ArticlePDF Available

Volcanic history of Macauley Island, Kermadec Ridge, New Zealand



An open access copy of this article is available from the publishers website. Macauley Island (3 km2) is the tiny emergent part of the large submarine Macauley volcano (c. 380 km2 at the 900 m isobath) on the Kermadec Ridge. It is composed mainly of arc tholeiite basalts, with a single interbedded dacite tephra. The oldest rocks seen are subaerial aa flows (North Cliff Lavas), overlain by basaltic tephra deposits (Boulder Beach Formation). Continued eruption of thin basalt flows (Annexation Lavas) built a large shield volcano, at least 150 m above sea level, with a crater in the vicinity of what is now Mt Haszard. A large eruption of dacite tephra (Sandy Bay Tephra) caused collapse of the flanks of the submarine volcano, to form a caldera immediately northwest of the present island. Renewed basaltic volcanism produced scoria cones, flows, and tephra (Haszard Formation), and the final stage of this eruptive phase was associated with the collapse of the northwest edge of the island into the caldera. The freshness of the exposed rocks on Macauley Island indicates a late Quaternary (possibly Holocene) age for the whole succession, and this is supported by a radiocarbon date of 6310 ᄆ 90 yr B.P. on the Sandy Bay Tephra.
Full Terms & Conditions of access and use can be found at
Download by: [] Date: 30 January 2016, At: 01:14
New Zealand Journal of Geology and Geophysics
ISSN: 0028-8306 (Print) 1175-8791 (Online) Journal homepage:
Volcanic history of Macauley Island, Kermadec
Ridge, New Zealand
E. F. Lloyd , Simon Nathan , I. E. M. Smith & R. B. Stewart
To cite this article: E. F. Lloyd , Simon Nathan , I. E. M. Smith & R. B. Stewart (1996) Volcanic
history of Macauley Island, Kermadec Ridge, New Zealand, New Zealand Journal of Geology
and Geophysics, 39:2, 295-308, DOI: 10.1080/00288306.1996.9514713
To link to this article:
Published online: 23 Mar 2010.
Submit your article to this journal
Article views: 163
View related articles
Citing articles: 21 View citing articles
New Zealand Journal of Geology and Geophysics, 1996, Vol. 39: 295—308
0028-8306/96/3902-0295 $2.50/0 © The Royal Society of New Zealand 1996295
Volcanic history of
Ridge, New
formerly DSIR Geology & Geophysics
Rotorua, New Zealand
Institute of Geological & Nuclear Sciences
Box 30368
Lower Hurt, New Zealand
Department of Geology
University of Auckland
Private Bag 92019
Auckland, New Zealand
Department of Soil Science
Massey University
Private Bag 11222
Palmerston North, New Zealand
Abstract Macauley Island (3 km2) is the tiny emergent
part of the large submarine Macauley volcano (c. 380 km2
at the 900 m isobath) on the Kermadec Ridge. It is composed
mainly of arc tholeiite basalts, with a single interbedded
dacite tephra.
The oldest rocks seen are subaerial aa flows (North Cliff
overlain by basaltic tephra deposits (Boulder Beach
Formation). Continued eruption of thin basalt flows
(Annexation Lavas) built a large shield volcano, at least
150 m above sea level, with a crater in the vicinity of what
is now Mt Haszard.
A large eruption of dacite tephra (Sandy Bay Tephra)
caused collapse of the flanks of the submarine volcano, to
form a caldera immediately northwest of the present island.
Renewed basaltic volcanism produced scoria cones, flows,
and tephra (Haszard Formation), and the final stage of this
eruptive phase was associated with the collapse of the
northwest edge of the island into the caldera.
The freshness of the exposed rocks on Macauley Island
indicates a late Quaternary (possibly Holocene) age for the
whole succession, and this is supported by a radiocarbon
date of 6310 ± 90 yr B.P. on the Sandy Bay Tephra.
Keywords volcano; arc tholeiite; basalt; dacite; pumice;
caldera; tephra; crater; phreatomagmatic; radiocarbon;
Macauley I.
Received 12 May 1995; accepted 14 December 1995
The Kermadec Islands are a chain of young island arc
volcanoes surmounting the NNE-trending Kermadec Ridge
(Fig. 1). As noted by Lloyd & Nathan
p. 11), detailed
bathymetry shows that individual islands lie onESE-trending
ridges, normal to the overall NNE trend of the Kermadec
Ridge, suggesting that structural features such as cross-
fractures may control the location of the major volcanoes.
Macauley Island, the second largest of the Kermadec
group, has a total area above sea level of only 3 km2
(including nearby Haszard Islet and Newcombe Rock). This
represents the emergent part of a large submarine volcano
(Macauley volcano), covering c. 380 km2 at the 900 m
isobath. Immediately north of Macauley Island there is a
roughly circular submarine depression, c. 12 km in diameter
and up to 1.1 km deep, which we interpret as a caldera into
which part of Macauley volcano subsided.
Much of Macauley Island is a rolling meadow, only
c. 100 m above sea level. The land rises sharply northwards
to 238 m at Mt Haszard, west of which a steep cliff (named
"Perpendicular Cliff by Brothers & Martin 1970) drops into
deep water near the southern end of the caldera (Latter et al.
fig. 9a). This cliff exposes a section through part of a
mainly basaltic shield volcano, including fragments of two
coalescing craters, lava flow sequences, and tephras (Fig. 2).
Macauley Island is an excellent example of a complex
oceanic shield volcano with a degree of exposure around
the coastal cliffs seldom seen elsewhere. This paper aims to
interpret the volcanic history of Macauley volcano, using
both the onshore geology and bathymetric investigations of
the surrounding part of the Kermadec Ridge.
The first geological description of Macauley Island was
given by Smith (1888). Further brief accounts, mainly rock
descriptions, were given by Thomas (1888), Speight (1896),
and Oliver (1911).
In 1966, K. R. Martin mapped the island in detail, and
subsequently Brothers & Martin (1970) gave a detailed
stratigraphic description and the first modern account of the
petrography. Further petrological and geochemical
information is given in Ewart et al. (1977), Brothers &
Hawke (1981), and Gamble et al. (1990, 1993).
EFL and SN spent 12 days on Macauley Island in
November 1980, as members of an expedition organised by
the former Wildlife Service of the Department of Internal
Affairs (now Department of Conservation), and produced a
revised geological map (Fig. 3). Our interpretation of the
onshore geology and scanty bathymetry suggested the
presence of a submarine caldera immediately northwest of
Macauley Island. In March 1988, EFL and B. J. Scott joined
a cruise on RV Vulkanolog, which obtained detailed
Downloaded by [] at 01:14 30 January 2016
296New Zealand Journal of Geology and Geophysics, 1996, Vol. 39
^ ( 1DecJ867
/ Is. \ [ ) \ \( I
Fig. 1 Bathymetry (in metres) around Macauley volcano, showing the tiny emergent part that constitutes Macauley Island, and the
submarine caldera that is interpreted to have collapsed during eruption of the Sandy Bay Ignimbrite. "V" indicates the inferred locus
of submarine volcanism terminating the Haszard eruption (Parakeet Tuff). Inset: Location of Macauley Island on the Kermadec Ridge.
The locations of two possible historic volcanic eruptions near Macauley volcano are shown by asterisks (*).
bathymetry over the subsea Macauley volcano and defined
the extent of the Macauley caldera (Scott & Lloyd 1989).
IEMS and RBS visited Macauley Island for 5 days in
April 1993, particularly to examine the Sandy Bay Tephra
and collect material for C-14 dating.
Petrology and geochemistry
Previous petrological work has established that Macauley
Island is composed of uniform arc tholeiite basalt (S1O2 =
48.30-50.03%), with a single interbedded composite tephra
unit of dacitic composition (SiO2 = 67.38-68.39%). The
rocks of Macauley Island appear to be distinct from those
of Raoul Island, for example, in having generally higher K2O
at given SiC>2 content. Petrological nomenclature in this
paper follows that used by Lloyd & Nathan (1981) for Raoul
Brothers & Martin (1970) divided the Macauley succession
into eight formations, which formed two major groups
(Table 1). The justification for recognising the two groups
was "a marked erosional break at the top of the Sandy Bay
taken as the boundary between the two groups"
(Brothers & Martin 1970, p. 333).
We recognise the Macauley Island sequence as
essentially conformable, but with local erosional breaks
between most units. Erosion at the top of the Sandy Bay
Tephra could have resulted from heavy rain in only days to
and we do not believe that there is evidence for a
major break. For this reason, we decided to abandon the use
of group nomenclature.
We have slightly modified the stratigraphic classification
used by Brothers & Martin (1970), as summarised in Table 1.
We prefer to group the deposits of each individual eruptive
episode together as part of a single formation, and for this
reason we have made Cascade Lavas a member of Haszard
Formation. A geological map of Macauley Island, based on
these stratigraphic units, is shown in Fig. 3.
Description of onshore sequence
(Brothers & Martin 1970)
DESCRIPTION: North Cliff Lavas, the oldest rocks on
Macauley Island, are exposed only at the bottom of the cliff
section at the northern part of the island, and their base is
not seen. They consist of highly vesicular aa flows of olivine
basalt, which apparently formed part of a low emergent
Downloaded by [] at 01:14 30 January 2016
Bay Tephra
Annexation Lavas
Boulder Beach Formation
= tephra
= scoria
= lava
Mt Haszard
2 A,
of the
at the
end of
Maeauley Island.
Overlay, showing
major units exposed
in the
cliff section. Maximum height
is 230 m.
Downloaded by [] at 01:14 30 January 2016
298New Zealand Journal of Geology and Geophysics, 1996, Vol.
Annexation Point
Sandy Bay
Rk V
„„ .
Haszara Tuff + Haszard
Sandy Bay
0.51.0 km
Fig. 3 Geological map of Macauley Island.
shield. The lavas have an apparent dip c. 3°E. In the cliff
beneath Mt Haszard they extend up to 15 m above sea level.
Together with overlying Boulder Beach Formation and
Annexation Lavas, they are truncated by a steep
discontinuity, which we interpret as a crater wall. Close to
the discontinuity, low-temperature (<105°C) vapour phase
alteration of North Cliff Lavas has changed the colour of
the rocks to pink and red-brown, and produced secondary
natroalunite, gypsum, kaolinite, tridymite, montmoriUonite,
and hematite (C. P. Wood pers. comm.). In 1980 there was
a faint sulphurous smell near the crater wall discontinuity
Brothers & Martin (1970) described North Cliff Lavas
as horizontal and separated from the overlying Boulder
Beach Formation by a low angular unconformity. We
interpret the two formations as essentially conformable,
although bedding is locally nonparallel at the western end
of the outcrop in a zone disrupted by small-scale normal
faulting (Fig. 4).
Downloaded by [] at 01:14 30 January 2016
Lloyd et al.—Volcanic history of Macauley I.299
Fig. 4 Reference locality for
the North Cliff Lavas and
Boulder Beach Formation on the
coast at the northern end of
Macauley Island. The North
Cliff Lavas at bottom left are
overlain by bedded tephras of
Boulder Beach Formation and
capped by Annexation Lavas.
The Annexation Lavas at upper-
most left are thin flows and
tephra of the shield, whereas
those at centre are crater-con-
fined lavas, previously inter-
preted as a stock by Brothers &
Martin (1970). The flow
sequence at far upper right is
Haszard Lavas filling the
western crater exposed in
Perpendicular Cliff (see Fig. 5).
Small faults are downthrown
right (north) near the left of the
view, with dikes intruding the
fault planes. The steep contact
between Annexation Lavas
(right) and Boulder Beach
Formation (left), is interpreted as
a crater wall.
INTERPRETATION: Effusive volcanism from a centre not far
north of Macauley Island built the low subaerial shield of
North Cliff Lavas. That the shield survived marine erosion
reflects the resistant nature of the basalt flows, and suggests
that there is unlikely to have been a long time break before
the eruption of the tephras in the overlying Boulder Beach
(Brothers & Martin 1970)
This unit crops out only in the lower part of the cliffs at the
northern end of the island, where it overlies the North Cliff
Lavas and underlies Annexation Lavas (Fig. 4).
DESCRIPTION: Boulder Beach Formation mantles the lava
flows of the underlying North Cliff Lavas, and consists of
bedded, variably sorted, brown palagonitised primary (and
possibly reworked) basaltic tephras, with a few interbedded
thin basalt flows. Maximum thickness (of c. 25 m) and
highest clast concentration occur at the western end of the
cliff section where the tephras are truncated by the old crater
wall, and the unit thins down-dip to the east until it is 8 m
thick where its base disappears beneath the boulder beach.
At this locality, bedding is planar in the basal 6 m, but cross-
bedding is developed within the sequence, and there are also
scattered disruptions of the bedding created by bomb
impacts. Accretionary lapilli are scattered through the deposit
but are particularly common near the top. Dikes penetrate
the tephra sequence. Most appear to have fed lava to the
overlying Annexation Lavas, but one fed a small cone and
lava flow on the surface of the primary tephra. The
uppermost metre of Boulder Beach Formation is baked dark
red by the overlying Annexation Lavas.
INTERPRETATION: The primary Boulder Beach tephras are a
mixture of basaltic fall, surge, and ballistic deposits produced
by a dominantly phreatomagmatic style of volcanism. This
Table 1 Comparison of rock units used by Brothers & Martin (1970) with those described here.
UnitBrothers & Martin (1970)This paper
Haszard Group
Macauley Group
Grand Canyon Formation
Haszard Formation
Cascade Lavas
erosional break
Parakeet Tuff Member
Haszard Scoria Member
Sandy Bay Tuff
Perpendicular Cliff Intrusion
Annexation Lavas
Boulder Beach Formation
North Cliff Lavas
Grand Canyon Formation
Haszard Formation
Sandy Bay Tephra
Annexation Lavas
Boulder Beach Formation
North Cliff Lavas
Parakeet Tuff Member
Haszard Scoria Member
Cascade Lava Member
Downloaded by [] at 01:14 30 January 2016
300New Zealand Journal
Geophysics, 1996,
Vol. 3S
" has
1 /
an ^^
end of
icular Cliff section, looking south
(drawn from
colour slide). Two
coalescing craters
older (on left) filled with massive
crater-ponded Annexation Lavas
on the
by ;i
younger crater, infilled with
scoria cone
Mt Haszard
in the
east and with Cascade Lava which
has issued from
cone vent and
accumulated around
it. The
of the fossil craters
a combination line/dot symbol.
Same letter symbols
as in
2, 3.
water access
to the
magma since
Cliff eruption, though
vent location north
Island may
have changed. Dikes were intruded,
basaltic lava flows discharged from them.
(after Brothers
is a
widespread unit
Macauley Island, cropping
at the
end of
the island,
the west,
also forming
shore platform around most
of the southern part
the island. Grouped with Annexation
lava flows
proximal scoria deposits
at the
tip of
Haszard Islet,
Newcombe Rock, which,
seen from
as a
lava plug.
Annexation Lavas
aa and
pahoehoe flows
porphyritic olivine-plagioclase basalt, with associated minor
lava flows greatly increased
shield which emerged above
level during
North Cliff
individual flows, including
scoriaceous base
and top, is
1 m or
and rarely exceeds
3 m.
Annexation Lavas conformably
overlie Boulder Beach Formation
the northern cliffs below
Mt Haszard,
dip c. 3°
thin southwards. They total
115 m
at the
end of
these cliffs,
two thin
distinctive beds
brown tephra which divide
the formation into three roughly equal parts. Their maximum
at the
northwest headland, where they
extend from below
to 150 m
Numerous dikes, mainly
<1 m
wide, exposed along
northern cliffs, supplied lava
to the
Annexation sequence,
and show that,
to the
main lava source, lava also
erupted from flank vents. Newcombe Rock
the largest
these flank vents.
& Martin (1970,
fig. 3)
mapped Perpendicular Cliff
Annexation Lavas intruded
by a
stock, which they named
the Perpendicular Cliff Intrusion. We reinterpret this section
overlapping craters cutting through Annexation
filled with crater-confined lava flows
5). The
older, eastern crater contains
basalt "stock"
of Brothers
Martin (1970), which
is in
crater wall contact
with thin shield-forming Annexation Lavas,
and is
by Sandy
Tephra. Intrusive emplacement
this basalt
ruled out
by an
almost horizontal scoriaceous parting
lava flow boundary, exposed
at the
accessible outcrop
at the
northeastern base
"Perpendicular Cliff Intrusion"
crater-confined, ponded lavas. Since there
is no
major volcanism
Macauley Island between
visualise these thick lava>
as products
late-stage Annexation volcanism.
of the
Annexation lavas
essentially effusive Hawaiian-style volcanism from vents
to be
immediately north
Perpendicular Cliff
About 1 km3
lava was added
to the
terrestrial shield, which
maximum height
150 m
a diameter
4 km.
normal faulting, first
evident during
Boulder Beach eruption,
widespread during Annexation volcanism. Faults
in the
the crater, probably
summit inflation,
were occupied
magma which produced small flank lava
flows. More distant flank vents also formed, such
Newcombe Rock.
eruption style
changed briefly from effusive
least twice
Annexation event when thin beds
fine tepbxa
were deposited
on the
shield. Towards
the end of the
eruption, lava solidified
in the
deep summit crater
by the
Haszard eruption commenced,
probable that this formed
at the end of
Annexation volcanism.
(modified from Brothers
Martin (1970) named this unit Sandy
and described
it as a
"light grey highly pumiceous vitrk
it as a
composite tephra deposited
multiple pyroclastic flows
propose that
in the
coastal cliff 0.6
Jims Gully (Fig.
type locality.
is the
most conspicuous
Macauley and Haszard Islands
as its
pale-grey colour
contrasts strongly with
other dark basaltic rocks.
poorly sorted, juvenile dacitic pumice,
predominantly finely comminuted
scattered pumice clasts <100
mm in
diameter, together with
scattered lithic clasts
rare plutonic blocks.
Downloaded by [] at 01:14 30 January 2016
Lloyd et al.—Volcanic history of Macauley I.301
Fig. 6 Reference locality for
Sandy Bay Tephra, 0.6 km east
of Jims Gully. At least 30 flow
units comprise the section. The
shore platform is Annexation
lithic clasts are unaccompanied by bomb sags and apparently
were rafted into position. Their large size—largest observed
is 1.5 m by Brothers & Martin (1970), and 0.3 m clasts are
common—suggest that Macauley Island is near to the source
of Sandy Bay Tephra.
Bedding in the deposit is distinct and highly irregular
(Fig. 6, 7). Bedforms characteristic of moderate energy
turbulent deposition (Fisher & Schmincke 1984), such as
antidunes, pinch and swell bedding, and cross-bedding,
dominate low-energy bedforms, and we attribute the bulk
of the unit to turbulent pyroclastic flows which not only
deposited tephra but also eroded deposits of preceding flows
to create the irregular internal unconformities. At the type
locality there are at least 30 flow units, and the basal deposit
is pumiceous ignimbrite, which is overlain by surge deposits
with internal bedding of millimetre-centimetre scale,
distinctly enriched in sand-lapilli sized lithic basalt (Fig. 7,
The overlying deposits of pyroclastic flows are typically
massive, poorly sorted, reverse graded, and pass up
gradationally, or less commonly with sharp contact, into fine
ash. The fine ash is interpreted as either co-ignimbrite ash
or fall deposit. Both interpretations imply discontinuous
pyroclastic flow production, with sufficient time between
flows for the co-ignimbrite ash to settle and/or ash fallout
from the eruption cloud to accumulate. That significant time
elapsed between successive flows is also deduced from the
sharp definition of bed and flow boundaries, indicating that
a deposit had time to deflate before being overridden by the
next pyroclastic flow or surge.
Sandy Bay Tephra is up to 100 m thick in the cliffs at
the southern end of Macauley Island, but thins northwards
to c. 15 m on the upper parts of the Annexation lava shield.
Brothers & Martin (1970) attributed thinning of Sandy Bay
Tephra at higher elevations to "extensive subaerial erosion".
We recognise evidence for only relatively minor erosion,
and interpret both upper and lower contacts of the Sandy
Fig. 7 Bedding detail of the lower part of Sandy Bay Tephra at
the type locality, including pinch and swell and cross-bedding, as
well as fine, co-ignimbrite ash separating some flow units. A
detailed view of the lower part of the section is shown in Fig. 8.
Downloaded by [] at 01:14 30 January 2016
302New Zealand Journal of Geology and Geophysics, 1996, Vol.
%^& Fig. 8 The base of Sandy Ba
~j( Tephra at the type locality consisi
of fine pumiceous ignimbrit
fine-bedded lithi
basalt-rich surge deposits. This
St* >s*fr* layer contains subhorizontalw
, «. .' - ' fragments, mostly represented b
|| 'J|p casts (one occupied bythr
geological hammer). Sample*
taken for C-14 dating are from a
equivalent stratigraphic horizon
Bay Tephra as paraconformable. Thickening of the
ignimbrite into subaerial topographic lows is deposit ional,
and its absence from Mt Haszard cone is because the cone
is younger.
overlies Annexation Lavas conformably, with a sharp change
in lithology but little indication of erosion. The basal beds
of the Sandy Bay Tephra consist of a variety of thin,
discontinuous pyroclastic layers, which are overlain by
thicker ignimbrite layers, typical of the main part of the
Sandy Bay Tephra. The most complete sequence of units is
exposed near the base of South Stack, at the southern end of
the island (Fig. 9). The base of the exposure is a pahoehoe
flow top. Overlying this is a lapilli unit 7.5—20 cm thick and
an ash bed up to 20 cm thick. Both are chocolate brown in
colour, and are interpreted as a basaltic tuff with associated
soil overlying the Annexation Lavas. They are overlain by
a unit characterised by basaltic clasts in a pumice-dominated
matrix and an essentially pumiceous ash layer, both showing
well-developed cross-bedding. These are interpreted as early
surge deposits of the Sandy Bay eruption, which incor-
porated surficial basaltic material. The upper layer in this
section is an irregular bed, up to 15 cm thick, consisting of
a reversely graded, fines-depleted diamictite containing
rounded basaltic clasts, some up to 5 cm in diameter, in a
pumiceous matrix. This is interpreted as a ground layer of
pyroclastic flow deposits.
At some localities along the southeast coast of Macauley
Island all these units are absent, and massive ignimbrite rests
directly on Annexation Lavas. At other localities, only parts
of the sequence are present.
At several localities in the vicinity of South Stack and in
basal cliff exposures northwest to Jims Gully there is
evidence for vegetation, mainly as casts of wood (?branches
of trees) in the cross-bedded surge layer (Fig. 8, 9). These
wood remnants range up to 30 cm in diameter, are typically
subhorizontal, and have orientations of 100—140° true. The
casts are generally empty, but some contain amorphou
carbonaceous material that was collected for C-14 dating
This vegetation is interpreted as the remains of scrubby fores;
which was overwhelmed by early explosive events at the
start of the Sandy Bay eruption.
AGE: A very small sample of plant fragments from the base
of the cross-bedded tephra, collected c. 30 m southeast of
the mouth of Jims Gully, gave a conventional radiocarbon
date (KE/f50; NZ5110A, B) of <500 years B.P., but with a
very large experimental error because the sample contained
only 27 mg of elemental carbon.
A nearby locality, on the eastern side of South Stack,
was recollected in 1993, and gave an AMS radiocarbon date
(NZA4624) of 6310 + 190 years B.P. The AMS date is
accepted as the more reliable, and is interpreted as dating
the initial phase of the Sandy Bay eruption, when existing
vegetation was knocked over.
INTERPRETATION: The tephra succession preserved on
Macauley and Haszard Islands consists of at least 30
pyroclastic flow units separated by thin deposits of co-
ignimbrite and airfall tephras, with a maximum preserved
thickness of 100 m. The bedded nature of the deposit:-
indicates discontinuous eruption of pyroclastic flows, eithei
by a series of direct explosive eruptions or as a result o!
intermittent collapse of a semi-continuous, high eruptior
column. The earliest ignimbrites to overwhelm Macauley
Island did not char the vegetation, which implies deposition
at temperatures not much above 280°C, nor did subsequent
deposits weld. Sea-water ingestion into the erupting column
is a likely cause of the relatively low temperatures.
Nevertheless, the pyroclastic flows were of moderate energy,
and their eruption and dispersal probably generated
widespread airfall deposits.
The onshore distribution of tephra indicates a source
north of Macauley Island. Eruption of a large amount of
pyroclastic material is likely to result in caldera collapse. A
Downloaded by [] at 01:14 30 January 2016
304New Zealand Journal of Geology and Geophysics, 1996, Vol. 3
Fig. 10 Lava Cascade, a ravim
cut in Sandy Bay Tephra, ant
filled with thick flows typical
the Cascade Lavas.
indication of the time gap between the Sandy Bay and
Haszard eruptions.
Cascade Lavas Member {half
The distribution of Cascade Lavas is shown in Fig. 3.
Cascade Lavas overlie minor aggradational fluvial deposits
(not mapped) on the floors of paleocanyons cut in Sandy
Bay Tephra down which the lavas flowed to the sea. The
inaccessible canyon exposure in the cliff at Annexation Point,
0.2 km WNW of Lava Cascade, shows c. 1 m of cross-
bedded alluvium beneath Cascade Lavas. The uppermost
200—300 mm of alluvium is speckled with black (?basaltic)
fragments compared with the white lower part.
Basalt flows of the Cascade Lavas fill the western crater
in Perpendicular Cliff crater from below sea level to c. 150 m
high. The section is inaccessible except in its uppermost part.
Viewed looking south from the sea (Fig. 5) the lower 45 m
consists of lava with prominent vertical joints which inter-
sect apparent faint horizontal banding, interpreted as a
single cooling unit composed of several flow units.
Overlying aa flows rarely exceed 2 m thick (including their
prominent scoriaceous bases and tops). Scoriaceous parts
of these flows are vapour-phase altered, indicating that the
flows retained heat for considerable time and hence probably
accumulated rapidly. In the west, the flows are horizontal,
but beneath Mt Haszard they steepen, bedding becomes
confused, and the proportion of clastic material increases.
This confused region is interpreted as a tephra and lava cone
which grew around a vent in the crater floor to produce Mt
Cascade Lavas discharged at the southeast rim of the
western crater and cover c. 1 km3 in the southwestern part
of Macauley Island. A dike-fed fissure eruption above the
western cliff produced a small volume of lava which flowed
southwest. Most of these flows were captured by narrow
canyons in the Sandy Bay Tephra. Lavas flowing eastwards
from the vicinity of Mt Haszard were also captured by
canyons. The lava-filled canyon called Lava Cascade
(Fig. 10) is selected as the type locality for Cascade Lavas
Cascade Lavas probably also occurred on northern flank s
of Macauley volcano, which subsequently collapsed into the
sea. The thickness and number of Cascade lava flows on
the flanks of Macauley Island vary. Flows on open slope
are c. 1-2 m thick, and where confined within canyons, u
to 8 m thick. Six lava flows of combined thickness 25 m
infill the largest of two fossil valleys exposed high in r
northern cliffs.
Haszard Scoria Member {has}
Haszard Scoria is black, well sorted, vesicular basalt scoria
deposited during an open-vent magmatic eruption of
Strombolian or possibly sub-Plinian type. The scoria is
conformable on Cascade Lavas or, where lava is absent,
paraconformable on Sandy Bay Tephra. Because of its
widespread exposure, it is possible to compile an isopach
map for this unit (Fig.
A), as well as isopleths on
maximum size of scoria clasts.
Proximal Haszard Scoria is exposed at the top of the
northern cliffs, but close inspection is hazardous because of
the precipitous terrain. The first gully south of the cliffs is
suggested as the reference locality, although the base is not
exposed. More distal Haszard Scoria is exposed at the top
of the cliff behind Sandy Bay (Fig. 12).
At the southeast end of Macauley Island, basal Haszard
Scoria is distinctly shower bedded (Fig. 12), with individuai
beds conformable and normally graded, and overlain by a
thicker, well sorted, massive scoria deposit. Shower bedding
is also seen in Haszard Scoria in the northern cliff exposure s
Most other outcrops expose only the upper massive part of
the sequence which, in proximal outcrops, has rare vague
alignments created by scattered large scoria. Lithic blocks
become conspicuous and abundant towards the top. This
scoria-supported lithic-block deposit changes upwards int<
an ash/lapilli-supported lithic-block deposit; the scoria t<
Downloaded by [] at 01:14 30 January 2016
Lloyd et al.—Volcanic history of Macauley I.305
(h metma) on upper
unit of t*
(in matma) oo Ihe Parakael
ton m
n Ih«
Fig. 11 Variations in properties of the Haszard Scoria and
Parakeet Tuff A, Isopachs (in metres) on upper massive unit of
the Haszard Scoria. B, Isopachs (in metres) on the Parakeet
C, Variation in maximum scoria size (in millimetres) in the Haszard
ash transition marks the division of Haszard Scoria and
The lithic blocks, some hydrothermally altered, are
poorly vesicular to vesicular older basaltic lavas. Maximum
diameter is up to 2 m around Mt Haszard. Although lack of
bedding in the upper Haszard Scoria makes detection of
bomb sags difficult, the blocks are not accompanied by
disruptions to the enclosing scoria deposit, and appear to
have "rafted" into position. Cross-bedded lithic-rich deposits
at the transition from Haszard Scoria to Parakeet Tuff in a
. ••?-;
Fig 12 Haszard Scoria exposed at the southeast end of Macauley
Island above Sandy Bay. The basal 1.28 m is shower bedded and
is overlain by 2.65 m of massive scoria.
cliff north of Mt Haszard are interpreted as surge deposits,
and it appears that dual emplacement mechanisms were
operative at this time.
Parakeet Tuff Member {pat}
Parakeet Tuff gradationally overlies Haszard Scoria and is
the soil-forming deposit over most of Macauley Island. From
a maximum thickness of c. 25 m in the northern cliff just
east of Mt Haszard, the deposit thins southwards and is
barely detectable near the south coast (Fig. 11B).
Parakeet Tuff consists of well-bedded, poorly sorted,
indurated, fine grey ash and lapilli generated by submarine
phreatomagmatic eruption and emplaced as airfall tephra and
subordinate surge deposits. The ash and lapilli are mostly
poorly vesicular to non-vesicular and subrounded.
Vesiculation of fine ash beds at the top of the northern cliff
section indicates that the tephra fell wet. The contact with
underlying Haszard Scoria is transitional, with scoria-
dominated Haszard Scoria grading over several centimetres
to ash/lapilli-dominated Parakeet
Lithic blocks up to
2 m in diameter are common in the transitional zone close
to Mt Haszard. The lithic-rich zone reaches maximum
thickness of 4 m in the northern cliffs east of Mt Haszard,
and generally is thicker and the blocks larger in Haszard
Scoria than in Parakeet
The blocks were probably
transported by surges. Above the lithic zone, Parakeet Tuff
Downloaded by [] at 01:14 30 January 2016
306New Zealand Journal of Geology and Geophysics, 1996, Vol. 3')
consists of vesiculated and indurated fine ash with thicker
interbeds of lapilli.
maps of Haszard Scoria and Parakeet Tuff are shown in
A, B, respectively, and the maximum scoria size for
the upper part of the Haszard Scoria Member is shown in
Fig. 11C. The thickness patterns for Haszard Scoria and
Parakeet Tuff are similar, suggesting that both units were
erupted from the same source, to the north of the present
island. By extrapolation northwards from the isopachs
determined on land, the vent location appears to have been
0.75 km northwest of Mt Haszard where the sea depth is
now c. 100 m.
volcanism apparently commenced in the collapse crater
formed at termination of the Annexation eruption near the
summit of the Annexation shield. The crater extended north
and south of Perpendicular Cliff and to below sea level. Early
activity was dominantly effusive; lava spread over the crater
floor mainly as thin aa flows, and a lava and scoria cone
formed around the active vent. Ultimately, lava filled the
crater and overflowed the rim to southeast, east, and probably
elsewhere. The lava and scoria cone overtopped the rim,
creating Mt Haszard. Pyroclastic ejecta from this phase of
the eruption is recognised only at Mt Haszard cone. It is
tentatively correlated with fine tephra incorporated high in
the alluvium sequence deposited in canyons cut into Sandy
Bay Tephra.
The next phase of the eruption commenced when flank
fissures opened through which magma erupted farther north
on the Annexation shield. The small dike-fed lava field at
the western side of Macauley Island is evidence of this, but
most of the evidence for what may have been a widespread
and significant event was destroyed by subsequent sub-
sidence. That this phase of the eruption was focussed about
northwest of Mt Haszard is inferred from the isopachs on
Haszard Scoria and Parakeet Tuff (Fig.
A, B) and the
isopleths of maximum scoria size in Haszard Scoria (Fig.
The volcanism which produced Haszard Scoria was
Strombolian at the outset (producing the basal bedded part
of the Haszard Scoria) and rapidly escalated into a climactic,
continuous gas-streaming event which may have been sub-
Plinian, forming the overlying massive scoria deposit.
The gas-streaming eruption induced failure of the
northern sector of the Annexation shield, probably
structurally weakened by the earlier collapse of immediately
adjoining Macauley caldera. Pyroclastic surges accompanied
the sector collapse and emplaced large lithic blocks derived
from the foundering volcano into the Haszard Formation
tephra sequence. The vent was overwhelmed by the sea, and
the eruption style changed to Surtseyan (phreatomagmatic),
dispersing fine wet ash and lapilli over Macauley Island.
Young explosion craters
Four small northeast-aligned pits towards the south of
Macauley Island resemble phreatic explosion craters.
Graeme Taylor (Department of Conservation) inspected
these pits in September 1988 and reported shallow saucer-
shaped depressions, the largest 5-6 m deep and the others
2-3 m deep. Dense sedge covered the ground so that
exposure was poor, but the soil incorporated pumice clasts
from Sandy Bay Tephra, ?Haszard Scoria, and scattered
basalt clasts of 50—500 mm diameter (which is too large for
Haszard Scoria at this site). This mixture of lithologies
suggests reworking, and the deposit may well be a phreatic
explosion breccia. The pits postdate Sandy Bay Tephra and
probably relate to Haszard volcanism.
(Brothers & Martin 1970)
A 30 m thick sequence of lacustrine silt and sand in Grand
Canyon above its ravine-like lower section was deposited
in an ephemeral lake or swamp created by a blockage in the
lower ravine (Brothers & Martin 1970). There are similar
sediments, but of probable fluvial origin, flooring Access
Gully, Lava Cascade, and other similar-sized gullies, which
we also include in Grand Canyon Formation (although too
small to map separately). Grand Canyon sediments consist
predominantly of reworked Haszard Formation tephrae;.
indicating that most of the sediments were laid down after
the eruption of the Haszard Formation. Aggradation of the
Macauley Island drainage system may have been a response
to Haszard volcanism, which provided a source ot
unconsolidated, readily erodible tephra.
The steep cliff on the north side of Macauley Island, exposing
a cross-section through the volcanic succession, indicates
that part of the volcano is missing, and suggests that then1
was major caldera collapse to the north or northwest. This
was confirmed by preliminary bathymetric soundings in
and a detailed bathymetric survey was carried out b>
the RV Vulkanolog in March 1988 (Scott & Lloyd 19891
(Fig. 1).
Although Macauley Island covers only 3 km3, it is the
emergent part of an ESE-elongated volcanic edifice, c. 22 km
long and covering 380 km2 at the 900 m isobath. The
submarine depression immediately west of Macauley Island
is up to 1.1 km deep and has an area of c. 12 km2 at its
deepest part. We infer that this is the caldera that was formed
by the eruption of the Sandy Bay Tephra, with further
collapse on the northern side of Macauley Island in the later
stages of the Haszard eruption.
Unlike other walls of the caldera which have steep,
relatively smooth profiles, the southeast wall of the caldera
beneath Macauley Island is relatively rough and jumbled,
possibly by slumping. The isobaths project as a blunt convex
tongue onto the floor of Macauley caldera (Fig. 1), outlining
what is interpreted as the toe of a large landslide. The
southeast caldera wall and bathymetric relief on either side
of the caldera coincide with a strong NNE alignment. This
structure strikes parallel to dikes and faults on Macauley
Island (as well as the overall trend of the Kermadec Ridge),
and has three major magnetic anomalies associated with it
(Alexandar Ivanenko pers. comm.), suggesting that it could
be a regional fracture that acted as a channel for magma
feeding Macauley volcano. Although it is possible that
slumping of the southeast caldera wall is the result of
faulting, the abundance of large lithic basalt blocks within
the transition from Haszard Scoria to Parakeet Tuff when
the eruption changed from subaerial to submarine, is
Downloaded by [] at 01:14 30 January 2016
Lloyd et al.—Volcanic history of Macauley I.
Fig 13 Development of
Macauley Island. In all diagrams
the present outline of Macauley
Island is shown. A, Macauley
volcano appears above the sea as
a small shield of resistant North
Cliff Lavas. B, A cone of pyro-
clastic ejecta (Boulder Beach
Formation) built over the North
Cliff shield, and shown here with
vent breached by the sea. C,
Eruption of Annexation Lavas
from multiple vent and fissure
sources built a large lava shield,
probably elongated NNE in the
direction of the fault system
controlling upflow of magma.
Subsea eruption northwest of
Annexation shield produces thick
deposits of Sandy Bay Tephra
covering the older shield, and
extending the shoreline to the
vicinity of the 100 m isobath. The
eruption produced Macauley
caldera into which the north-
western flanks of the shield
foundered. E, Marine erosion
had reduced the island to an area
not greatly larger than at present
before commencement of the
Haszard eruption. The surface of
the ignimbrite had been incised
with deep slot canyons. Flows of
Haszard Lavas, probably from
multiple vents, were captured in
the canyons and channelled to the
sea. A scoria and lava cone of Mt
Haszard was constructed around
one vent; another vent located to
the north was the source of much
of the Haszard Scoria. Further
subsidence of northern flanks of
the island along zone X—X' term-
inated the Haszard Scoria erup-
tion, the vent migrated beneath the
sea, and the eruption ended with
a Surtseyan phase.
evidence that major collapse accompanied closing stages of
the Haszard eruption. The drowning of the vent may have
been caused by foundering of the northern flanks of the
Macauley Island is only the tiny emergent top of the large
submarine Macauley volcano, but the surface exposures,
especially on the northern cliffs, allow a reconstruction of
the later phases of volcanism.
Macauley Island came into existence with the eruption
of basaltic lava to form a low, resistant basaltic shield
volcano (North Cliff Lavas), probably north of the present
island (Fig.
Phreatomagmatic eruptions followed and
built a low pyroclastic cone (Boulder Beach Formation)
(Fig. 13B). This explosive magmatic activity implies that
water was able to enter the vent, probably because it was
beneath the sea. Magma then intruded the volcanic pile, and
a few lava flows discharged at the surface. Macauley volcano
then entered a brief quiescent phase when tephra eroded from
upper parts of the cone was redeposited lower down.
Downloaded by [] at 01:14 30 January 2016
308New Zealand Journal of Geology and Geophysics, 1996, Vol. 39
The succeeding episode of intrusion and lava outflow
(Annexation Lavas) built a large (4 km diameter) basaltic
shield to c. 150 m above sea level (Fig. 13C). Towards the
end of this eruptive phase, a deep crater developed in the
vicinity of Mt Haszard.
Explosive eruption of the dacitic Sandy Bay Tephra
caused collapse of part of the submarine Macauley volcano
massif (including the northwestern flanks of the Annexation
lava shield) to create the Macauley caldera. A thick blanket
of nonwelded, dacitic, pumiceous ignimbrite covered the
subaerial part of the volcano and shallow water surrounding
it, greatly expanding the area of the island for a short time
(Fig. 13D). Wave action rapidly eroded the soft ignimbrite
until harder lavas were exposed to form the shore platform.
Narrow slot canyons developed rapidly in Sandy Bay Tephra
during heavy rain, and a weak paleosol developed on the
ignimbrite surface.
Renewed basaltic volcanism commenced with eruption
of flows (Cascade Lavas) flooding the floor of the crater
formed late in the Annexation event, and building a
composite tephra and lava cone around the vent which
eventually overtopped the crater at Mt Haszard. Ponded
Cascade Lavas in the old crater overflowed southwest slopes
of the island; some flows were channelled to the sea down
canyons in the Sandy Bay Tephra. Lava also flowed from
sources on the flanks of the volcano. The eruption climaxed
with a powerful ?sub-Plinian phase from a source north of
the present island, which erupted massive unbedded scoria
(Fig. 13E). The northern part of Macauley Island collapsed
into the nearby Macauley caldera, and the vent was invaded
by the sea (Fig. 13E). During the brief transition from
subaerial to submarine volcanism, airfall, pyroclastic flow,
and base surge tephra emplacement mechanisms were
operative, and the eruption terminated with a Surtseyan
The Haszard eruption may not have been the youngest
eruption from Macauley volcano. A small, ring-shaped island
(Brimstone Island), reputed position 45 km west of
Macauley Island, was seen steaming when observed by
Captain Thazer on 6 September 1825, and a submarine
eruption was reported 22 km NNE of Macauley Island on 1
December 1887 (Wolff 1929). The cited locations of these
eruptions (Fig. 1), in deep water where there is no evidence
of bathymetric
suggests that the positions reported
were probably in error. Macauley is the nearest known
volcano to the reported positions, and it is possible that these
historic eruptions were from its submarine flanks.
EFL and SN are grateful to Brian Bell (formerly Wildlife Division,
Department of Internal Affairs) for an invitation to join the 1980
expedition to Macauley Island, and for efficient handling of the
expedition logistics. We were transported to the island by the New
Zealand Navy. We thank scientists and crew of the Russian research
vessel RV Vulkanolog. Without their co-operation the precise
location of Macauley caldera may have remained unknown.
Bradley Scott provided valuable assistance on the RV Vulkanolog
cruise, and produced several of the diagrams.
IEMS and RBS would like to acknowledge the New Zealand
Lottery Grants Board, who supported fieldwork on Macaule>
Island (Lottery Science grant no. SR 022590), and John Young of
the yacht Blackadder who provided logistic support.
The manuscript was reviewed by Jim Cole, John Latter, Bruce
Thompson, George Walker, Colin Wilson, and Ian Wright, who
are thanked for their constructive comments.
Brothers, R. N.; Hawke, M. M. 1981: The tholeiitic Kermadei
volcanic suite: additional field and petrological data
including iron-enriched plagioclase feldspars. New Zealand
journal of geology and geophysics 24: 167-175.
Brothers, R. N.; Martin, K. R. 1970: The geology of Macauley
Island, Kermadec Group, southwest Pacific. Bulletin:
volcanologique 34 (1): 330-346.
Ewart, A.; Brothers, R. N.; Mateen, A. 1977: An outline of the
geology and geochemistry, and possible petrogeneric
evolution of the volcanic rocks of the Tonga-Kermadec-
New Zealand island arc. Journal of volcanology and
geothermal research 2: 205-250.
Fisher, R. V.; Schmincke, H.-U. 1984: Pyroclastic rocks. Berlin.
Springer-Verlag. 472 p.
Gamble, J. A.; Smith, I. E. M.; Graham, I. J.; Kokelaar, B. P.
J. W.; Houghton, B. F.; Wilson, C. J. N. 1990: The
petrology, phase relations and tectonic setting of basalts
from the Taupo Volcanic Zone, New Zealand and the
Kermadec Island Arc-Havre trough, S.W. Pacific. Journal
of volcanological and geothermal research 43: 253-270
Gamble, J. A.; Smith, I. E. M; McCulloch, M. T.; Graham, I. J.
Kokelaar, B. P. 1993: The geochemistry and petrogenesis
of basalts from the Taupo Volcanic Zone and Kermadec
Island Arc, S.W. Pacific. Journal of volcanology and
geothermal research 54: 265-290.
Latter, J. H.; Lloyd, E. F.; Smith, I. E. M.; Nathan, S. 1992: Volcanic
hazards in the Kermadec Islands and at submarine
volcanoes between southern Tonga and New Zealand
Volcanic hazards information series 4. Wellington, New
Zealand. Ministry of Civil Defence. 44 p.
Lloyd, E. F.; Nathan, S. 1981: Geology and tephrochronology of
Raoul Island, Kermadec Group, New Zealand. New
Zealand Geological Survey bulletin 95: 105 p.
Oliver, W. R. B. 1911: The geology of the Kermadec Islands.
Transactions of the New Zealand Institute 43: 524-535.
Scott, B. J.; Lloyd, E. F. 1989: Scientific cruise of Kermadec Ridge:
R.V. "Vulkanolog". New Zealand volcanological record
17: 6-7.
Smith, S. P. 1888: Geological notes on the Kermadec Group
Transactions of the New Zealand Institute 20: 333-344.
Speight, R. 1896: Notes on rocks from the Kermadec Islands
Transactions of the New Zealand Institute 28: 625-627.
Thomas, A. P. W. 1888: Notes on the rocks of the Kermadec
Islands. Transactions of the New Zealand Institute 20.
F. von 1929: Der Vulkanismus. Stuttgart, Enke. 828 p.
Downloaded by [] at 01:14 30 January 2016
... All of these subaerial exposures have been studied intensively (e.g. Brothers and Martin, 1970;Searle, 1970, Lloyd andNathan, 1981;Smith et al., 1988Smith et al., , 2003aLloyd et al., 1996;Worthington et al., 1999). Apart from the original discovery of submarine volcanic activity in the 1960s (Kibblewhite, 1966) little was known of the largely submarine volcanoes of the Kermadec or Tonga arcs until a resurgence of ship-based science during the late 1980s and 1990s. ...
... Despite the intra-oceanic setting, felsic magmatism appears to be common in the Tonga-Kermadec arc. The occurrence of rhyolitic and dacitic rocks on Curtis and Macauley Islands in the Kermadecs has been known for some time (Lloyd and Nathan, 1981;Smith et al., 1988;Lloyd et al., 1996) and rhyolites, rhyolite tuffs, and pumice breccias were reported in drill core from ODP Leg 135 site 841 in the Tongan forearc (Bloomer et al., 1994). More recently, felsic volcanic rocks have also been dredged from several submarine volcanoes in the southern Kermadecs and multibeam mapping has revealed large, submarine caldera structures similar to those associated with large subaerial felsic volcanoes (Gamble et al., 1997;Wright and Gamble, 1999). ...
... Raoul consists of a series of coalescing stratocones and the centrally-located Den-ham Bay caldera has been the site of numerous eruptions since 2.2 ka, including three since 1814 (Worthington et al., 1999). Similarly, Macauley is dominated by a large, centrally-located submarine caldera that formed during a voluminous eruption at 6.3 ka (Lloyd et al., 1996). Curtis Island to the south of Macauley Island is the remnant of a large pyroclastic flow sheet from an unknown centre and in this respect differs from the other Kermadec Islands. ...
Full-text available
The convergent plate boundary marked by the Tonga–Kermadec arc system is one of the clearest examples of an intra-oceanic arc on the modern Earth. Arising from ground breaking research in the 1970's and continuing through the present, the arc has become a testing ground for defining and explaining models for the tectonic and magmatic evolution of intra-oceanic subduction systems. Despite its simple oceanic setting the arc shows significant variations along its length in terms of rock types and structure and because of the lack of involvement of continental crust these allow refinement of the details of or models of how arcs develop and evolve. Of particular importance are the variations in convergence rates and vectors between the major plates and the effect these have on the tectonic evolution of the arc and back-arc. The erupted rocks of the arc have provided the means of testing the contribution of subducted sediment, slab, mantle wedge and overlying crust and most recently tests of P–T–time paths for recycled sediments. Although the arc is dominated by basalts and basaltic andesites, there is increasing evidence for significant proportions of silicic magmas erupted in recent times. Silicic volcanism is also evidenced in the presence of caldera structures revealed by detailed multibeam mapping that have been a feature of research in the last decade. As we enter the 21st century, the Tonga–Kermadec arc continues as an archetypical natural laboratory for the testing of new ideas in subduction zone tectonics and petrology.
... The arc consists dominantly of submarine volcanoes, with only Raoul, Macauley, Curtis and L'Esperance volcanoes being partially emergent (Fig. 1). The occurrence of silicic explosive volcanism along the arc has long been apparent in subaerial deposits (e.g., Brothers and Martin, 1970;Lloyd and Nathan, 1981;Lloyd et al., 1996). However, not until detailed bathymetric mapping and submarine sampling was undertaken was it recognised that submarine caldera-related, silicic explosive volcanism is widespread (e.g., Wright and Gamble, 1999;Haase et al., 2002;Wright et al., 2002Wright et al., , 2003Wright et al., , 2006Graham et al., 2008). ...
... However, not until detailed bathymetric mapping and submarine sampling was undertaken was it recognised that submarine caldera-related, silicic explosive volcanism is widespread (e.g., Wright and Gamble, 1999;Haase et al., 2002;Wright et al., 2002Wright et al., , 2003Wright et al., , 2006Graham et al., 2008). The Kermadec arc volcanoes offer the opportunity to compare and contrast eruptions in the subaerial and deep submarine environments as they have explosively erupted silicic magmas within the last 10 kyr (Lloyd and Nathan, 1981;Lloyd et al., 1996;Worthington et al., 1999;Wright, 2001;Smith et al., 2003aSmith et al., ,b, 2006Wright et al., 2003Wright et al., , 2006Barker et al., 2012Barker et al., , 2013Carey et al., 2014;Rotella et al., 2014). Samples for this study were acquired during the 2007 voyage of the R.V. Tangaroa (TAN0706) by seafloor dredging at Healy and Raoul SW volcanoes, and by the H.M.N.Z.S. Canterbury shipboard crew and scientists on 9 August 2012 from a floating pumice raft which resulted from the 19 to 20 July eruption of Havre volcano. ...
... If the eruption jet breaches the sea surface while the pyroclasts are still above the glass transition temperature, such as would be the case for powerful eruptions in shallower water, then an additional sudden decrease in pressure would be experienced by the pyroclasts and an additional nucleation event could occur. Such may be the case for the 6.3 C 14 ka Sandy Bay Tephra eruption of Macauley volcano, which erupted through~400 m of water and emplaced up to 100 m of cross-bedded, non-welded, poorly stratified dacite ignimbrite, as seen on the 3 km 2 subaerial exposure of the volcano's eastern flank (Lloyd et al., 1996;Barker et al., 2012). The Sandy Bay eruption is interpreted to have interacted extensively with water, given the evidence for cool emplacement of ignimbrite on Macauley Island such as lack of oxidation coloration, ubiquitous ash coatings on pumice pyroclasts and lack of a pyroclastic fall deposit. ...
Despite increasing recognition of silicic pumice-bearing deposits in the deep marine environment, the processes involved in explosive silicic submarine eruptions remain in question. Here we present data on bubble sizes and number densities (number of bubbles per unit of melt matrix) for deep submarine erupted pumices from three volcanoes (Healy, Raoul SW and Havre) along the Kermadec arc (SW Pacific) to investigate the effects of a significant (>~1 km) overlying water column and the associated increased hydrostatic pressure on magma vesiculation and fragmentation. We compare these textural data with those from chemically similar, subaerially-erupted pyroclasts from nearby Raoul volcano as well as submarine-erupted ‘Tangaroan’ fragments derived by nonexplosive, buoyant detachment of foaming magma from Macauley volcano, also along the Kermadec arc. Deep submarine-erupted pumices are macroscopically similar (colour, density, texture) to subaerial or shallow submarine-erupted pumices, but show contrasting microscopic bubble textures. Deep submarine-erupted pyroclasts have fewer small (<10 μm diameter) bubbles and narrower bubble size distributions (BSDs) when compared to subaerially erupted pyroclasts from Raoul (35–55 μm vs. 20–70 μm range in volume based median bubble size, respectively). Bubble number density (BND) values are consistently lower than subaerial-erupted pyroclasts and do not display the same trends of decreasing BND with increasing vesicularity. We interpret these textural differences to result from deep submarine eruptions entering the water column at higher pressures than subaerial eruptions entering the atmosphere (~10 MPa vs. 0.1 MPa for a vent at 1000 mbsl). The presence of an overlying water column acts to suppress rapid acceleration of magma, as occurs in the upper conduit of subaerial eruptions, therefore suppressing coalescence, permeability development and gas loss, amounting to closed-system degassing conditions. The higher confining pressure environment of deep submarine settings hinders extensive post-fragmentation clast expansion, coalescence of bubbles, and thinning of bubble walls, causing clasts to have similar BND values regardless of their vesicularity. Although deep submarine-erupted pyroclasts are closely similar to their subaerial counterparts on the basis of bulk vesicularities and macroscopic appearance, they differ markedly in their microscopic textures, allowing them to be fingerprinted in modern and ancient pumiceous marine sediments.
... Geological setting Macauley Island is located in the intra-oceanic Kermadec Arc (Fig. 1) and is the uppermost subaerial edifice of an active submarine stratovolcano (Wright et al., 2006;Shane & Wright, 2011;Barker et al., 2012). Onshore Macauley Island, deposits are dominated by basaltic lavas and phreatomagmatic deposits (Lloyd et al., 1996). These are overlain by the Sandy Bay Tephra Formation and younger basaltic lavas (Lloyd et al., 1996). ...
... Onshore Macauley Island, deposits are dominated by basaltic lavas and phreatomagmatic deposits (Lloyd et al., 1996). These are overlain by the Sandy Bay Tephra Formation and younger basaltic lavas (Lloyd et al., 1996). The Sandy Bay Tephra is the only confidently known silicic activity related to this volcano and consists of a massive lithic basal unit overlain by multiple wet pyroclastic density current and surge deposits (Smith et al., 2003;Barker et al., 2012). ...
A comparative analysis of bedform fields along the submarine flanks of insular volcanoes, characterized by different morpho‐structural settings, volcanic and meteo‐marine regimes (Vanuatu, Kermadec, Bismark, Madeira and Aeolian archipelagos), is presented here to provide insights on the size distribution, morpho‐dynamic and genesis of such bedforms. Two main types of bedforms are recognized according to their size, location and preconditioning/triggering processes. Small‐scale bedforms have wavelengths of tens to hundreds of metres and wave heights of metres. Because of their small‐size, they are typically not recognizable at water depths greater than 400 m from vessel‐mounted bathymetric surveys. However, few examples of small‐scale bedforms are reported from upper volcanic flanks, where steep gradients commonly hinder their formation. Their recognition is mostly limited to the thalweg of shallow and flat‐bottomed channels that carve the insular shelf on slope gradients <15°. Small‐scale bedforms are mostly related to erosional–depositional processes due to sedimentary gravity flows that are often the result of a cascading effect between volcanic and non‐volcanic processes (for example, flood discharges and retrogressive landslides). Large‐scale bedforms occur at all water depths, having wavelengths of hundreds/thousands of metres and wave heights up to few hundreds of metres. The origin of large bedforms is more difficult to ascertain, especially if only bathymetric data are available. Some diagnostic criteria are presented to distinguish between bedforms associated with landslide deposits and those associated with density currents. In this latter case, relevant sediment sources and slope gradients (<8°) are key factors for bedform development. Erosional–depositional bedforms are typically related to eruption‐fed density flows formed during large caldera collapses or to large turbidity flows. Bedforms generated by turbidity flows are often observed in the lower volcanic flanks, where an abrupt decrease of gradients is present, often matching a change from confined to unconfined settings. In summary, this study provides insights to interpret bedforms in modern and ancient marine volcaniclastic settings elsewhere.
... Macauley Caldera is situated at 30° 12' S in the central part of the Kermadec Arc et al. (2006). The caldera has formed about 6200 years ago during the Sandy Bay tephra eruption (Lloyd et al., 1996). While the entire volcano is mainly composed of basaltic rocks (Lloyd et al., 1996), the caldera itself consists of rocks with dacitic to rhyolitic compositions. ...
... The caldera has formed about 6200 years ago during the Sandy Bay tephra eruption (Lloyd et al., 1996). While the entire volcano is mainly composed of basaltic rocks (Lloyd et al., 1996), the caldera itself consists of rocks with dacitic to rhyolitic compositions. ...
This work presents investigations of high-T hydrothermal vent fluids, associated mineral precipitates and volcanic rocks from three submarine volcanos in the South Kermadec intraoceanic arc (Brothers Caldera, Haungaroa Volcano, Macauley Caldera). Mineralogical investigations, vent fluid compositions and fluid inclusion studies are discussed. Reaction path models simulate the influence of host rock compositions and the integrated fluid flux on the formation of massive sulfides. Further models demonstrate how addition of magmatic SO2 to hydrothermal systems generates arc typical magmatic-hydrothermal fluids and influences their REE composition. Vent fluid compositions, as well as salinities and formation temperatures in mineral precipitates from seawater-dominated hydrothermal fluids, point to boiling processes of rising solutions. The investigations demonstrate how magma degassing and boiling are key processes responsible for the chemical diversity of arc hydrothermal fluids.
... 33 km 2 (Mortimer and Campbell 2014). The cones have had a highly dynamic recent history with explosive emergence and collapse (Lloyd et al. 1996). In spite of geothermal activity and remote- ness, the Kermadec Ridge and adjacent trench is inhabited by over 2,000 taxa (Duffy and Ahyong 2015) including 397 species of fishes (Te Papa unpublished records), of which nine are known to be endemic to the shelf (0-200 m depth). ...
Full-text available
Flexorincus , new genus and species, is described from 15 specimens (14.0–27.2 mm SL) collected from shallow (0–9 meters) intertidal and sub-tidal waters of the Rangitāhua Kermadec Islands, New Zealand. The new taxon is distinguished from all other members of the Gobiesocidae by a combination of characters, including a heterodont dentition comprising both conical and distinct incisiviform teeth that are laterally compressed with a strongly recurved cusp, an oval-shaped opening between premaxillae, a double adhesive disc with a well-developed articulation between basipterygia and ventral postcleithra, and many reductions in the cephalic lateral line canal system. The new taxon is tentatively placed within the subfamily Diplocrepinae but shares a number of characteristics of the oral jaws and the adhesive disc skeleton with certain members of the Aspasminae and Diademichthyinae.