Conference PaperPDF Available

Miocene fossils show that kiwi (Apteryx, Apterygidae) are probably not phyletic dwarves. Pp 63–80, In Göhlich, U.B. & Kroh, A. (eds): Paleornithological Research 2013 – Proceedings of the 8th International Meeting of the Society of Avian Paleontology and Evolution.

Abstract and Figures

Abstract — Until now, kiwi (Apteryx, Apterygidae) have had no pre-Quaternary fossil record to inform on the timing of their arrival in New Zealand or on their inter-ratite relationships. Here we describe two fossils in a new genus of apterygid from Early Miocene sediments at St Bathans, Central Otago, minimally dated to 19–16 Ma. The new fossils indicate a markedly smaller and possibly volant bird, supporting a possible overwater dispersal origin to New Zealand of kiwi independent of moa. If the common ancestor of this early Miocene apterygid species and extant kiwi was similarly small and volant, then the phyletic dwarfing hypothesis to explain relatively small body size of kiwi compared with other ratites is incorrect. Apteryx includes five extant species distributed on North, South, Stewart and the nearshore islands of New Zealand. They are nocturnal, flightless and comparatively large birds, 1–3 kg, with morphological attributes that reveal an affinity with ratites, but others, such as their long bill, that differ markedly from all extant members of that clade. Although kiwi were long considered most closely related to sympatric moa (Dinornithiformes), all recent analyses of molecular data support a closer affinity to Australian ratites (Casuariidae). Usually assumed to have a vicariant origin in New Zealand (ca 80–60 Ma), a casuariid sister group relationship for kiwi, wherein the common ancestor was volant, would more easily allow a more recent arrival via overwater dispersal.
Content may be subject to copyright.
published 10 Dec. 2013
© Verlag Naturhistorisches Museum Wien, 2013
Paleornithological Research 2013
Proceed. 8th Inter nat. Meeting Society of
Avian Paleontology and Evolution
Ursula B. Göhlich & Andreas Kroh (Eds)
– 63 –
Kiwi (Apteryx, Apterygidae) are the most iconic
of New Zealand birds. Smallest of the extant
ratites, the ve species of Apteryx now recognised
(Gill et al. 2010) are all chicken-sized, ightless,
nocturnal birds that are supercially convergent
on mammals with their fur-like plumage, burrow-
breeding behaviour, and dependence on olfactory
and tactile rather than optical senses (Calder
1978). One of the most extraordinary peculiarities
of kiwi (kiwi and moa may be both singular and
plural because in Māori there is no ‘s’ to denote the
plural) is the huge egg that they produce – more
than four times the size of that predicted from
their body weight – which allows the production
of an extremely precocial chick (Calder 1978).
This characteristic led to the hypothesis that kiwi
are phyletic dwarfs, as rst espoused by Calder
(1978, 1984) and championed by Gould (1986,
1991). This hypothesis suggested that extant kiwi
were the outcome of an evolutionary trajectory
of a reduction in body size based on a perceived
sister-group relationship with the giant moa
ZooBank LSID:
Miocene fossils show that kiwi (Apteryx, Apterygidae)
are probably not phyletic dwarves
1 School of Earth and Environmental Sciences, University of Adelaide, Adelaide, Australia;
E-mail: trevor.worthy@; Present address: School of Biologial Sciences, Flinders University
2 Museum of New Zealand Te Papa Tongarewa, Wellington, New Zealand
3 School of Biological Sciences, University of Queensland, Brisbane, Australia
4 School of Biological, Earth and Environmental Sciences, University of New South Wales, Australia
5 Canterbury Museum, Christchurch, 8013, New Zealand
Abstract — Until now, kiwi (Apteryx, Apterygidae) have had no pre-Quaternary fossil record to inform on the
timing of their arrival in New Zealand or on their inter-ratite relationships. Here we describe two fossils in a new
genus of apterygid from Early Miocene sediments at St Bathans, Central Otago, minimally dated to 19–16 Ma.
The new fossils indicate a markedly smaller and possibly volant bird, supporting a possible overwater dispersal
origin to New Zealand of kiwi independent of moa. If the common ancestor of this early Miocene apterygid spe-
cies and extant kiwi was similarly small and volant, then the phyletic dwarng hypothesis to explain relatively
small body size of kiwi compared with other ratites is incorrect. Apteryx includes ve extant species distributed
on North, South, Stewart and the nearshore islands of New Zealand. They are nocturnal, ightless and compara-
tively large birds, 1–3 kg, with morphological attributes that reveal an afnity with ratites, but others, such as
their long bill, that differ markedly from all extant members of that clade. Although kiwi were long considered
most closely related to sympatric moa (Dinornithiformes), all recent analyses of molecular data support a closer
afnity to Australian ratites (Casuariidae). Usually assumed to have a vicariant origin in New Zealand (ca 80–60
Ma), a casuariid sister group relationship for kiwi, wherein the common ancestor was volant, would more easily
allow a more recent arrival via overwater dispersal.
Key words: Apterygidae, fossil record, evolution, new species, ightlessness
SAPE Proceedings 2013
– 64 –
ent origins of ratites involving dispersal by volant
ancestors and subsequent convergent evolution
towards the ratite form (HarsHman et al. 2008;
PHilliPs et al. 2010), although morphological data
(JoHnston 2011; WortHy & sCofield 2012) still
supports a vicariant origin. It is therefore possible
that both moa and kiwi may have dispersed as
volant, and therefore small, birds to New Zealand
after its separation from East Gondwana.
Resolution of these contrasting hypotheses
will be helped by a fossil record that establishes
limits such as lineage presence and actual mor-
phological form at crucial times. The fossil
history of moa prior to the Quaternary remains
elusive, but eggshell and tantalising fragments
from the St Bathans Fauna, South Island (Wor-
tHy et al. 2007; tennyson et al. 2010) show that
moa ancestors were present and were large birds
in the Early Miocene.
The St Bathans Fauna has produced a diverse
assemblage of terrestrial vertebrates including
leiopelmatid frogs, reptiles including skinks,
geckos, turtles and crocodilians, and mammals
(Jones et al. 2009; lee et al. 2009; WortHy et
al. 2006, 2011a, 2011b). The terrestrial verte-
brate fauna (non-sh) is however dominated in
diversity and abundance by about 40 species of
birds, principally of waterfowl (Anatidae), with a
minimum of eight taxa in ve genera. It includes
moa (Dinornithiformes), a tubenose (Procellari-
iformes), birds of prey (Accipitriformes), several
gruiforms (Rallidae), a gull (?Laridae) and other
charadriiforms, herons (Ardeidae), a palaelodid
(Phoenicopteriformes), pigeons (Columbidae),
parrots (Psittaciformes), a swiftlet (Apodidae),
an owlet-nightjar (Aegothelidae), and passerines
(Passeriformes) (sCofield et al. 2010; tennyson
et al. 2010; WortHy et al. 2007, 2009, 2010a,
2010b, 2011c, 2011d). The fauna includes repre-
sentatives of all the quintessential endemic New
Zealand terrestrial vertebrates such as leiopelma-
tids (WortHy et al. 2011b), sphenodontids (Jones
et al. 2009), moa (tennyson et al. 2010), the
basal gruiform Aptornis (WortHy et al. 2011c),
and acanthisittid wrens (WortHy et al. 2010a). It
has, however, not revealed any evidence of that
most iconic of all New Zealand taxa, the kiwi.
This absence is now informed by the discovery of
two fossils referrable to Apterygidae. They allow
assessment of the phyletic dwarng hypothesis
(Dinornithiformes) and a lack of any Cenozoic
fossil record.
Ratites (ostrich, rhea, cassowary, emu,
elephant bird, moa and kiwi) are ightless pal-
aeognaths with greatly reduced wings, or, in the
case of moa, completely lost. In the absence of
an informative fossil record, interpretation of rat-
ite origins has been limited to inference from the
highly modied extant representatives. In recent
decades, ratites have come to be regarded as
one of the best vertebrate exemplars of a group
with vicariant origins e.g., roff (1994), follow-
ing initial promotion of the idea by CraCraft
(1974). However, analyses of molecular data,
e.g., CooPer et al. (2001) and HaddratH & Baker
(2001), cast doubt on the vicariant origin of vari-
ous ratite clades with unanimous support for kiwi
having a closer relationship to Australian ratites
(Casuariidae) than to moa. Thus the rst premise
underpinning the phyletic dwarng hypothesis
of kiwi origins – a sister-group relationship with
moa is now doubtful. With the divergence of
kiwi from casuariids then calculated to have
occurred at about 60 Ma, it has been suggested
that the occurrence of kiwi in New Zealand
required dispersal over a signicant oceanic bar-
rier following Zealandia’s separation from East
Gondwana approximately 80 Ma (mCnaB 1994;
CooPer et al. 2001). However, it is now recog-
nised that the unzipping of Zealandia (inclusive
of New Zealand) from East Gondwana took
over 27 Ma, commencing 82 Ma and nishing
approximately 55 Ma (Gaina et al. 1998; sCHel-
lart et al. 2006). If this was indeed the case, then
a vicariant origin for kiwi remains a possibility
(tennyson 2010). This is especially so given that
the most recent estimates for the divergence of
kiwi from casuariids continue to support their
ancient origin, e.g., 53.5 (95 % CIs 36.9–72.1)
Ma, estimated with both external and internal fos-
sil calibrations (PHilliPs et al. 2010), and 73 (95 %
CIs 50–100) Ma as estimated in BEAST by Had-
dratH & Baker (2012). But the question arises,
was the common ancestor of kiwi and casuariids
large and ightless, or was it small and volant as
might be predicted by the multiple loss of ight
hypothesis invoked for ratites (HarsHman et al.
2008; PHilliPs et al. 2010)? More recently, even
the vicariant origin of moa has been questioned,
as molecular analyses suggest multiple independ-
WORTHY ET AL.: Kiwi are probably not phyletic dwarves
– 65 –
OR.14964; NMNZ OR.17206; NMNZ OR.17207;
NMNZ OR.17208; NMNZ OR.17209; NMNZ
OR.17210; NMNZ OR.17211; NMNZ OR.17212;
NMNZ OR.17213; NMNZ OR.24640; NMNZ
OR.24984; A. rowi: CM Av16691, CM Av16717,
CM Av16718.
Crypturellus obsoletus, SMF 2148; C. nocti-
vagus noctivagus, SMF 11394; C. tataupa, SMF
11392; C. parvirostris, SMF 2164, SMF 9357,
SMF 8184; C. cinnamomeus, SMF 2537; C.
undulatus vermiculatus, SMF 2149; Eudromia
elegans, NMW 3.071, SMF 9306, SMF 9260,
SMF 6111, SMF 5415, SMF 6416, SMF 6298;
Nothoprocta perdicaria, NMNZ OR.22983,
NMW 4.068; Nothura maculosa, NMW 1061;
Tinamus major robusta, SAM B.31339; T. major
major, NMNZ OR.1433; T. major, NMW 4.559;
T. (Trachpilmus Cab.) robustus, NMB C.2004;
T. solitarius, SMF 2150; SMF 2146; Rhynchotus
fasciatus =rufescens, NMW 161, NMW 160; R.
rufescens, SMF 2147, NMB 5537.
We describe two fossil bones from the Early
Miocene St Bathans Fauna, from Central Otago,
New Zealand. The locality details and general
stratigraphy of the sites producing this fauna
have been described already (WortHy et al.
2007; sCHWarzHans et al. 2012).
We estimated mass for the fossil kiwi based
on its femoral circumference using six different
algorithms, including four from the literature
based on all kinds of birds (anderson et al. 1985;
CamPBell & marCus 1992) and ratites (diCkin-
son 2007), as well as two newly derived ones
based on kiwi and tinamous. Circumference was
determined by wrapping a narrow and thin piece
of cellotape (or string for extant kiwi specimens
in NMNZ) around the mid-shaft and marking the
point where it overlapped itself, then measuring
the length with callipers. Given apterygids are
rather atypical palaeognaths, we also computed
algorithms for mass from data for tinamous and
for apterygids separately, and used these to esti-
mate the mass of the fossil kiwi. We reasoned
that if the fossil taxon was volant the value
based on tinamou may be more pertinent but if
it were ightless then that based on kiwi would
be relevant. Measurements were taken with dial
callipers or a graticule in a binocular microscope
and rounded to the nearest 0.1 mm.
and have bearing on when kiwi joined the New
Zealand biota.
Material and methods
Nomenclature: We follow Gill et al. (2010)
for nomenclature of kiwi and use names from
specimen labels interpreted via davies (2002) for
tinamous. We use the anatomical nomenclature
given in Baumel et al. (1993) and elzanoWski
& stidHam (2010) and abbreviate common terms
as follows: artic., articularis; cond., condyle; m.,
musculus; proc., processus; tuber., tuberculum.
Abbreviations: AM, Australian Museum,
Sydney, New South Wales, Australia; CM, Can-
terbury Museum, Christchurch, New Zealand;
MV, Museum Victoria, Melbourne, Victoria, Aus-
tralia; NMB, Naturhistorisches Museum, Basel,
Switzerland; NMNZ, Museum of New Zealand
Te Papa Tongarewa, Wellington, New Zealand;
NMW, Naturhistorisches Museum, Vienna, Aus-
tria; SAM, South Australian Museum, Adelaide,
South Australia, Australia; SMF, Forschungsinsti-
tut Senckenberg, Sektion Ornithologie, Frankfurt
am Main, Germany. Ma, million years ago.
Comparative material examined: Comparisons
were made widely among birds using the skeletal
collections of the Australian Museum and South
Australian Museum. Following determination
of the apterygid afnity of the fossils, detailed
observations were made of the following kiwi
and tinamou specimens.
Apteryx owenii: SAM B.5051, MV B56009,
AMS535, AM A.1980, AM A.1992, AM
A.4570; NMNZ OR.22815a; NMNZ OR.23044;
NMNZ OR.23717a; NMNZ OR.24415; NMNZ
OR.24416; A. haastii: MV B40905, CM Av31538;
NMNZ OR.19773a; NMNZ OR.23022a;
NMNZ OR.23038; NMNZ OR.23045; NMNZ
OR.23648a; NMNZ OR.27983; NMNZ
OR.28010a; A. australis: CM Av14447, CM
Av32404, CM Av36637, CM Av36638, CM
Av39065, AM O. unregistered; A. a. australis:
NMNZ OR.22089a; NMNZ OR.27761a; NMNZ
OR.27965; A. a. lawryi: NMNZ OR.23591;
NMNZ OR.23756; A. mantelli: CM Av5492;
NMNZ DM.909-S; NMNZ OR.13588; NMNZ
SAPE Proceedings 2013
– 66 –
cranially (10), merges gradually to the corpus
femoris distally (11), and lateromedially broad
adjacent to the collum femoris such that a nar-
row groove, less than the diameter of the caput,
connects the pretrochanteric surface to the facies
artic. antitrochanterica (12); the pretrochanteric
surface is shallowly concave cranially, lacking
pneumatic foramina (13); the linea intermuscula-
ris cranialis extends to the distal end of the crista
trochanteris (14); the area for the insertion of the
m. obturatorius lateralis on an elevated bulge, not
marked by any scar (15); large sulcus centred on
the lateral facies proximally for insertion areas of
the m. iliotrochantericus caudalis cranially and
the m. iliofemoralis externus caudally (mCGoWan
1979), is elongate, extending about half the length
of the crista trochanteris (16); insertion areas for
the mm. iliotrochanterici medius et cranialis form
a narrow elongate groove, slightly separated from
the insertion area for the m. iliofemoralis exter-
nus, extending distally to point level with end of
the crista trochanteris (17); insertion area of the
m. ischiofemoralis forms a short, broad sulcus,
caudal of and overlapping, in the proximodistal
plane, the distal end of the insertion area for the
m. iliofemoralis externus and the proximal end
of the insertion area for the m. iliotrochantericus
Systematic Palaeontology
Order Casuariiformes: Cassowaries, Emus
and Kiwi
Family Apterygidae G.R. Gray, 1840: Kiwi
The fossil is identied as an apterygid by the fol-
lowing combination of femoral characters (Figs
1, 2): the facies artic. antitrochanterica is convex
in cranial-caudal section (1) and lateromedially
about same width as the caput femoris (2); in
caudal view, the proximal prole has a marked
notch between the caput femoris and the facies
artic. antitrochanterica (3); the collum femoris is
constricted proximodistally and craniocaudally
(4); the caudal facies distal to the facies artic.
antitrochanterica is at (5), forming a near right
angle with the lateral facies (6), not a curved tran-
sition, as in e.g. galliforms; the insertion area for
the major part of m. obturatorius medialis is on a
distinct bulge traversing the caudal facies disto-
laterally, ending laterally level with the insertion
area for m. ischiofemoralis (7); the depth of the
crista trochanteris is about twice the depth of
the caput femoris (8); the crista trochanteris not
extending proximad of the facies artic. antitro-
chanterica, no fossa trochanteris (9), rounded
FIGURE 1. Apterygid right femora. Proapteryx micromeros (NMNZ S.53324, A-C, E) and Apteryx owenii (MV
56009, D, F), in medial (A), cranial (B), caudal (C, D), and lateral (E, F) views. Scale bars are 10 mm. Numbers
refer to family attribution characters.
WORTHY ET AL.: Kiwi are probably not phyletic dwarves
– 67 –
stromatolites, Site FF1 (lindqvist 1994), a fos-
sil stromatolite bed at 44.90359°S, 169.85840°E,
Manuherikia River, Otago, New Zealand. Fos-
sil Record Number (FRN) in the archival Fossil
Record File of the Geological Society of New
Zealand is H41/f0058 (stromatolites) and H41/
f0059 (clay draping stromatolites).
Formation, Manuherikia Group, Early Miocene
(Altonian); 19–16 Ma; St Bathans Fauna. The
stratigraphic relationship of site FF1 to other
Bannockburn Formation exposures is presently
unknown but the associated faunas are similar.
Measurements of holotype: Preserved
(incomplete) length 42.2 mm, proximal width
8.4 mm, maximum proximal depth 7.6 mm ; shaft
width at former mid length 3.6 mm, shaft depth at
former mid length 4.0 mm.
Comparison and description: In addition
to the listed diagnostic characters, there are few
differences between Proapteryx and Apteryx.
The nutrient foramen caudally on the shaft is
distinctly proximal to mid-length in Apteryx, but
near original mid-length in Proapteryx. The linea
intermuscularis caudalis extends proximally
towards the lateral facies laterad of an elongate
prominence for the insertion of m. puboischi-
ofemoralis pars medialis (as in Casuariidae). This
medius (18); the corpus femoris is elongate rela-
tive to its proximal width (19), arched dorsally at
mid-length (20), and bent medially at the distal
end of the crista trochanteris, such that the lateral
facies beside the trochanter is markedly inclined
medially relative to more distal parts (21).
The insertion area for the m. iliotrochanteri-
cus caudalis being centred craniocaudally on
the lateral facies, not more cranially, may be a
synapomorphy of Casuariiformes. Characters 15
and 16 are considered apterygid autapomorphies.
Femora of all other birds are further distinguished
from those of apterygids by numerous features
(see Appendix 1).
Proapteryx gen. nov.
Types species: Proapteryx micromeros spec. nov.
Diagnosis: An apterygid distinguished from
Apteryx by the facies artic. antitrochanterica of
the femur having a well-formed lobe overhang-
ing the caudal facies; insertion area for the minor
part of the m. obturatorius medialis a marked
scar about a third the length of and located
proximocaudal to the insertion area of the m.
ischiofemoralis, not immediately caudal to it; and
by its markedly smaller size with a femoral shaft
diameter about half that of A. owenii, the smallest
Apteryx species.
Etymology: Addition of the Latinised Clas-
sical Greek prex προ- (pro-), meaning before,
to the scientic name of kiwi (Apteryx). Apteryx
is Latinised Classical Greek and derives from
the Greek “α”, a prex indicating to be without
or absent, and “πτέρυγας” = wings; neuter noun.
Denoting that this taxon precedes Apteryx in the
geological record.
Proapteryx micromeros spec. nov.
(Figs 1–3)
Holotype: NMNZ S.53324 (Figs 1, 2), right
femur missing distal condyles; collected 20 April
Diagnosis: As for genus.
Etymology: Latinised Classical Greek μικρός
(mikros) for small or little and μηρία (meros)
for thigh; neuter noun. For the markedly smaller
femur than in extant apterygids.
Type locality: In a clay layer enveloping
FIGURE 2. Apterygid right femora. Proapteryx mi-
cromeros (NMNZ S.53324, A, B) and Apteryx owenii
(a small example, SAM B.5095, C), in lateral (A, C),
and caudal (B), views. Numbers refer to family attri-
bution characters. Abbreviations: mic, insertion area
of m. iliotrochantericus caudalis; mie, insertion area
of m. iliofemoralis externus; mom, insertion area of m.
obturatorius medialis pars minor.
SAPE Proceedings 2013
– 68 –
10–15 cm thick sand and cobble layer 9.5–9.58
m above the base of the Bannockburn Formation,
Trench Excavation, at 44.90780° S 169.85844° E,
Manuherikia River section; FRN is H41/f0103.
Stratigraphy/Age/Fauna: As for holotype.
Measurements of referred specimen: Total
height from proc. oticum to cond. medialis 9.2
mm; height above cotyla quadratojugalis 4.9 mm.
prominence in Proapteryx is relatively more
proximally located, ending level with the caudal
bulge that is the insertion area for the major part
of m. obturatorius medialis, rather than distal to
the crista trochanteris.
Referred specimen: NMNZ S.53209, a left
quadrate (Fig. 3).
Locality of referred specimen: Bed HH1b,
FIGURE 3. Apterygid left quadrates. Apteryx owenii (SAM 5051 A-C, G, I) and Proapteryx micromeros (NMNZ
S.53209, D-F, H, J) in medial (A, D), lateral (B, E), anterolateral (C, F), anterior (G, H) and ventral (I, J) views.
Abbreviations: cm, crista medialis; cml, cond. mandibularis lateralis; cmm, cond. mandibularis medialis; co, ca-
pitulum oticum; cp, cond. pterygoideus; cqj, cotyla quadrateojugalis; cr, crista; cs, capitulum squamosum; faqv,
facies artic. quadrateojugalis ventralis; fqj, fovea quadrateojugalis; mcpc, medial process cond. pterygoideus; po,
proc. orbitalis; pot, proc. oticus; ts, tuber. subcapitulare.
WORTHY ET AL.: Kiwi are probably not phyletic dwarves
– 69 –
the cond. mandibularis lateralis is laterocaudal
of, narrower than, subparallel to, and overlaps
the cond. mandibularis medialis; 14, the cond.
mandibularis medialis has two distinct parts: a
ventrally convex medial part, and an anterocau-
dally broader lateral part that extends dorsally
onto the lateral facies on the anterior side of the
cotyla quadratojugalis.
In addition, NMNZ S.53209 reveals the fol-
lowing features. The preserved dorsal surface of
the capitilum squamosum is level with the eroded
surface of the capitulum oticum, indicating that
the latter was originally slightly more prominent
dorsally. Medially, the tuber. subcapitulare abuts
a crest linking the capituli to the proc. orbitalis.
A deep partially pneumatic sulcus at the base of
the capitulum oticum lies medial to this crest.
The proc. oticus is robust, extending dorsad of
the cotyla quadratojugalis by slightly over half
(53 %) of total quadrate height, anterior and cau-
dal borders are subparallel in anterolateral view,
being formed respectively from crests extending
from the proc. orbitalis to the capituli anteriorly
and from the cotyla quadratojugalis to the capitu-
lum squamosum posteriorly. The anterior margin
of the proc. oticus meets the dorsal margin of the
proc. orbitalis in a broad (approximately 140º)
angle level with the dorsal side of the cotyla
quadratojugalis. Anterior to the cond. mandibu-
laris lateralis and lateral to the cond. medialis a
shallow sulcus undercuts the cotyla quadratoju-
galis ventrally.
Quadrates of Apteryx species differ from
NMNZ S.53209 as follows. (1) They are rela-
tively more robust, although the smallest extant
species, A. owenii, has a similar height (Table 1;
Appendix 2). (2) They lack a pneumatic fossa
anteriorly under the capitulum oticum. (3) The
proc. oticus is relatively shorter with its height
above the cotyla quadratojugalis slightly less
than half of the maximum dorsoventral height.
(4) The laterocaudal margin of the proc. oticus
above the cotyla quadratojugalis is more rounded
(in NMNZ S.53209, the corpus is notably cranio-
caudally compressed forming a ridge extending
from under the capitulum squamosum down
nearly to the cotyla quadratojugalis). (5) The
cond. mandibularis medialis is proportionally
larger but relatively less protuberant ventrally,
and less offset ventrally from the ventral margin
Comparison and description: The fossil is
worn and damaged with the proc. orbitalis lost
anterior to the cond. pterygoideus, loss of much
of the cond. mandibularis lateralis including the
caudoventral rim of the fovea quadratojugalis,
and the anteriomedial tip of the cond. mandibula-
ris medialis is worn.
NMNZ S.53209 differs markedly from
quadrates of all birds except those of Apteryx
with which it shares the following combina-
tion of features (Fig. 3) and so it is referred to
Apterygidae: 1, the head of the proc. oticus is
expanded laterally and medially into a broad
‘dumbell’ shape, about three times wider than
long, aligned at right angles to the proc. orbit-
alis; 2, caudally, there are pneumatic foramina at
the base of the capitulum oticum; 3, the capituli
oticum et squamosum are linked by a craniocau-
dally narrow articular surface lacking an incisura
intercapitalis; 4, the capitulum squamosum wid-
ens laterally and its articular surface extends
caudoventrally as a small oval lobe protuberant
over the caudal facies; 5, the tuber. subcapitulare
on the anterior side of the capitulum squamosum
is prominent, robust, and about twice as wide as
high; 6, the proc. oticus has a at caudal facies,
straight in lateral aspect, centred above the cond.
mandibularis medialis, and forming an angle of
approximately 100 degrees with the proc. orbit-
alis; 7, the crista medialis is acute, extending
from the medial prominence of the cond. ptery-
goideus towards the capitulum oticum, forming
the caudal boundary to a deep triangular sulcus
without a foramen pneumaticum rostromediale;
8, the corpus lacks pneumatic foramina both cau-
dally and laterally; 9, the cotyla quadratojugalis
is very prominent laterally, the fovea quadratoju-
galis is deep, and the facies artic. quadratojugalis
ventralis is proportionately large; 10, the proc.
orbitalis is lateromedially thin with a shallow
sulcus medially; 11, the cond. pterygoideus is
rectilinear, separated by a sulcus from and lies
dorsal to the medial half of the cond. mandibu-
laris medialis, broader than high, dorsally convex
in section, lateromedially concave, and extends
continuously from a prominence medially onto
the ventromedial part of the proc. orbitalis; 12,
in ventral view, the condyli mandibularis media-
lis et lateralis are aligned roughly at right angles
to the proc. orbitalis; 13, the articular surface of
SAPE Proceedings 2013
– 70 –
fossil therefore has little similarity with dinorni-
thiform quadrates.
Lithornithidae. Lithornithids are volant pal-
aeognaths found in the Northern Hemisphere
in the Early Tertiary (Houde 1988) and could
conceivably have been ancestral to Casuari-
iformes, which are not known to be older than
Late Oligocene ca. 24 Ma (Boles 2001). If so, a
closer relationship and hence greater osteological
similarity with Apteryx than to most other ratites
might be predicted. We compared images of quad-
rates of Lithornis celetius Houde, 1988 (USNM
290601) with Apteryx quadrates. Similarities
include: the proc. oticus is dorsally expanded
with poorly separated capituli aligned at right
angles to the proc. orbitalis, the cond. mandibu-
laris medialis is medially prominent below the
saddle-like cond. pterygoideus whose articular
facet curves around onto the ventral part of the
proc. orbitalis; the proc. orbitalis is relatively
low (dorsoventrally) with a sulcus medially.
However, the Lithornis quadrate differed from
Apteryx quadrates and the fossil as follows: the
proc. oticus is dorsally convex across the capituli,
rather than the capituli being separated by a slight
hollow; the capitulum oticum is markedly con-
vex dorsally, does not overhang the corpus either
posteriorly or anteriorly (not comparable in fos-
sil); the capitulum squamosum is less protuberant
laterally; the tuber. subcapitulare is lacking or
poorly developed; the cotyla quadratojugalis is
located relatively more ventrally, its ventral mar-
gin aligned with the ventral margin of the proc.
orbitalis, has subequal ventral extent with the
cond. mandibularis medialis; the cotyla quadra-
tojugalis is not as protuberant laterally, has a crest
extending from the dorsal margin to the corpus
at slightly above mid height; the rostromediale
facies of the corpus lacks a distinct sulcus at mid-
height (present in Apteryx) where the foramen
of the base of the proc. orbitalis, creating a deeper
sulcus under the cond. pterygoideus. (6) In lateral
view, the proc. orbitalis meets the proc. oticus
below the dorsal margin of the cotyla quadrato-
jugalis, a result of the more robust ventral half
of the quadrate compared to NMNZ S.53209. (7)
The tuber. subcapitulare is sometimes (A. owenii)
not bound medially by a ridge extending from
the proc. orbitalis, but in other species, e.g., A.
haastii and A. australis, a ridge is present, as in
NMNZ S.53209.
While the fossil quadrate was only similar to
those of apterygids among a large range of com-
pared taxa, we present detailed comparisons with
three other palaeognath groups, the dinornithi-
forms because they were previously considered
the sister group of kiwi (CraCraft 1974) and the
tinamids and extinct lithornithids, because their
volant nature makes them or taxa in their line-
ages potential candidates for the ancestral kiwi in
New Zealand via dispersal across oceans (Houde
1988; PHilliPs et al. 2010).
Dinornithiformes. Apart from being consid-
erably larger, moa quadrates differ markedly from
those of the fossil and Apteryx as follows: proc.
orbitalis short, relatively about half the length
in kiwi, robust, dorsally convex (not laminar
in nature, elongate and concave dorsally); they
have a large foramen pneumaticum rostromedi-
ale (lacking in Apteryx); the capitulum oticum
is relatively smaller, more poorly differentiated
from the capitulum squamosum, not pneumatic
caudally; in ventral view, the articular surfaces of
the condyles differ markedly such that the artic-
ular facies of the cond. mandibularis medialis
has very little medial prominence, and so rather
than being lateromedially broad and sub-parallel
to the cond. mandibularis lateralis as in kiwi, is
craniocaudally elongate and aligned at near right
angles to the cond. mandibularis lateralis. The
A. mantelli A. rowi A. australis A. lawyri A. owenii A. haastii
Mean 48.6 45.8 47.1 45.0 48.2 45.4
Standard Error 1.08 1.14 1.47 1.34 1.04 1.13
Standard Deviation 3.42 1.98 3.90 3.00 3.28 2.53
Minimum 44.4 43.6 40.9 41.7 42.6 42.5
Maximum 53.5 47.5 50.8 49.4 54.8 49.1
Count 10 3 7 5 10 5
TABLE 1. Summary statistics for Apteryx species of % TH above cotyla quadrateojugalis from Appendix 2.
WORTHY ET AL.: Kiwi are probably not phyletic dwarves
– 71 –
Mass estimate for Proapteryx
The measured femoral circumference of the
holotype femur of Proapteryx is 12.4 mm. We
assessed body mass with several techniques.
Assessment of mass using algorithms based
on a range of birds
1. Body mass (W) in g was estimated using
the power function W = 0.16C^2.73 (anderson
et al. 1985; murray & viCkers-riCH 2004). This
equation suggests that a bird with a femoral cir-
cumference of 12.4 mm would weigh 154.6 g.
2. We also used equations from CamP-
Bell & marCus (1992) based on Group
AL (all 795 species from diverse families),
using Ordinary Least Squares regression
(OLS), the intercept is −0.065 and the slope is
2.411: thus log10W = 2.411log10C + 0.065 or
W = 1.1645C^2.411, and the estimated weight is
502 g.
3. Using the equation based on heavy-bodied
(HB) birds using Reduced Major Axis regres-
sion (RMA) from CamPBell & marCus (1992),
where the intercept is 0.11 and the slope is
2.268: thus log10W = 2.268log10C 0.11 or
W = 0.775427C^2.268, the estimated weight is
234.1 g (95 % CIs 166–337.3 g.
Assessment of mass using only palaeognaths:
Because kiwi are palaeognaths and not typical
birds and because the above values varied widely,
we assessed mass using formulae based only on
palaeognaths. Assessing the mass of ightless
ratites and the estimating the mass of extinct
forms has a sizable literature (see review in diCk-
ison 2007), however most studies found mass
estimates of Apterygidae are not well predicted
by algorithms based on other ratites, probably
because kiwi measurements lie well outside of
the data generating those equations.
4. Ratite-specic algorithms. First we
used diCkisons (2007) formula from OSL
regression of known ratite body mass on bone
measurements: W = 0.114815C^2.83. This gives a
predicted mass of 142.69 g. This estimate suffers
from being based on a data range that does not
encompass that for the fossil and further which is
biased towards large size of extant ratites and is
considerably smaller than other estimates.
5. Palaeognaths – tinamous. It is possible
Proapteryx was volant, so for this reason
pneumaticum is present in dinornithiforms; and
the cond. pterygoideus is not so prominent above
the medial side of the cond. mandibularis media-
lis. Thus, Lithornis quadrates differ substantially
and in the same way from both Apteryx quadrates
and the fossil quadrate referred to Proapteryx.
Tinamidae. The South American small vol-
ant palaeognaths in Tinamidae potentially could
be related to the bird represented by this fos-
sil as tinamous render ratites paraphyletic in
recent analyses (HarsHman et al. 2008; PHilliPs
et al. 2010) and potentially are the sister taxon to
moa (PHilliPs et al. 2010). Moreover, the oldest
tinamou fossils, clearly recognisable as similar to
modern tinamous, are of similar Early Miocene
age to the St Bathans Fauna (Bertelli & CHiaPPe
2005). However, tinamou quadrates differ mark-
edly from those of the fossil and Apteryx in having
the cotyla quadratojugalis and condyli mandibula-
ris medialis et lateralis displaced caudally relative
to the proc. oticus, and the cotyla quadratojugalis
more separated vertically from the cond. mandib-
ularis medialis, being located above the base of
the proc. orbitalis (see silveira & HöflinG 2007:
gs 40, 41). Tinamous or their recent ancestors
can thus be ruled out as being closely related to
the fossil or as ancestral to apterygids.
Assignment to Proapteryx
In summary, NMNZ S.53209 is more similar to
Apteryx than to any other palaeognath group,
which supports its referral to the Apteryx lineage.
While NMNZ S.53209 is most similar to kiwi
quadrates among known birds, the above features
2–5 are notable departures from the quadrate
form of all extant Apteryx species and support the
generic distinction based on femoral differences.
We tentatively refer NMNZ S.53209 to Proap-
teryx micromeros because it represents a kiwi of
similar size to that estimated for the holotype (see
below) and it is presently most parsimonious to
consider that only one such species is represented
in the St Bathans Fauna. With several thousand
bird bones having been collected from various
sites sourcing this fauna, it seems unlikely that the
two apterygid elements collected thus far would
belong to separate taxa. The collection of addi-
tional material will hopefully conrm this idea.
SAPE Proceedings 2013
– 72 –
of 12.4 mm would weigh 281.9 g (95 % CIs
141.2–339.5 g).
These various calculations suggest that the
predicted mass of Proapteryx probably lay within
the range of 234.1 g (95 % CIs 166–337.3 g),
using CamPBell & marCus’s (1992) equation
based on heavy-bodied birds, and 377 g (95 % CIs
307.2–463.6 g), assuming it was volant and based
on tinamou. We note that the equation based on
extant kiwi gave an intermediate value of 281.9 g
(95 % CIs 141.2–339.5 g). It seems the algorithm
based on all birds from CamPBell & marCus
(1992) probably over-estimated the weight of
Proapteryx at 502 g. Therefore, Proapteryx was
markedly smaller than all extant Apteryx species
(Appendix 4) and similar in mass to the banded
rail Gallirallus philippensis (Linnaeus).
The fossils we describe as Proapteryx micromeros
reveal, minimally, that a small apterygid species
was present in New Zealand, about 19–16 Ma.
The holotype femur derives from site FF1, an
isolated outlier of the Bannockburn Formation
(lindqvist 1994) whose stratigraphic relation-
ships to bed HH1b 9.5–9.58 m above the base
of the Bannockburn Formation in the extensive
we compared femoral diameter with mass in
the volant and similar-sized palaeognaths,
the tinamous (Tinamidae). Using the data in
Appendix 3, and Fig. 4, we computed a RMA
regression of log10W = 2.4639log10C 0.1173
thus W = 0.763308 × C ^2.4639. This equation
suggests the fossil femur with circumference of
12.4 mm was from a bird weighing 377 g (95 %
CIs 307.2–463.6 g). This predicted mass is greater
than the result from the heavy-birds algorithm of
CamPBell & marCus (1992), which is consistent
with the observation that Tinamidae include
some of the more bulky birds among those listed
as heavy birds by CamPBell & marCus (1992).
We note, however, that this estimate would only
be valid if Proapteryx was volant and of similar
proportions to tinamous.
6. Palaeognaths – apterygids. If Proapteryx
had body proportions similar to Apteryx then
an algorithm based on kiwi would be the most
accurate way of estimating its mass. We took meas-
urements of femora from 30 individuals of kiwi
of known weight in the collection of the National
Museum of New Zealand Te Papa Tongarewa
(Appendix 5) and generated a kiwi-specic algo-
rithm. The calculated RMA regression equation
for kiwi was log10W = 2.654307log10C 0.56
108 thus W = 0.274739C^2.1496. This equation
suggests that a bird with a femoral circumference
FIGURE 4. The least-squares regression of the raw data between the log circumference and log body mass in a
sample of 28 tinamous of 14 species from Appendix 3. The coefcient of determination (R2) was 0.8613.
WORTHY ET AL.: Kiwi are probably not phyletic dwarves
– 73 –
interpret the fossils, only a single taxon is rep-
resented, then this taxon was of very small size
compared to extant kiwi, being only about 0.195–
0.314 times the mass of A. owenii, the smallest
extant kiwi. We do not know whether Proapteryx
was the only apterygid present in Zealandia dur-
ing the Early Miocene, but the large sample size
of birds from the St Bathans Fauna, some 5000
specimens, makes the undiscovered presence
of another larger apterygid in this local fauna
unlikely. However, our samples for this period
derive from a single local fauna from one lacus-
trine environment in a rather large and diverse
landscape, so there is a reasonable possibility
that other forms existed elsewhere in Zealan-
dia at that time. Despite this, given Proapteryx
is an undoubted apterygid, it seems reasonable
to assume that if its morphology was similar to
that of a shared (hypothetical) common ancestor
with extant apterygids, then small size was ple-
siomorphic for the clade. If so, it then follows
that since the early Miocene, kiwi have evolved
into larger birds with proportionately larger legs.
Most terrestrial ightless birds have smaller vol-
ant relatives with proportionately smaller legs
(roff 1994; mCnaB 1994). The small size and
slenderness of the femur makes it distinctly pos-
sible that Proapteryx was volant, supporting an
overwater dispersal origin to New Zealand of
kiwi that was independent of moa (roff 1994;
CooPer et al. 2001; PHilliPs et al. 2010). Further
fossils will be required to conrm this sugges-
tion. The divergence of kiwi from casuariids, for
which estimates range from 53.5 Ma (PHilliPs et
al. 2010) to 73 Ma (HaddratH & Baker 2012)
long preceded Proapteryx leaving a large gap in
the lineage history. This is signicant because if
Proapteryx was volant then the common ances-
tor of kiwi and Australian casuariids was also
likely to have been volant, as is predicted by the
multiple loss of ight hypothesis for ratites (PHil-
liPs et al. 2010). Given that the oldest Australian
casuariid fossil presently known, Emuarius gidju
a species similar to a small emu, is about 25 Ma
(Boles 1992), then a ghost lineage in Australia
of between 25 and 50 Ma, is inferred, providing
more than enough time for the lineage to produce
ightless and large species. If Proapteryx was
ightless, a ightless lineage minimally spanning
16 Ma has to be invoked. The crown radiation of
Manuherikia River section (sCHWarzHans et
al. 2012) which lies 300 m south, are presently
undeterminable, so the two fossils may represent
two species. However, we favour conspecicity
of these fossils because they indicate a similar-
sized bird (see below) and the associated fossils
from FF1 are of species found in bed HH1b indi-
cating the same source fauna. The addition of
Proapteryx to the St Bathans Fauna reveals the
assembly of all extant iconic terrestrial vertebrates
of New Zealand, e.g., leiopelmatid frogs, Spheno-
don, moa, Aptornis, and acanthisittid wrens (see
above), was complete by the Early Miocene: not
one has arrived in the subsequent 16 Ma.
The holotype femur reveals that, at an esti-
mated 234.1 g (95 % CIs 166.0–337.3 g) – 377 g
(95 % CIs 307.2–463.6 g), Proapteryx was only
0.27–0.43 times the mass of the smallest individ-
ual (880 g) of the smallest extant kiwi species (A.
owenii), or 0.2–0.3 times the mass of the approxi-
mate modal size (1200 g) of A. owenii (Appendix
5). In contrast, the quadrate is about the size of that
in a small individual of A. owenii, and assuming it
reects skull size, might indicate that Proapteryx
and A. owenii had similar sized skulls. However,
Proapteryx has a more gracile proc. oticus, which
may indicate a shorter bill than in Apteryx. The
proportion of femur size to quadrate size seen
in Proapteryx lies intermediate between those
observed in similar-sized but distantly related
birds such as Banded Rail, Gallirallus philippen-
sis, e.g., SAM B36299, and the Australian Little
Bittern, Ixobrychus dubius (a species that has a
relatively large head), e.g., SAM B48804, (height
quadrate 7.1 mm and 7.5 mm, respectively, ver-
sus 9.2 mm; femur proximal width 8.6 mm and
5.9 mm, respectively, versus 8.4 mm; femur mid
shaft width 3.5 mm and 2.3 mm, respectively, vs
3.6 mm). Thus assuming the two bones belong
to the same species, Proapteryx had a quadrate
to femur proportion not greatly different from
the Banded Rail. The two fossils, if conspecic,
point to a bird with a head only slightly smaller
than A. owenii, but with proportionally much
smaller, more gracile legs, more like those of an
average terrestrial bird, rather than with the rela-
tively large legs modern kiwi have.
The presence of Proapteryx in the Early Mio-
cene of New Zealand places the apterygid lineage
in New Zealand at this time. If, as we prefer to
SAPE Proceedings 2013
– 74 –
during the Neogene and that their large eggs are
an evolutionary novelty resulting from develop-
ment towards extreme precociality.
We acknowledge generous support from the land-
owners A. and E. JoHnstone (Home Hills Station,
St Bathans) from whose land the fossils were
excavated. We thank Walter Boles (Australian
Museum), Wayne lonGmore (Museum Victo-
ria), Philippa Horton (South Australia Museum),
Gerald mayr (Forschungsinstitut Senckenberg,
Frankfurt am Main, Germany), Vanesa de Pietri
and Loic Costeur (Naturhistorisches Museum,
Basel, Switzerland), and Anita Gamauf (Naturhis-
torisches Museum, Vienna, Austria) for access
to comparative material. We thank Helen James
(Smithsonian Institution) for images of Lithornis.
We acknowledge the comments of Andrei zino-
viev and another anonymous reviewer that helped
to improve the text. This research supported by
the Australian Research Council (DP120100486)
and a UNSW Goldstar award (PS22963). R.P.S.
was funded by a grant from The Brian Mason
anderson, J., Hall-martin, a, & russell, d. (1985):
Long bone circumference and weight in mammals,
birds and dinosaurs. – Journal of Zoology, 207:
Baumel, J.J., kinG, a.s., Breazile, J.e., evans, H.e.&
vanden BerGe, J.C. (1993): Handbook of avian
anatomy: Nomina Anatomica Avium. 2nd edition.
Publications of the Nuttall Ornithological Club,
23: 1–779 pp.
Bertelli, s. & CHiaPPe, l.m. (2005): Earliest tina-
mous (Aves: Palaeognathae) from the Miocene of
Argentina and their phylogenetic position. – Con-
tributions in Science, Natural History Museum Los
Angeles County, 502: 1–20.
Boles, W.e. (1992): Revision of Dromaius gidju Pat-
terson and Rich, 1987, with a reassessment of its
generic position. – In: CamPBell K.E., Jr, (ed.):
Papers in avian paleontology honoring Pierce
Brodkorb. – Natural History Museum of Los Ange-
les County, Sciences Series, 36: 195–208.
Boles, W.e. (2001): A new emu (Dromaiinae) from
the Late Oligocene Etadunna Formation. – Emu,
kiwi has a recent basal divergence most recently
estimated at 6.3 Ma (2.3–1.4 Ma, 95 % CIs) in
BEAST by HaddratH & Baker (2012). There-
fore, if Proapteryx was similar to the shared
common ancestor with the crown radiation of
kiwi and if it were volant, there was minimally a
10 Ma period during which the ightless condi-
tion could have been attained and body size of
taxa increased. Even more rapid loss of ight
and attainment of similar large body size is advo-
cated for such divergent taxa as Porphyrio rails
in New Zealand over only 1 Ma (treWiCk 1997)
and for ancestral ducks becoming moa-nalos in
Hawaii in just the last 3.6 Ma (sorenson et al.
1999).The phyletic dwarng hypothesis advo-
cated by Calder (1978, 1984) and Gould (1986,
1991) explains the large size of kiwi eggs as the
result of an emu-sized ancestor evolving smaller
body size while retaining the large size of the
ancestor’s egg. Regardless of its position on the
volant–ightless continuum, Proapteryx reveals
that ancestral kiwi were probably small in the
Early Miocene. There remains no evidence in the
fossil record to support the idea that geologically
older kiwi were large like the emu and the cas-
sowary, which are the sister taxa of Apterygidae.
If this was so, then whether the ancestor of kiwi
arrived in New Zealand vicariantly or by over-
water dispersal is irrelevant; it was already small
and probably volant it cannot have been large
and ightless like modern casuariids. This rea-
soning further suggests that the Apteryx lineage
evolved larger size over the last 19–16 Ma. Thus
the super-large egg of kiwi is probably not an his-
toric holdover, but rather is more likely to be the
result of adaptive advantage. It results in a highly
precocial chick, which on emerging from the egg
is independent of adults. We contend that this fea-
ture and the nocturnal habits of kiwi are perhaps
best interpreted as adaptations to avoid avian
predators, including terrestrial Aptornis and aer-
ial raptors, both present in the Early Miocene and
subsequently (WortHy et al. 2007; WortHy et al.
2011b), to which the evolution of ightlessness
made them more vulnerable. The discovery of
Proapteryx provides strong evidence to suggest
that the origin of Apteryx by phyletic dwarng is
probably incorrect and that the opposite is instead
more likely, that kiwi are the result of an evolu-
tionary trajectory towards increased body mass
WORTHY ET AL.: Kiwi are probably not phyletic dwarves
– 75 –
dispersal of the palaeognaths, and an Early Cre-
taceous origin of modern birds. – Proceedings of
the Royal Society of London, B, 279: 4617–4625.
HarsHman, J., Braun, e.l., Braun, m.J., Huddleston,
C.J., BoWie, r.C.k., CHoJnoWski, J.l., HaCkett,
s.J., Han, k.-l., kimBall, r.t., marks, B.d,
miGlia, k.J., moore, W.s., reddy, s., sHeldon,
f.H., steadman, d.W., stePPan, s.J., Witt, C.C.,
yuri, t. (2008): Phylogenomic evidence for mul-
tiple losses of ight in ratite birds. Proceedings
of the National Academy of Sciences USA, 105:
Houde, P.W. (1988): Paleognathous birds from the
early Tertiary of the Northern Hemisphere. – Pub-
lications of the Nuttall Ornithological Club, 22:
JoHnston, P. (2011): New morphological evidence
supports congruent phylogenies and Gondwana
vicariance for palaeognathous birds. – Zoological
Journal of the Linnean Society, 163: 959–982.
Jones, m.e.H., tennyson, a.J.d., WortHy, J.P., evans,
s.e. & WortHy, t.H. (2009): A sphenodontine
(Rhynchocephalia) from the Miocene of New
Zealand and palaeobiogeography of the tuatara
(Sphenodon) – Proceedings of the Royal Society of
London, B, 276: 1385–1390.
lee, m.s.y., HutCHinson, m.n., WortHy, t.H.,
arCHer, m., tennyson, a.J.d., WortHy, J.P. &
sCofield, r.P. (2009): Miocene skinks and geckos
reveal long-term conservatism of New Zealand’s
lizard fauna – Biology Letters, 5: 833–837.
lindqvist, J.k. (1994): Lacustrine stromatolites and
oncoids: Manuherikia Group (Miocene), New
Zealand. – In: Bertrand-sarfati, J. & monty, C.
(eds): Phanerozoic Stromatolites II. – pp. 227–254,
Dordrecht (Kluwer Academic Publishers).
marCHant, s. & HiGGins, P.J. (co-ordinators) (1990):
Handbook of Australian, New Zealand and Antarc-
tic Birds. Volume 1, Ratites to Ducks. – 1400 pp.
Melbourne (Oxford Univ. Press).
mCGoWan, C. (1979): The hind limb musculature of
the Brown kiwi Apteryx australis mantelli – Jour-
nal of Morphology, 160: 33–74.
mCnaB, B.k. (1994): Energy-conservation and the
evolution of ightlessness in birds. – American
Naturalist, 144: 628–642.
murray, P.f. & viCkers-riCH, P. (2004): Magnicent
mihirungs. The colossal ightless birds of the
Australian Dreamtime. – 416 pp. Bloomington
(Indiana University Press).
Patak, a.e. & BaldWin, J. (1998): Pelvic limb mus-
culature in the Emu Dromaius novaehollandiae
(Aves: Struthioniformes: Dromaiidae): adaptations
to high-speed running. – Journal of Morphology,
238: 23–37.
PHilliPs, m.J., GiBB, G.C., CrimP, e.a. & Penny, d.
(2010): Tinamous and moa ock together: mito-
101: 317–321.
Calder, W.a. (1978): The Kiwi. – Scientic Ameri-
can, 239/1: 132–142.
Calder, W.a. (1984): Size Function and Life His-
tory. – 448 pp. Cambridge (Harvard Univ. Press).
CamPBell, k.e. & marCus, l. (1992): The relation-
ship of hindlimb bone dimensions to body weight
in birds. – In: CamPBell, k.e. (ed.): Papers in avian
paleontology honoring Pierce Brodkorb. – Natural
History Museum of Los Angeles County, Science
Series 36: 395–412.
CooPer, a., lalueza-fox, C., anderson, s., ramBaut,
a., austin, J. & Ward, r. (2001): Complete mito-
chondrial genome sequences of two extinct moas
clarify ratite evolution. – Nature, 409: 704–707.
CraCraft, J. (1974): Phylogeny and evolution of the
ratite birds. – Ibis, 116: 494–521.
davies, s.J.J.f. (2002): Ratites and Tinamous:
Tinamidae, Rheidae, Dromaiidae, Casuariidae,
Apterygidae, Struthionidae (Bird Families of the
World). – 360 pp. Oxford (Oxford University
de queiroz, s.a. & CooPer, r.G. (2011): Gender-
based differences in stride and limb dimensions
between healthy red-wing tinamou (Rhynchotus
rufescens) Temminck, 1815. – Turkish Journal of
Zoology, 35: 103–108.
diCkison, m.J. (2007): Allometry of giant ightless
birds. Unpublished PhD thesis. – 114 pp. Durham,
USA (Duke University).
elzanoWski, a. & stidHam, t.a. (2010): Morphology
of the quadrate in the Eocene anseriform Presby-
ornis and the extant Galloanserine birds. – Journal
of Morphology, 271: 305–323.
Gaina, C., roest, W.r., müller, r.d. & symonds, P.
(1998): The opening of the Tasman Sea: a gravity
anomaly animation. – Earth Interactions, 2–004:
Gill, B.J., Bell, B.d., CHamBers, G.k., medWay,
d.G., Palma, r.l., sCofield, r.P., tennyson,
a.J.d. & WortHy, t.H. (2010): Checklist of the
Birds of New Zealand, Norfolk and Macquarie
Islands, and the Ross Dependency, Antarctica.
4th Edition. – 500 pp. Wellington (Ornithological
Society of NZ & Te Papa Press).
Gould, s.J. (1986): Of Kiwi eggs and the Liberty
Bell. – Natural History, 95(11): 20–29.
Gould, s. J. (1991): Of Kiwi Eggs and the Liberty
Bell. – In: Gould, s.J. (ed.): Bully for Brontosau-
rus. – pp. 109–123, New York (W. Norton).
HaddratH, o. & Baker, a.J. (2001): Complete mito-
chondrial DNA genome sequences of extinct birds:
Ratite phylogenetics and the vicariance biogeogra-
phy hypothesis. – Proceedings of the Royal Society
of London, B, 268: 939–945.
HaddratH, o. & Baker, a.J. (2012): Multiple nuclear
genes and retroposons support vicariance and
SAPE Proceedings 2013
– 76 –
cal Club, 23: 189–247.
WortHy, t.H., Hand, s.J., nGuyen, J.m.t., tennyson,
a.J.d., WortHy, J.P., sCofield, r.P., Boles, W.e.
& arCHer, m. (2010a): Biogeographical and phy-
logenetic implications of an early Miocene wren
(Aves: Passeriformes: Acanthisittidae) from New
Zealand. – Journal of Vertebrate Paleontology, 30:
WortHy, t.H., Hand, s.J., WortHy, J.P., tennyson,
a.J.d. & sCofield, r.P. (2009): A large fruit
pigeon (Columbidae) from the early Miocene of
New Zealand. – The Auk, 126: 649–656.
WortHy, t.H. & sCofield, r.P. (2012): Twenty-rst
century advances in knowledge of the biology of
moa (Aves: Dinornithiformes): a new morphologi-
cal analysis and diagnoses revised. – New Zealand
Journal of Zoology, 39: 87–153.
WortHy, t.H., tennyson, a.J.d., arCHer, m., musser,
a.m., Hand, s.J., Jones, C., douGlas, B.J., mCna-
mara, J.a. & BeCk, r.m.d. (2006): Miocene
mammal reveals a Mesozoic ghost lineage on insu-
lar New Zealand, southwest Pacic. – Proceedings
of the National Academy of Sciences USA, 103:
WortHy, t.H., tennyson, a.J.d., arCHer, m. & sCo-
field, r.P. (2010b): First record of Palaelodus
(Aves: Phoenicopteriformes) from New Zealand. –
Records of the Australian Museum, 62: 77–88.
WortHy, t.H., tennyson, a.J.d., Hand, s.J., God-
tHelP, H. & sCofield, r.P. (2011a): Terrestrial
turtle fossils from New Zealand reoat Moa’s
Ark. – Copeia, 2011: 72–76.
WortHy, t.H., tennyson, a.J.d., Jones, C., mCna-
mara, J.a. & douGlas B.J. (2007): Miocene
waterfowl and other birds from Central Otago,
New Zealand. – Journal of Systematic Palaeontol-
ogy, 5: 1–39.
WortHy, t.H., tennyson, a.J.d. & sCofield, r.P.
(2011c): Fossils reveal an early Miocene pres-
ence of the aberrant gruiform Aves: Aptornithidae
in New Zealand. – Journal of Ornithology, 152:
WortHy, t.H., tennyson, a.J.d. & sCofield, r.P.
(2011d): An Early Miocene diversity of parrots
(Aves, Strigopidae, Nestorinae) from New Zea-
land. – Journal of Vertebrate Paleontology, 31:
WortHy, t.H., WortHy, J.P., arCHer, m., Hand, s.J.,
sCofield, r.P., marsHall, B.a. & tennyson,
a.J.d. (2011b): A decade on, what the St Bathans
Fauna reveals about the Early Miocene terrestrial
biota of Zealandia. – In: litCHfield, n.J. & Clark,
k. (eds), Abstract volume, Geosciences 2011
Conference, Nelson, New Zealand. – Geoscience
Society of New Zealand Miscellaneous Publica-
tion, 130A: 120.
chondrial genome sequence analysis reveals
independent losses of ight among ratites. Sys-
tematic Biology, 59: 90–107.
roff, d.a. (1994): The evolution of ightlessness:
is history important? – Evolutionary Ecology, 8:
sCHellart, W.P., lister, G.s. & toy, v.G. (2006): A
Late Cretaceous and Cenozoic reconstruction of
the Southwest Pacic region: tectonics controlled
by subduction and slab rollback processes. – Earth
Science Revues, 76: 191–233.
sCHWarzHans, W., sCofield, r.P., tennyson, a.J.d.,
WortHy, J.P. & WortHy, t.H. (2012): Fish
remains, mostly otoliths, from the non-marine
Early Miocene of Otago, New Zealand. – Acta
Palaeontologica Polonica, 57: 319–350.
sCofield, r.P., WortHy, t.H. & tennyson, a.J.d.
(2010): A heron (Aves: Ardeidae) from the Early
Miocene St Bathans Fauna of southern New Zea-
land. – Records of the Australian Museum, 62:
silveira, l.f. & HöflinG, e. (2007): Osteologia cra-
niana dos Tinamidae (Aves: Tinamiformes), com
implicações sistemáticas. – Boletim do Museu Par-
aense Emílio Goeldi Ciências Naturais, Belém, 2:
sorenson, m.d., CooPer, a., Paxinos, e.e., quinn,
t.W., James, H.f., olson, s.l., fleisCHer, r.C.
(1999): Relationships of the extinct moa-nalos,
ightless Hawaiian waterfowl, based on ancient
DNA. – Proceedings of the Royal Society of Lon-
don, B, 266: 2187–2193.
tennyson, a.J.d. (2010): The origin and history of
New Zealand’s terrestrial vertebrates. – New Zea-
land Journal of Ecology, 34: 6–27
tennyson, a.J.d., Palma, r.l., roBertson, H., Wor-
tHy, t.H., & Gill B.J. (2003): A new species of
kiwi (Aves, Apterygiformes) from Okarito, New
Zealand. – Records of the Auckland Museum, 40:
tennyson, a.J.d., WortHy, t.H., Jones, C.m., sCo-
field, r.P. & Hand, s.J. (2010): Moa’s Ark:
Miocene fossils reveal the great antiquity of moa
(Aves: Dinornithiformes) in Zealandia. – Records
of the Australian Museum, 62: 105–114.
treWiCk, s.a. (1997): Flightlessness and phylogeny
amongst endemic rails (Aves: Rallidae) of the New
Zealand region. – Philosophical Transactions:
Biological Sciences, 352: 429–446.
vanden BerGe, J.C. (1982): Notes on the myology
of the pelvic limb in kiwi (Apteryx) and in other
birds. – The Auk, 99: 309–315.
vanden BerGe, J.C. & sWeers, G.a. (1993): Myolo-
gia. – In: Baumel, J.J., kinG, a.s., Breazile, J.e.,
evans, H.e. & vanden BerGe, C. (eds): Handbook
of avian anatomy: Nomina anatomica avium, 2nd
edition. – Publications of the Nuttall Ornithologi-
WORTHY ET AL.: Kiwi are probably not phyletic dwarves
– 77 –
All the following taxa lack a marked notch
proximally separating the caput from facies artic. anti-
trochanterica, have a well-marked impression for m.
obturatorius lateralis, and are further distinguished as
Gruiformes (Gruiidae, Otididae, Rhynochetos
jubatus, and Rallidae). Proportions similar, but have
a fossa trochanteris; and lateral facies is markedly con-
vex adjacent to the crista trochanteris.
Ardeidae. Proportions similar, but the shaft lacks
dorsal curvature; crista trochanteris craniocaudal
depth is much shallower and proximally is lateromedi-
ally narrow, so a broad at groove connects the cranial
surface to the facies artic. antitrochanterica.
Anseriforms. Femoral shaft relatively much
shorter; caput on shorter neck.
Podicipedidae, Spheniscidae, Procellari-
iformes, Anhingidae, Phalacrocoracidae, Sulidae,
Phaethontidae. Femoral shaft much shorter; cranio-
caudal depth crista trochanteris subequal or only
slightly deeper than caput depth; neck with little or no
Pelecanidae. Femoral shaft relatively shorter, thin-
walled, no dorsal curvature, crista trochanteris shorter,
arrangement of insertion areas laterally differs.
Threskiornithidae. Femoral shaft relatively
shorter, no dorsal curvature, crista trochanteris rela-
tively short (proximodistally) with little cranial
Ciconiidae, Phoenicopteridae, Accipitri-
formes, Cathartidae and Falconidae. Femora as
for Theskiornithids, but with pneumatic foramina in
cranial pretrochanteric area and linea intermuscularis
cranialis extends mesad of crista trochanteris.
Charadriiformes. Shaft lacks dorsal curvature;
fossa trochanteris present. In addition, in Haemato-
podidae, Recurvirostridae, Charadriidae, Laridae, and
Glareolidae, linea intermuscularis cranialis extends
mesad of and parallel to crista trochanteris.
Columbidae, Psittaciformes, Caprimulgi-
formes, Cuculidae, Strigiformes, Coraciiformes
and Passeriformes. Crista trochanteris cranially low,
often with craniocaudal depth subequal to caput depth.
Columbidae further differ with a fossa trochanteris
and Caprimulgiformes and Strigiformes have crista
trochanteris proximodistally shorter, and passed medi-
ally by well-marked crista intermuscularis cranialis.
Appendix 1
Femoral features distinguishing Apterygidae from
other birds.
Other modern birds, with an emphasis on those
likely to be related for reasons of geographic proxim-
ity, have femora that differ from those of apterygids by
the following features, which are considered sufcient
to distinguish them, but which are not intended to be a
comprehensive list of differences.
Dinornithiformes. All species of moa are vastly
larger; facies artic. antitrochantica concave; pattern
of ligament insertions proximolaterally differ greatly
as follows (terminology after mCGoWan (1979) with
preferred synonyms from vanden BerGe (1982) and
vanden BerGe & sWeers (1993) in brackets, although
we note mCGoWan’s caveat that size and the pres-
ence of few muscles in Apteryx can be correlated
with impressions on bones): large elongate sulcus
for insertions for m. iliotrochantericus posterior (m.
iliotrochantericus caudalis), m. gluteus medius et
minimus (m. iliofemoralis externus), and m. iliotro-
chantericus medius et anterior (mm. iliotrochanterici
medius et cranialis) shallower, located well craniad,
nearly adjacent to cranial margin of crista trochanteris;
insertion area for m. ischiofemoralis broad, well sepa-
rated from the latter, just caudad of centre; impression
for m. obturator internus (m. obturatorius medialis)
closer to caudal margin, small, just caudad of the
insertion area for m. ischiofemoralis; impression for
M. obturator externus (m. obturatorius lateralis), shal-
low, broad.
Casuariidae. Similarities include lack of fossa
trochanteris, notably convex facies artic. antitro-
chantica separated from caput by marked notch and
ligamental insertion for m. iliotrochantericus cauda-
lis largest and at mid craniocaudal depth. Differences
include markedly larger size and a distinct pattern
of sulci for ligamental insertions proximolaterally,
interpreted by Patak & BaldWin (1998) as follows:
a large oval sulcus at mid-craniocaudal depth for m.
iliofemoralis externus, but the bipartite appearance of
this sulcus suggests it also houses the insertion for m.
iliotrochantericus caudalis; distal to this sulcus lies a
circular and deep impression for m. ischiofemoralis;
impression for m. obturator externus (=m. obturatorius
lateralis), shallow, relatively small, stronger-marked in
Tinamidae, Galliformes. Proportions similar, but
differ with a well-marked fossa trochanteris; facies
artic. antitrochantica concave; impressio m. iliotro-
chantericus caudalis shallower and more cranially
located; insertion area of m. obturatorius lateralis well
marked; proximally, caudal and lateral facies meet in
even curve; facies artic. antitrochanterica connected
via broad groove to cranial surface.
SAPE Proceedings 2013
– 78 –
Catalogue number Taxon TH (capitulum
squamosum – base of
cond. medialis)
TH above cotyla
% TH above cotyla
DM.492-S australis 11.1 5.6 50.5
OR.4738 australis 11.0 4.5 40.9
OR.21035 australis 10.5 5.3 50.5
OR.22114 australis 11.6 5.5 47.4
OR.22115 australis 11.9 5.5 46.2
OR.27965 australis 12.0 6.1 50.8
CM Av 39065 australis 11.5 5.0 43.2
OR.21415 haastii 11.3 5.1 45.1
OR.23045 haastii 11.4 5.6 49.1
OR.27983 haastii 12.3 5.7 46.3
MV B40905 haastii 12.1 5.3 43.8
CM Av31538 haastii 12.4 5.3 42.5
OR.21832 lawryi 12.1 5.3 43.8
CM Av 14447 lawryi 11.7 4.9 41.7
CM Av 32404 lawryi 13.5 6.7 49.4
CM Av 36637 lawryi 13.0 5.7 43.6
CM Av 36638 lawryi 12.0 5.6 46.4
DM.909-S mantelli 11.1 5.9 53.2
OR.23048 mantelli 11.1 5.4 48.6
OR.24640 mantelli 10.1 5.4 53.5
OR.24984 mantelli 11.3 5.8 51.3
OR.27604 mantelli 11.3 5.2 46.0
OR.28614 mantelli 11.3 5.1 45.1
OR.28615 mantelli 11.7 5.2 44.4
OR.28616 mantelli 10.8 5.5 50.9
OR.29374 mantelli 11.3 5.1 45.1
CM Av 5492 mantelli 10.8 5.2 48.1
OR.20990 owenii 9.6 4.6 47.9
OR.22369 owenii 9.3 4.8 51.6
OR.23214 owenii 9.3 5.1 54.8
OR.24414 owenii 8.4 4.0 47.6
OR.24415 owenii 8.9 4.4 49.4
OR.24416 owenii 9.0 4.3 47.8
OR.25100 owenii 10.0 4.8 48.0
OR.25794 owenii 9.4 4.0 42.6
MV B56009 owenii 10.1 4.7 46.2
SAM B5051 owenii 9.3 4.3 46.5
CM Av 16691 rowi 12.2 5.3 43.6
CM Av 16717 rowi 11.0 5.2 47.5
CM Av 16718 rowi 12.3 5.7 46.2
USNM 290601 Lithornis celetius 10.8 7.6 70.7
S.53209 Proapteryx 9.2 4.9 53.3
Appendix 2
Measurements (mm) of quadrates of Apteryx species. TH is total height. Catalogue numbers starting
with OR, DM, and S are all prexed by NMNZ.
WORTHY ET AL.: Kiwi are probably not phyletic dwarves
– 79 –
Reg. No. Taxa SW SD C Weight log C log mass
SAM B.31339 Tinamus major robusta 5.6 5.8 17.9 1140 1.253092 3.056905
NMNZ OR.1433 Tinamus major major 4.7 5.5 16.1 1028.5 1.206052 3.012204
NMNZ OR.22983 Nothoprocta perdicaria 3.9 4.5 13.2 458 1.121504 2.660865
NMW 4.559 Tinamus major 6.1 6.2 19.3 1140 1.286039 3.056905
NMW 1061 Nothura maculosa 4.2 4.0 12.9 300 1.110063 2.477121
NMW 3.071 Eudromia elegans 5.5 5.0 16.5 703.5 1.217801 2.847264
NMW 160 Rhynchotus fasciatus = rufescens 5.6 5.7 17.8 770 1.249215 2.886491
NMW 161 Rhynchotus fasciatus = rufescens 6.1 5.9 18.9 770 1.275361 2.886491
NMW 4.068 Nothoprocta perdicaria 4.6 4.1 13.7 458 1.136355 2.660865
NMB 5537 Rhynchotus rufescens 5.8 5.3 17.3 770 1.238748 2.886491
NMB C.2004 Tinamus (Trachypilmus Cab.) robustus 5.9 5.8 18.4 1140 1.26507 3.056905
SMF 2148 Crypturellus obsoletus 4.1 3.7 12.3 480 1.089952 2.681241
SMF 11394 Crypturellus noctivagus noctivagus 4.2 4.2 13.2 800 1.119375 2.90309
SMF 11392 Crypturellus tataupa 3.4 3.8 11.3 202 1.053995 2.305351
SMF 2164 Crypturellus parvirostris 3.0 2.8 9.0 180 0.955935 2.255273
SMF 8184 Crypturellus parvirostris 3.5 3.1 10.4 220 1.015069 2.342423
SMF 9357 Crypturellus parvirostris 3.4 3.1 10.2 220 1.0081 2.342423
SMF 2537 Crypturellus cinnamomeus 3.9 4.0 12.4 419 1.092696 2.622214
SMF 2149 Crypturellus undulatus vermiculatus 4.4 4.6 14.2 540 1.151456 2.732394
SMF 2146 Tinamus solitarius 5.7 5.9 18.2 1200 1.260642 3.079181
SMF 2150 Tinamus solitarius 6.5 6.9 21.1 1500 1.324752 3.176091
SMF 5415 Eudromia elegans 5.5 4.9 16.5 703.5 1.216318 2.847264
SMF 6111 Eudromia elegans 5.8 5.1 17.1 703.5 1.233381 2.847264
SMF 6416 Eudromia elegans 5.6 5.0 16.7 703.5 1.222983 2.847264
SMF 9260 Eudromia elegans 5.6 5.1 16.8 703.5 1.225977 2.847264
SMF 9306 Eudromia elegans 5.0 4.5 14.9 703.5 1.172044 2.847264
SMF 6298 Eudromia elegans 5.4 5.4 16.9 703.5 1.227935 2.847264
SMF 2147 Rhynchotus rufescens 5.2 5.2 16.3 747 1.213153 2.873321
Species Males Females
Apteryx owenii 880–1356 1000–1400
Apteryx mantelli 1820–2590 2090–3270
Apteryx rowi 1575–2250 1950–3570
Apteryx australis lawryi 2300–3060 2700–3600
Apteryx haastii 1215–2320 1530–2718
Appendix 3
Relationship of Tinamidae estimated femoral circumference (using the formula circumference = PI*
SQRT(2*((POWER((1/2*SD),2)) + (POWER((1/2*SW),2)))) to average weight. SW is width at mid
shaft, SD is shaft depth at mid shaft, C is circumference, measurements in mm. Weights were taken
from davies (2002) except for SMF 2147, which was from queiroz & CooPer (2011).
Appendix 4
The range in mass values (grams) by sex for Apteryx species from marCHant & HiGGins (1990) and
for A. rowi from tennyson et al. (2003).
SAPE Proceedings 2013
– 80 –
Appendix 5
Relationship of Apteryx femoral circumference to weight of individual recorded at death. (SW is width
at mid shaft, SD is shaft depth at mid shaft, circumference is estimated using the formulae PI*SQRT
(2*((POWER((1/2*SD),2)) + (POWER((1/2*SW),2)))). All specimens are from the Museum of New
Zealand Te Papa Tongarewa.
Taxon Reg no. SD SW Circumference weight (g)
A. mantelli DM.909-S 9.6 8.5 25.1607 785
A. mantelli OR.13588 11 11 29.6656 2428
A. mantelli OR.14964 10 9 27.0852 1988
A. mantelli OR.17206 10 9.6 27.0332 1366
A. mantelli OR.17207 11 10 28.4908 2563
A. mantelli OR.17208 9.8 9.2 26.13 1729
A. mantelli OR.17209 11 10 28.1212 2336
A. mantelli OR.17210 12 11 30.7656 2919
A. mantelli OR.17211 12 12 32.0138 2495
A. mantelli OR.17212 11 11 29.9376 2526
A. mantelli OR.17213 10 8.9 26.4356 930
A. mantelli OR.24640 9.1 8.6 24.3134 815
A. mantelli OR.24984 9 9.5 24.948 1055
A. a. australis OR.22089a 14 12 36.9381 2488
A. a. australis OR.27761a 13 12 33.6873 3400
A. a. australis OR.27965 11 10 28.842 2200
A. a. lawryi OR.23591 11 11 28.8394 2800
A. a. lawryi OR.23756 12 11 31.1584 4335
A. owenii OR.22815a 8.4 8 22.497 910
A. owenii OR.23044 7.8 7.4 20.865 885
A. owenii OR.23717a 7.7 7.4 20.6809 1285
A. owenii OR.24415 6.8 6.9 18.5919 670
A. owenii OR.24416 7.8 7.8 21.2215 875
A. haastii OR.19773a 9.6 8.5 25.1607 1215
A. haastii OR.23022a 12 12 32.6507 2232
A. haastii OR.23038 9.5 9.5 25.8466 1890
A. haastii OR.23045 12 10 29.9895 2435
A. haastii OR.23648a 11 11 29.8414 2015
A. haastii OR.27983 9.6 9.2 25.7611 2550
A. haastii OR.28010a 12 11 31.3853 2843
... Here, we describe a new species of Mystacina from the early Miocene Ma) St Bathans Fauna of Central Otago, South Island (Fig 1). The St Bathans deposit, which consists of near-shore freshwater lacustrine sediments of Miocene Lake Manuherikia, has also produced New Zealand's oldest frogs, lizards, sphenodontine, land birds (including kiwi, moa and parrots), its only crocodilian, terrestrial turtle and a non-volant mammal [21][22][23][24][25][26][27][28][29]. We also review the fossil record for plants that are pollinated, dispersed, eaten or used as roosts by modern Mystacina, as well as their arthropod prey, and note potentially deep-rooted ecological associations between New Zealand's endemic bats, vegetation and arthropod prey. ...
... It confirms that Mystacina has had a long presence in New Zealand, in common with other genera of terrestrial vertebrates in the St Bathans Fauna including frogs (e.g. Leiopelma), lizards (Hoplodactylus, Oligosoma) and some birds (Aptornis, Pelecanoides, Aegotheles) [25,28,29,48]. The family Mystacinidae itself is an ancient lineage that molecular data indicate diverged from other bats c.51-41 Ma [49][50][51][52]. ...
Full-text available
The New Zealand endemic bat family Mystacinidae comprises just two Recent species referred to a single genus, Mystacina. The family was once more diverse and widespread, with an additional six extinct taxa recorded from Australia and New Zealand. Here, a new mystacinid is described from the early Miocene (19-16 Ma) St Bathans Fauna of Central Otago, South Island, New Zealand. It is the first pre-Pleistocene record of the modern genus and it extends the evolutionary history of Mystacina back at least 16 million years. Extant Mystacina species occupy old-growth rainforest and are semi-terrestrial with an exceptionally broad omnivorous diet. The majority of the plants inhabited, pollinated, dispersed or eaten by modern Mystacina were well-established in southern New Zealand in the early Miocene, based on the fossil record from sites at or near where the bat fossils are found. Similarly, many of the arthropod prey of living Mystacina are recorded as fossils in the same area. Although none of the Miocene plant and arthropod species is extant, most are closely related to modern taxa, demonstrating potentially long-standing ecological associations with Mystacina.
... The closest relatives of kiwi are Madagascar's extinct elephant birds (Phillips et al. 2010;Mitchell et al. 2014), with the split between their ancestors thought to be about 50 million years ago, which supports the suggestion that the ancestors of kiwi flew to Zealandia and subsequently evolved flightlessness. Further evidence of ancestral kiwi flying to Zealandia was provided by kiwi fossils from the Early Miocene (16-19 million years ago) of New Zealand, which represent much smaller and possibly volant birds (Worthy et al. 2013). Subsequent studies have supported the sister relationship between kiwi and elephant birds but have suggested that the split between these taxa occurred about 54 or 62 million years ago (Grealy et al. 2017;Yonezawa et al. 2017), again raising the possibility that ancestral kiwi did not need to reach Zealandia by flight (Tennyson 2010). ...
We describe Apteryx littoralis sp. nov., a new species of kiwi based on a 1-million-year-old tarsometatarsus from shallow marine sediment in the North Island of New Zealand. The fossil is very similar to the tarsometatarsi of living kiwi species, most closely resembling Apteryx rowi and A. mantelli in size and shape, but differs in being stouter, with proportionally narrower proximal and distal ends. The new fossil is the second oldest known record of kiwi. It demonstrates a relatively conservative kiwi morphology since the mid-Pleistocene.
... Rather than evolving from a large flightless bird, as had previously been thought, this suggests that kiwi were always small and had never become large and herbivorous (like other flightless ratites) because that niche had already been occupied in New Zealand by moa ). The hypothesis is supported by the presence of small fossil kiwi bones ( Worthy et al. 2013) and contemporaneous large moa bones and eggshell ) from the Early Miocene of New Zealand. ...
In the 25 years since the first DNA sequences were obtained from the extinct moa, ancient DNA analyses have significantly advanced our understanding of New Zealand's unique fauna. Here, we review how DNA extracted from ancient faunal remains has provided new insights into the evolutionary histories and phylogenetic relationships of New Zealand animals, and the impacts of human activities upon their populations. Moreover, we review how ancient DNA has played a key role in improving our ability to taxonomically identify fragmentary animal remains, determine biological function within extinct species, reconstruct past faunas and communities based on DNA preserved in sediments, resolve aspects of the ecology of extinct animals and characterising prehistoric parasite faunas. As ancient DNA analyses continue to become increasingly applied, and sequencing technologies continue to improve, the next 25 years promises to provide many more exciting new insights and discoveries about New Zealand's unique fauna.
... The abundant avifauna is dominated by waterfowl (Anseriformes), with a minimum of eight taxa in five genera (Worthy et al. 2007(Worthy et al. , 2008. The avifauna also includes such diverse taxa as moas (Dinornithiformes), a kiwi (Apterygidae), a tubenose (Procellariiformes), birds of prey (Accipitriformes), rails (Rallidae), an endemic gruiform (Aptornithidae), a gull and other waders (Charadriiformes), herons (Ardeidae), a palaelodid (Phoenicopteriformes), pigeons (Columbidae), parrots (Psittaciformes), a swiftlet (Apodidae), an owlet-nightjar (Aegothelidae), and passerines (Passeriformes) (De Pietri et al. 2016a, 2016bWorthy et al. 2007Worthy et al. , 2009aWorthy et al. , 2009bWorthy et al. , 2010aWorthy et al. , 2010bWorthy et al. , 2011aWorthy et al. , 2011bWorthy et al. , 2011cWorthy et al. , 2013aWorthy et al. , 2013bWorthy et al. , 2013cScofield et al. 2010). Fish, frogs, reptiles and mammals are also represented (Worthy et al. 2006(Worthy et al. , 2011d(Worthy et al. , 2013dJones et al. 2009;Lee et al. 2009;Schwarzhans et al. 2012;Hand et al. 2013Hand et al. , 2015. ...
Full-text available
Eight species of terrestrial Mollusca are recorded from Early–Middle Miocene sediments from palaeolake Manuherikia, near St Bathans, central Otago, New Zealand. Five new charopid species are described in Cavellia Iredale, 1915, Charopa Martens, 1860 and Fectola Iredale, 1915, a new genus Dendropa, based on the Recent species Flammulina pilsbryi Suter, 1894, Neophenacohelix Cumber, 1961, which is resurrected from synonymy under Phenacohelix Suter, 1892, and two new species of punctid are described in Paralaoma Iredale, 1913 and Atactolaoma n. gen. All genera involved are endemic to New Zealand and are the first pre-Quaternary records. A rhytidid is also recorded, which, though indeterminable, is the earliest record of the family. The three other confirmed pre-Quaternary (Late Pliocene) records of land snails are briefly discussed.
... One of the most striking discoveries since 2010 is that of a small kiwi (Apterygidae) in the St Bathans Fauna, Proapteryx micromeros Worthy TH, Worthy JP, Tennyson, Salisbury, Hand and Scofield, 2013. It was a tiny bird, about a third the size of the Little-spotted Kiwi (Apteryx oweni) and is rare, with just a femur and a quadrate known to date (Worthy et al. 2013b). Given its very small size, the authors speculated that this, the oldest member of the kiwi lineage, could have been volant and if indeed a representative of ancestral kiwi, it showed that Gould's hypothesis (Gould 1986(Gould , 1991) that the kiwi ancestor was a moa-sized taxon that had dwarfed over time was wrong. ...
Full-text available
New Zealand, long recognised as a land where birds dominate the terrestrial vertebrate biota, lacked an informative fossil record for the non-marine pre-Pleistocene avifauna until the twenty-first century. Here we review recent research that alters the known diversity of the fossil Paleogene–Neogene birds and our understanding of the origin of New Zealand’s recent or modern biota. Since 2010, there has been a 50% increase in the number of described fossil bird species (now 45) for the pre-Quaternary period. Many represent higher taxa that are new or listed for New Zealand for the first time, including 12 genera (35 total), nine family-level taxa (18 total), and seven ordinal taxa. We also review recent multidisciplinary research integrating DNA and morphological analyses affecting the taxonomic diversity of the Quaternary avifauna and present revised diversity metrics. The Holocene avifauna contained 217 indigenous breeding species (67% endemic) of which 54 (25%) are extinct.
... The fossil record of other ratites groups is otherwise mostly incomplete: Struthio coppensi, the oldest known fossil in the genus Struthio , is dated from the Lower Miocene, and most fossil Apterygidae and Dinornithidae are found only in quaternary deposits (Worthy & Holdaway, 2002), although some Miocene kiwi fossils have recently been described (Worthy et al., 2013). Oldest fossil tinamous are 17.5-16.5 ...
Full-text available
Archosaurs are a clade of vertebrates that includes birds, crocodiles, and numerous fossil groups. This clade has been a matter of debate among paleontologists for decades concerning the evolution of thermometabolism in its different lineages. The classical hypothesis considers that only modern birds are true endotherms, whereas all other archosaurs are ectotherms. Bone histology allows to study several traits linked to bone growth rate and thermometabolism, otherwise impossible to estimate on fossil specimens; for this reason, we used characters measured on long bone histological sections.In the first section, we extensively reviewed the measure of phylogenetic signal for osteohistological features in two clades of vertebrates, which was then used to define the methodology for building our predictive models.After a preliminary study during which we built a predictive model for bone growth rate, we built a global model to predict the metabolic rate of our fossil specimens, using both histological features and phylogenetic information for each specimen. Our results show that a majority of archosaurs in our sample were endotherms. This implies that the last common ancestor of archosaurs was likely an endotherm, and that modern crocodiles became secondarily ectothermic, probably in response to their aquatic environment. More specific studies on pseudosuchians should allow to precisely identify the level of the phylogenetic tree at which the ectothermic state was acquired, as well as adaptive constraints behind this acquisition.
Eight species of terrestrial Mollusca are recorded from Early–Middle Miocene sediments from palaeolake Manuherikia, near St Bathans, central Otago, New Zealand. Five new charopid species are described in Cavellia Iredale, 1915, Charopa Martens, 1860 and Fectola Iredale, 1915, a new genus Dendropa, based on the Recent species Flammulina pilsbryi Suter, 1894, Neophenacohelix Cumber, 1961, which is resurrected from synonymy under Phenacohelix Suter, 1892, and two new species of punctid are described in Paralaoma Iredale, 1913 and Atactolaoma n. gen. All genera involved are endemic to New Zealand and are the first pre-Quaternary records. A rhytidid is also recorded, which, though indeterminable, is the earliest record of the family. The three other confirmed pre-Quaternary (Late Pliocene) records of land snails are briefly discussed.
The fossil record of squamates in New Zealand is scant, and this chapter represents the first systematic review of the available information for this reptile group in New Zealand. The oldest fossil squamates are found in the Early Miocene (19–16 mya) St Bathans Fauna. The material represents skinks referred to Eugongylinae similar to extant species of Oligosoma and geckos referred to Diplodactylidae, which differ little from extant New Zealand geckos. No other squamates are represented in the St Bathans Fauna. The Early Miocene St Bathans skinks and geckos formed part of a fauna that was similar to the recent prehuman New Zealand fauna as it was dominated by birds and included sphenodontids and leiopelmatid frogs, but differed markedly by the additional presence of a crocodilian, terrestrial turtles and greater mammalian diversity. A squamate fossil record is then unknown until the last 50,000 years of the late Quaternary. This late Pleistocene to Holocene fauna, however, documents the natural, undisrupted biota that was encountered and decimated by humans and the species they introduced. Fossil squamates are relatively common, but little studied. Notably, they document the former widespread presence of a suite of large forms. Among skinks, these include two extinct taxa in the Northland region of North Island, including Oligosoma northlandi, the largest skink known from New Zealand. In Northland and elsewhere on the North Island, fossils attest to the more widespread presence of Oligosoma alani, O. whitakeri, O. macgregori and O. oliveri. In both North and South Islands, Hoplodactylus duvaucelii was widespread. The available information for the kawekaweau (Hoplodactylus delcourti) is reviewed, and the lack of fossil evidence for it is discussed.
Molecular dating largely overturned the paradigm that global cooling during recent Pleistocene glacial cycles resulted in a burst of species diversification although some evidence exists that speciation was commonly promoted in habitats near the expanding and retracting ice sheets. Here, we used a genome-wide dataset of more than half a million base pairs of DNA to test for a glacially induced burst of diversification in kiwi, an avian family distributed within several hundred kilometers of the expanding and retracting glaciers of the Southern Alps of New Zealand. By sampling across the geographic range of the five kiwi species, we discovered many cryptic lineages, bringing the total number of kiwi taxa that currently exist to 11 and the number that existed just before human arrival to 16 or 17. We found that 80% of kiwi diversification events date to the major glacial advances of the Middle and Late Pleistocene. During this period, New Zealand was repeatedly fragmented by glaciers into a series of refugia, with the tiny geographic ranges of many kiwi lineages currently distributed in areas adjacent to these refugia. Estimates of effective population size through time show a dramatic bottleneck during the last glacial cycle in all but one kiwi lineage, as expected if kiwi were isolated in glacially induced refugia. Our results support a fivefold increase in diversification rates during key glacial periods, comparable with levels observed in classic adaptive radiations, and confirm that at least some lineages distributed near glaciated regions underwent rapid ice age diversification.
Full-text available
Panbiogeographic analysis is now used by many authors, but it has been criticised in recent reviews, with some critics even suggesting that studies using the method should not be accepted for publication. The critics have argued that panbiogeography is creationist, that it rejects dispersal, that its analyses are disingenuous, and that it deliberately ignores or misrepresents key evidence. These claims are examined here, and are all shown to be without foundation. The distributions of the molecular clades of ratites have not been mapped before, and they are considered here in some more detail as a case study illustrating panbiogeographic methodology.
Full-text available
Since the 1980s, morphological and molecular research has resulted in significant advances in understanding the relationships and origins of the recent terrestrial vertebrate fauna in the New Zealand biogeographic region. This research has led to many taxonomic changes, with a significant increase in the number of bird and reptile species recognised. It has also resulted in the recognition of several more Holocene (<10 000 years ago) bird species extinctions. The conclusion that Holocene extinctions were primarily caused by human-hunting and predation by other introduced mammals (particularly rats and cats) has been supported by new data. Despite many local eradications of introduced pests, the number of introduced species has increased, with the establishment of five more foreign birds and (on Norfolk Island) the house gecko (Hemidactylus frenatus). Many new, significant New Zealand vertebrate fossils have been reported, including more dinosaurs from the Cretaceous, and the first Tertiary records of frogs, rhynchocephalids, lizards, crocodylians, bats and a terrestrial "Mesozoic ghost" mammal from the Early Miocene near St Bathans. For birds, the earliest known penguins in the world have been discovered, and there are intriguing Late Cretaceous - Early Paleocene remains still awaiting detailed description. Other significant Tertiary bird fossils reported include a rich avifauna from the Early Miocene St Bathans sites and a small terrestrial fauna from the Early Pleistocene near Marton. In line with the traditional theory, new research has supported the vicariant Gondwanan origin of some distinctive New Zealand terrestrial vertebrates, such as leiopelmatid frogs, tuatara and moa, and the immigration of many others, including New Zealand wattlebirds and piopio, during the Cenozoic. Extinctions caused by an asteroid impact and climate fluctuations probably explain the absence of many groups, such as crocodylians, dinosaurs, monotremes, palaelodids and swiftlets, from the modern fauna.
Full-text available
Dissertation Despite our intuition, birds are no smaller than mammals when the constraints of a flying body plan are taken into account. Nevertheless, the largest mammals are ten times the mass of the largest birds. Allometric equations generated for anseriforms and ratites suggest mid-shaft femur circumference is the best measure to use in estimating avian body mass. The small sample size of extant ratites makes mass estimate extrapolation to larger extinct species inaccurate. The division of ratites into cursorial and graviportal groups is supported. Aepyornithids do not show atypical femoral shaft asymmetry. New and more accurate estimates of egg masses, and separate male and female body masses for sexually-dimorphic ratites are generated. Egg mass scaling exponents for individual bird orders differ from that Aves as a whole, probably due to between-taxa effects. Ratite egg mass does not scale with the same exponent as other avian orders, whether kiwi are included or excluded. Total clutch mass in ratites, however, scales similarly to egg mass in other birds, perhaps as a consequence of the extreme variation in ratite clutch size. Kiwi and elephant bird eggs are consistent with the allometric trend for ratites as a whole, taking clutch size into account. Thus kiwi egg mass is probably an adaptation for a precocial life history, not a side effect of their being a dwarfed descendant of a moa-sized ancestor. Relatively small body size in ancestral kiwis is consistent with a trans-oceanic dispersal to New Zealand in the Tertiary, as suggested by recent molecular trees. This implies multiple loss of flight in Tertiary ratite lineages, which is supported by biogeographic, molecular, paleontological, and osteological evidence, but which is not the currently prevailing hypothesis.
Full-text available
. Several isolated bones of tinamous from Miocene deposits of Santa Cruz Province (southern Patagonia, Argentina) are the oldest known remains of this paleognath lineage. The specimens include an incomplete coracoid, proximal end of four coracoids, distal ends of two tibiotarsi, and distal ends of two humeri. They represent at least two species but cannot be assigned to any known taxon. A detailed description and phylogenetic interpretation of this material is provided here. Morphological data of the fossils are included in a matrix of 63 osteological characters and 34 terminal taxa incorporating 24 living species of Tinamidae in addition to the fossils under study. The cladistic analysis produced 81 optimal trees, in which the fossils are more closely related to the open-area tinamous (Nothurinae). Placement of the Santa Cruz fossil tinamous between the open-area (Nothurinae) and the forest-dwelling (‘‘Tinaminae’’) tinamous is consistent with the paleoenvironmental conditions inferred from the associated fossil fauna.
Full-text available
A new emu (Emuarius guljaruba, sp. nov.) is described from the Late Oligocene Etadunna Formation (Ngama Local Fauna), based on a complete tarsometatarsus. While exhibiting evidence of cursorial abilities advanced over those of cassowaries (Casuarius), this taxon was not as cursorially adapted as the living Emu (Dromaius novaehollandiae). This taxon is provisionally referred to the genus Emuarius, although a definite generic assignment cannot be made.
Full-text available
I examine the hypothesis that energy conservation contributes to the evolution of a flightless condition in birds by comparing the factors that correlate with basal rate of metabolism in kiwis and flighted and flightless rails and ducks. Flightless rails have low basal rates, the level of which decreases with pectoral muscle mass. Kiwis also have low basal rates and small pectoral masses. The small pectoral masses found in flightless grebes, the flightless cormorant, and the flightless parrot suggest that these species have low basal rates. Penguins and flightless ducks, in contrast, have neither low basal rates nor small pectoral masses because these birds use their wings for locomotion. These data are compatible with the hypothesis that energy conservation contributes to the evolution of flightlessness in species in which pectoral muscle mass is reduced. On oceanic islands, rails have evolved a flightless condition repeatedly, usually in association with a small body size. Both adjustments reduce energy expenditure, which thereby facilitates the persistence of rails in environments with limited resources. The evolution of flightlessness in insects may also be a response to a restricted resource availability, especially in persistent habitats characterized by low rates of production.
The red-wing tinamou (Rhynchotus rufescens) is economically important as food. The current study investigated the limb and trunk characteristics in age-matched [year-of-hatch (yoh) 2004 and 2005], gender segregated birds, and determined differences in stride between cocks and hens. The locomotion trial was completed in a corridor of 0.6 × 2.36 m dimension. The girth was significantly higher in cocks than in hens, while body weight was slightly higher in hens. Cocks had a greater height than hens. The time to walk 2.36 m was longer for cocks than for hens. This was related to an attenuated speed and stride length in cocks by comparison with hens. The increased number of strides and reduced stride lengths in cocks suggested their defensive posture by lagging behind the fleeing hens or juveniles. Subtle differences in yoh may have been related to small variances in nutrition, weather, management, age, and reproductive traits. Although the birds born in distinct years were submitted, the net sum of these effects could have potentiated the different performances observed.