH thanks the New Zealand Foundation
for Research, Science & Technology for support for this work,
undertaken as part of Contract PLC501. Thanks are also due to
Alan Tennyson and Geoff Tunnicliffe for access to the collections
in their care at the Museum of New Zealand Te Papa Tongarewa
and Canterbury Museum, respectively. Alan Tennyson and Trevor
Worthy provided most helpful advice and discussion on the sorrows
and joys of the identification of petrel bones and the species that
may have bred on the island.
References
Anderson, A.J., 1980. The archaeology of Raoul Island (Kermadecs)
and its place in the settlement history of Polynesia. Archaeology
and Physical Anthropology in Oceania 15: 131–141.
Anderson, A.J., 1996. Discovery of a prehistoric habitation site on
Norfolk Island. Journal of the Polynesian Society 105: 479–486.
Anderson, A.J., in press. No meat on that beautiful shore: the
prehistoric abandonment of subtropical Polynesian islands. In
Zooarchaeology of Oceanic Coasts and Islands , ed. A.J.
Anderson and B.F. Leach. International Journal of Osteo-
archaeology.
Bartle, J.A., D. Hu, J-C. Stahl, P. Pyle, T.R. Simons and D. Woodby,
1993. Status and ecology of gadfly petrels in the temperate North
Pacific. In The Status, Ecology, and Conservation of Marine Birds
of the North Pacific, ed. K. Vermeer, K.T. Briggs, K.H. Morgan
and D. Siegel-Causey, pp. 101–111. Ottawa: Canadian Wildlife
Service special publication.
Bell, B.D., 1978. The Big South Cape Islands rat irruption. In
The Ecology and Control of Rodents in New Zealand Nature
Reserves, ed. P.R. Dingwall, I.A.E. Atkinson and C. Hay, pp.
33–40. Wellington: New Zealand Department of Lands and
Survey Information Series no. 4.
Booth, A.M., E.O. Minot, R.A. Fordham and J.G. Innes, 1996. Kiore
(Rattus exulans) predation on the eggs of the little shearwater
(Puffinus assimilis haurakiensis). Notornis 43: 147–153.
Brooke, M. de L., and G. Rowe, 1996. Behavioural and molecular
evidence for specific status of pale and dark morphs of the
Herald petrel Pterodroma heraldica. Ibis 138: 420–432.
Connors, P.G., 1983. Taxonomy, distribution, and evolution of
golden plovers ( Pluvialis dominica  and Pluvialis fulva ). Auk
100: 607–620.
Gurney, J.H., Sr., 1854. Note on Nestor productus , the extinct
parrot of Philip Island. The Zoologist 1854 : 4298.
Harrison, P ., 1983. Seabirds: An Identification Guide. Wellington:
Croom Helm and A.H. and A.W. Reed.
Hawkins, J., 1988. The Farewell Spit gannetry—A new sea level
colony. Notornis 35: 249–260.
Heather, B.D., and H.A. Robertson, 1996. Field Guide to the Birds
of New Zealand. Auckland: Viking.
Hermes, N., O. Evans and B. Evans, 1986. Norfolk Island birds: a
review 1985. Notornis 33: 141–149.
Hindwood, K.A., 1940. The birds of Lord Howe Island. Emu
50: 1–86.
Hindwood, K.A., 1965. The “Sydney” bird paintings. The
Australian Zoologist  13: 83–92.
Hindwood, K.A., and D.L. Serventy, 1943. Further notes on
Pterodroma leucoptera. Emu 42: 153–155.
Hoare, M., 1987. Norfolk Island: An Outline of its History 1774–
1987. St Lucia: University of Queensland Press.
Holdaway, R.N., 1999. Introduced predators and avifaunal
extinction in New Zealand. In Extinctions in Near Time:
Causes, Contexts, and Consequences , ed. R.D.E. MacPhee,
pp. 189–238. New York: Kluwer Academic/Plenum Press.
Holdaway, R.N., T.H. Worthy and A.J.D. Tennyson, 2001. A working
list of breeding bird species of the New Zealand region at first
human contact. New Zealand Journal of Zoology 28: 119–187.
Kirch, P.V ., 1988. Polynesia’s mystery islands. Archaeology May/
June: 26–31.
Marchant, S., and P .J. Higgins, co-ordinators, 1990. Handbook of
Australian, New Zealand & Antarctic Birds, vol. 1—Ratites to
Ducks. Melbourne: Oxford University Press.
Mathews, G.M., 1912. The geographic relationships of the birds
of Lord Howe, Norfolk and Kermadec Islands. Australasian
Record 1: 121–122.
Meredith, C.W., 1985. The fossil fauna of Norfolk Island, South-
west Pacific, and a review of the phylogeny of the genus
Pterodroma. Unpublished PhD thesis, Department of Zoology,
Monash University, Clayton, Victoria, Australia.
Meredith, C.W., 1991. Vertebrate fossil faunas from islands in
Australasia and the southwest Pacific. In Vertebrate
Palaeontology of Australasia , ed. P. Vickers-Rich, J.M.
Monaghan, R.F. Baird and T.H. Rich, pp. 1345–1382.
Melbourne: Monash University Publications Committee.
Moore, J.L., 1999. Norfolk Island bird notes, 1977–1997. Notornis
46: 354–364.
North, A.J., 1890. Nests and eggs of birds found breeding in
Australia and Lord Howe and Norfolk Islands. Australian
Museum Catalogue 12: 372–379, 412–414. Sydney: Australian
Museum.
Oliver, W.R.B., 1955. New Zealand Birds . Wellington: A.H. &
A.W. Reed.
Ramsay, E.P., 1888. Tabular List of All the Australian Birds at
Present Known to the Author, Showing the Distribution of the
Species Over the Continent of Australia and Adjacent Islands.
Sydney: The author.
Ripley, S.D., 1977. Rails of the World: a Monograph of the Family
Rallidae. Boston: Godine.
Saunders, H., and O. Salvin, 1896. Catalogue of the Gaviae and
Tubinares in the Collection of the British Museum (Catalogue
of the Birds in the British Museum Vol. XXV). London: British
Museum (Natural History).
Schodde, R., P. Fullagar and N. Hermes, 1983. A Review of Norfolk
Island Birds: Past and Present . Australian National Parks &
Wildlife Service, Special Publication 8. Canberra: Australian
National Parks & Wildlife Service.
Serventy, D.L., V. Serventy and J. Warham, 1971. The Handbook
of Australian Sea-birds. Sydney: A.H. & A.W. Reed.
Steadman, D.W., 1988. A new species of Porphyrio (Aves: Rallidae)
from archaeological sites in the Marquesas Islands. Proceedings
of the Biological Society of Washington 101: 162–170.

Holdaway and Anderson: Avifauna       99
Steadman, D.W., 1997. Extinctions of Polynesian birds: reciprocal
impacts of birds and people. In Historical Ecology in the Pacific
Islands: Prehistoric Environmental and Landscape Change ,
ed. P.V . Kirch and T.L. Hunt, pp. 51–79. New Haven and
London: Yale University Press.
Steadman, D.W., and S.L. Olson, 1985. Bird remains from an
archaeological site on Henderson Island, South Pacific: Man-
caused extinctions on an “uninhabited” island. Proceedings of
the National Academy of Sciences, U.S.A.  82: 6191–6195.
Tennyson, A.J.D., 1991. The black-winged petrel on Mangere
Island, Chathams Islands. Notornis 38: 111–116.
Tennyson, A.J.D., and G.A. Taylor, 1990. Behaviour of Pterodroma
petrels in response to “war-whoops”. Notornis 37: 121–128.
van Tets, G.F., C.W. Meredith, P.J. Fullagar and P.J. Davidson,
1988. Osteological differences between Sula and Morus, and a
description of an extinct new species of Sula from Lord Howe
and Norfolk Islands, Tasman Sea. Notornis 35: 35–57.
Wakelin, H., 1968. Some notes on the birds of Norfolk Island.
Notornis 15: 156–174.
Whitley, H.P., 1938. Naturalists of the First Fleet. Australian
Museum Magazine 6: 291–304.
Worthy, T.H., 1998. A remarkable fossil and archaeological
avifauna from Marfells Beach, Lake Grassmere, South Island,
New Zealand. Records of the Canterbury Museum 12: 79–176.
Problems of identification, reasons for inclusion or exclusion of
taxa from the Norfolk Island avifauna, and comments on the time
of some extinctions.
Petrels. There have long been problems in identifying the
species of petrel that originally bred on Norfolk Island,
notwithstanding the copious fossil remains and the existence of
paintings from the early period of European occupation. All the
Pterodroma petrels are difficult to separate on fragmentary material
such as is normal in middens. It is usually necessary to examine
crania of species that are closely similar in post-cranial dimensions.
Meredith (1991) recorded three species of Pterodroma —
Pterodroma solandri, P. pycrofti , and an undescribed species—
from Norfolk Island. P . solandri and P . pycrofti are now extinct
on Norfolk Island itself. Pterodroma solandri breeds at Lord Howe
Island 900 km to the west, and a small breeding population has recently
been found on Philip Island (Hermes et al., 1986). It is not known
whether the Philip Island colony represents a recolonization of the
Norfolk Island group from the Lord Howe population or is a remnant
of the population that was otherwise extirpated on the Norfolk Island
group during the early years of European settlement.
The only large Pterodroma to be recorded as definitely breeding
on Norfolk Island is Pterodroma solandri, the Bird of Providence,
Providence Petrel, or Solander’s Petrel. Meredith (1985, 1991)
recorded this species as being common in fossil deposits on
Norfolk Island. He also recorded an unnamed Pterodroma of
intermediate size (Pterodroma new species), but listed mainly leg
elements; wing elements from other collections were referred. In
the present study, it became apparent that, although all Pterodroma
humeri larger than those of P . pycrofti were of a size range
consistent with specimens of P . solandri, there were two size classes
in the leg elements. The longer femora, tibiotarsi, and tarsometatarsi
were long enough to be of P . solandri but others were more comparable
to elements of Pterodroma inexpectata and therefore were at first
attributed to Pterodroma new species of Meredith (1985, 1991). The
absence of wing bones in the intermediate size range was problematic,
and other possibilities were explored.
Two other species of Pterodroma in the Pacific have humeri of
about the same length as that of Pterodroma solandri, but shorter,
thinner legs. These are P . neglecta (Kermadec Petrel) and P.
arminjoniana (Herald Petrel). The Pacific populations of Herald
Petrel have been recognized as a separate species ( P . heraldica)
and Brooke and Rowe (1996) split that species into white-bellied
(P . heraldica) and dark-plumaged birds (P . atrata). The presence
of any of these species would explain the anomalous pattern of
smaller leg elements but larger humeri and ulnae. Measurements
of leg elements of both these species (T.H. Worthy, pers. comm.)
are similar to those in the collection considered here. The material
is referred to P . neglecta on the basis of present breeding range.
Harrison (1983) gave the breeding range of P . arminjoniana (P.
heraldica) in the Pacific as including Chesterfield Reef, Tonga,
Marquesas, Tuamotus, Gambier Islands, Pitcairn group, and Easter
Island. The putative P . atrata is confined to the Pitcairn Islands.
These islands are all either north of, or are very close to, the Tropic
of Capricorn: none of those in the western South Pacific is south
of 23°S. The range of P . neglecta includes Lord Howe, Kermadecs,
Austral, Pitcairn, and Easter groups, and islands off Chile. In the
southwestern Pacific its range is well south of that of P.
arminjoniana (P . heraldica) (Lord Howe is at 32°S, and Raoul
Island is at 28°S, roughly the same as Norfolk Island). As Norfolk
Island lies between two of the present breeding stations of the
Kermadec Petrel, it is reasonable to suspect that P . neglecta
formerly bred there. Hindwood (1940) pointed out that two
specimens taken by Dr P.H. Metcalfe in the 1880s referred to P.
solandri by North (1890) were in turn referred by Mathews (1912)
to P . neglecta, which, if correct, would constitute the first record
of that species from Norfolk Island, and add weight to the
conclusion that bones referred to an “intermediate” Pterodroma
(Meredith, 1991) are actually of P . neglecta.
A further point that may aid in distinguishing between the two
species in old accounts is that Pterodroma neglecta usually nests
on the surface and does not burrow like P . solandri (Bartle et al.,
1993). Records of surface-nesting petrels in early accounts are likely
to refer to P. neglecta. A confounding factor may be the former presence
on the islands of at least one species of Accipiter, whose predation
could conceivably have forced P. neglecta to nest under cover.
The identity of the small species of Pterodroma that formerly
bred in large numbers on Norfolk Island has been especially
problematic. Until Meredith (1985, 1991) established that the small
Pterodroma  bones from the island represented a previously
unknown and very large population of P . pycrofti, it was thought
that P . nigripennis was probably present at the time of European
settlement. Pterodroma pycrofti is now absent from the Norfolk
Island group: it breeds on small islands off the northeastern coast
of the North Island of New Zealand (Heather and Robertson, 1996).
There are too few P . pycrofti on the present breeding islands (1500+
pairs, Heather and Robertson, 1996) for there to be any pressure
for young birds to find new breeding sites.
Part of the reason for suspecting the former presence of P.
nigripennis was that, by analogy with the present small population
of P . solandri, the P . nigripennis presently breeding on Philip Island
and attempting to breed on Norfolk Island have been taken to be a
recolonization after extermination in the 1790s (Schodde et al.,
1983). In fact, P . nigripennis was first identified on Norfolk Island
in 1965 (Serventy et al., 1971). It has not been identified among
the fossils in either natural or archaeological contexts (Meredith,
1985, 1991; this study), so it is likely to be instead a recent colonist,
as it is in northern New Zealand and the Chatham Islands
(Tennyson, 1991) where there also no records of former breeding.
The present populations on Lord Howe Island and Balls Pyramid
cannot be used as evidence of a former more extensive breeding range
of P. nigripennis, as can be proposed for P . neglecta. The breeding
colonies of P. nigripennis (Philip Island, Lord Howe Island, Balls
Pyramid) are part of the recent and on-going south and southwestward
expansion (Marchant and Higgins, 1990). The source is probably the
large population (2–3 million pairs, Heather and Robertson, 1996) on
Macauley Island in the Kermadecs, 1300 km to the east. Pterodroma
nigripennis is not mentioned by Hindwood (1940) as being part of
the Lord Howe Island fauna, although he records two skins of
Cookilaria [=Pterodroma] cookii (under the common name “Blue-
footed Petrel”), citing a breeding range that includes the Kermadecs,
where P . cookii does not breed. The 1965 specimen of P . nigripennis
on Norfolk Island had blue instead of the typical fleshy pink legs
(Serventy et al., 1971). A small proportion of P . nigripennis individuals
has blue legs (Serventy et al., 1971) so it is possible that the birds
seen by Hindwood were of this form rather than P . cookii. The
specimens should be re-examined.
Appendix

100       Records of the Australian Museum, Supplement 27 (2001)
The identity of the small Pterodroma petrel portrayed in a
coloured drawing done at the time of the First Settlement (no. 96
in the “Sydney” collection, Hindwood, 1965) remains an enigma.
Whitley (1938) described it as a new species Cookilaria hindwoodi.
The discovery of a specimen of P . nigripennis on Norfolk led
Hindwood (1965: 90) to suggest that the “Norfolk Island Dove-
Petrel [had] been re-discovered”. Earlier, Hindwood and Serventy
(1943) considered that the bird “differs from all known species of
Cookilaria in the brown colour of the upper parts” and considered
it to be incertae sedis . As painted, the bird appears to lack the
“powdering of grey… from the nape down the sides of the upper
breast” (Serventy et al., 1971) that is characteristic of P . nigripennis
and the dimensions (if the bird was indeed painted life size) fit
with those of P . pycrofti (Oliver, 1955). It is possible, contra
Serventy et al. (1971: 103), that the bird in the painting was P.
pycrofti. In favour of this interpretation are the dimensions taken
by G.M. Mathews from the painting (Hindwood and Serventy,
1943), which are all within the ranges for P . pycrofti. In addition,
the colour pattern is similar to that species, although the colour
itself apparently differs in having a brown back, which could be
an artefact of the paint used. The reference in Heather and
Robertson (1996) to a “huge” colony of P . nigripennis on Norfolk
Island being destroyed by cats and rats ignores the fact that the
birds were not known to breed there before 1965 and have not
been identified in the large fossil collections (Meredith, 1991, and
see above). The simplest view, and the one adopted here, is that
the late Holocene avifauna of Norfolk Island included one species
of small Pterodroma petrel, P . pycrofti, which became extinct
shortly after European settlement, and whose place is now being
taken—for whatever reason, and in the absence of a large source
population of P . pycrofti—by immigrant P . nigripennis from the
thriving populations on the southern Kermadec Islands.
A recent summary of birds reported from the Norfolk Island group
(Moore, 1999) includes references to several other petrels that may
well have been part of the original avifauna. These include the Flesh-
footed Shearwater (Puffinus carneipes: in a burrow on Philip Island),
Newells Shearwater ( Puffinus auricularis newelli : captured,
photographed and released on Philip Island), Tahiti Petrel
(Pseudobulweria rostrata: at sea within 15 km of the island, breeds
in the South Pacific), Goulds Petrel (Pterodroma leucoptera: one race
of which breeds on New Caledonia), White-necked Petrel (Pterodroma
cervicalis: breeding in large numbers on Macauley Island in the
Kermadec group, 1200 km east of Norfolk, and a nesting pair found
on Philip Island in 1992), and Kermadec Petrel (Pterodroma neglecta:
breeding on islets off Raoul Island, Kermadecs, and at Lord Howe
Island, found nesting on Philip Island in 1992). Other species, such
as the Cape Petrel (Daption capense), have been seen near the islands,
but their breeding grounds are in the subantarctic and it is highly
unlikely that there were breeding populations of these species in the
Norfolk Island area at any time in the past.
There appear to be previous records of the occurrence of two
of these species on the island. A specimen of Puffinus carneipes
taken on Norfolk Island by E.H. Saunders (Saunders and Salvin,
1896) is apparently the first record from the island. Meredith (1991)
did not list P . carneipes in the fossil fauna while Schodde et al.
(1983) recorded it only as a vagrant before the recent breeding
record (Moore, 1999). The breeding distribution of the species
includes Lord Howe Island as well as islands off northern New
Zealand and many around Australia (Hindwood, 1940; Serventy
et al., 1971). Hence, as with P . neglecta, it is likely on the grounds
of a gap in an otherwise continuous distribution that the species
once bred at Norfolk Island. For this reason, and as the
measurements of P . carneipes overlap with those of P . pacificus
(Oliver, 1955; Serventy et al., 1971), it is possible that material of
P . carneipes exists unrecognized in the fossil collections. If so, it
is included here in the unresolved “other petrel” category.
In addition to the petrels dealt with above, it is possible that at
least one other subtropical petrel may have had a breeding
population in the group: four individuals of the Phoenix Petrel
(Pterodroma alba) were found ashore on Raoul Island in 1913
(Oliver, 1955), and it is likely that the species bred there before
rats and cats were introduced (Holdaway et al., 2001). That some
of the subtropical species have now been found breeding, or
attempting to breed, on Philip Island is evidence that they could
have included the group in their breeding range in the past. The
possibility of their former presence adds complexity to an already
difficult identification problem.
Sulids. The sulid presently breeding in the Norfolk Island group
is the Masked Booby Sula dactylatra. An apparently extinct species
of booby (Sula tasmani) has been described from fossil remains
collected on Norfolk and Lord Howe Islands (van Tets et al., 1988).
The material in the present collection is attributed to S. dactylatra
because the mensural differences listed by van Tets et al. (1988)
are not sufficient to support recognition of a separate taxon and
instead represent the upper size range of S. dactylatra (Holdaway
and Anderson, unpubl. data).
Other differences proposed included choice of nesting habitat, S.
tasmani apparently differing from S. dactylatra in nesting on sand
beaches where they were vulnerable to predation by humans (van
Tets et al., 1988). When undisturbed by humans, even Australasian
Gannets (Morus serrator), which typically nest on or above high sea
cliffs, nest on sand dunes at sea level (Hawkins, 1988).
Waders. Although several species of charadriiform have been
identified from the island, all are vagrants or regular migrants.
Only two species, the Bar-tailed Godwit (Limosa lapponica) and
Pacific Golden Plover (Pluvialis fulva) are regular in numbers on
the island. Although Whimbrels (races of Numenius phaeopus )
have been reported live and as fossil from Norfolk Island, none
was recorded in the archaeological collection.
Hawks. The only predatory bird on the island today is the
Australian Kestrel (Falconidae: Falco cenchroides), which became
established as a breeding species in the 1970s (Schodde et al .,
1983). Reports of the presence of “hawks” in the 1790s were
confirmed by the discovery of remains of an Accipiter very closely
related to, if not identical to, the Brown Goshawk Accipiter
fasciatus of Australia and some islands to the north of Norfolk
Island, including New Caledonia (Meredith, 1985, 1991). It has
been thought that hawks died out on Norfolk Island very soon
after Europeans arrived or were vagrants (Schodde et al., 1983),
but Gurney (1854) referred to a report by F. Strange that hawks
were on Philip Island some time before 1853, presumably during
the residence of Strange’s informant on Norfolk Island, which
may mean that the goshawk survived on Philip Island for several
decades after its demise on the main island. Philip Island was still
vegetated at that time, and Strange records having met the man
“who exterminated the Nestor productus of Philip Island”. After
describing the way that the large parrot used its bill in climbing,
Strange reports that “He likewise informed me that there was a
large species of hawk that used to commit great havoc amongst
them [the parrots], but what species it was he could not tell me.”
Rails . The flightless endemic Gallirallus  new species
discovered by Meredith (1985) may have survived into the
European period. A rail painted on Norfolk Island in the 1790s
(no. 79 in the “Sydney” series, Hindwood, 1965) has a plumage
pattern similar to that of G. philippensis , but it may represent
Gallirallus  new species rather than the extant Gallirallus
philippensis as has been assumed. Ripley (1977) identified Rallus
tenebrosus (Gray, 1862), a small rail described from Norfolk Island
in 1824, as the Spotless Crake ( Porzana tabuensis ), which is
widespread in Australia and the South Pacific. At present, both P.
tabuensis and G. philippensis are vagrants on Norfolk Island, rarely
breed there, and their former status has been uncertain (Schodde
et al., 1983).
Parrots. Nestor productus  survived until the late 1840s on
Philip Island (Strange, in Gurney, 1854) after being extirpated on
the main island in the early 1800s. By late 1853 the species was
known only from Philip Island. No mention of its former presence
on the main island is made, although it figures prominently in the
collections of paintings made during the first convict settlement.
Strange’s informant said that “they rarely made use of their wings,
except when closely pressed” and that when he went to the island
to shoot them, he “would invariably find them on the ground”.
Such habits are not unusual in species confined to uninhabited
islands without mammalian predators and would have made the
birds easy prey for people.

Full-text PDF of each one of the works in this volume are available at the following links :  
 
 
Anderson and White, vol. eds, 2001, Rec. Aust. Mus., Suppl. 27: 1–143 
http://dx.doi.org/10.3853/j.0812-7387.27.2001.1334 
 
Anderson and White, 2001, Rec. Aust. Mus., Suppl. 27: 1–9 
http://dx.doi.org/10.3853/j.0812-7387.27.2001.1335 
 
Anderson et al., 2001, Rec. Aust. Mus., Suppl. 27: 11–32 
http://dx.doi.org/10.3853/j.0812-7387.27.2001.1336 
 
Anderson et al., 2001, Rec. Aust. Mus., Suppl. 27: 33–42 
http://dx.doi.org/10.3853/j.0812-7387.27.2001.1337 
 
Anderson and Green, 2001, Rec. Aust. Mus., Suppl. 27: 43–51 
http://dx.doi.org/10.3853/j.0812-7387.27.2001.1338 
 
Marianne et al., 2001, Rec. Aust. Mus., Suppl. 27: 53–66 
http://dx.doi.org/10.3853/j.0812-7387.27.2001.1339 
 
Schmidt et al., 2001, Rec. Aust. Mus., Suppl. 27: 67–74 
http://dx.doi.org/10.3853/j.0812-7387.27.2001.1340 
 
Smith et al., 2001, Rec. Aust. Mus., Suppl. 27: 75–79 
http://dx.doi.org/10.3853/j.0812-7387.27.2001.1341 
 
Matisoo-Smith et al., 2001, Rec. Aust. Mus., Suppl. 27: 81–84 
http://dx.doi.org/10.3853/j.0812-7387.27.2001.1342 
 
Holdaway and Anderson, 2001, Rec. Aust. Mus., Suppl. 27: 85–100 
http://dx.doi.org/10.3853/j.0812-7387.27.2001.1343 
 
Walter and Anderson, 2001, Rec. Aust. Mus., Suppl. 27: 101–108 
http://dx.doi.org/10.3853/j.0812-7387.27.2001.1344 
 
Campbell and Schmidt, 2001, Rec. Aust. Mus., Suppl. 27: 109–114 
http://dx.doi.org/10.3853/j.0812-7387.27.2001.1345 
 
Neuweger et al., 2001, Rec. Aust. Mus., Suppl. 27: 115–122 
http://dx.doi.org/10.3853/j.0812-7387.27.2001.1346 
 
Macphail et al., 2001, Rec. Aust. Mus., Suppl. 27: 123–134 
http://dx.doi.org/10.3853/j.0812-7387.27.2001.1347 
 
Anderson and White, 2001, Rec. Aust. Mus., Suppl. 27: 135–141 
http://dx.doi.org/10.3853/j.0812-7387.27.2001.1348