Volcanoes buried in Te Riu-a-Māui/Zealandia sedimentary basins

ABSTRACT

Many volcanoes are buried within Te Riu-a-Māui/Zealandia sedimentary basins. Based on a large collection of seismic reflection and drillhole data, this paper outlines the morphology, location and age of over 400 volcanoes buried and preserved in offshore New Zealand basins. We discuss the relationships of these volcanoes to the timing and locations of the main tectonic events that shaped Zealandia in the last 105 Myr. Late Cretaceous volcanism mainly formed large (>20 km³) composite cones and volcanic complexes that erupted along rift faults related to the break-up of Zealandia from south-eastern Gondwana. Cenozoic intraplate volcanic activity was widespread, typically forming clusters of scattered small-volume (<1 km³) craters and cones. In contrast, large Cenozoic volcanic complexes, composite cones and shield volcanoes erupted along linear arcs associated with subduction along the Pacific-Australian plate boundary. In detail, the presence of pre-existing structures and types of host rocks also influence the passage of magma in the shallow (<5 km) crust and the location of eruptive centres, regardless the tectonic setting of eruption. Insights from this work demonstrate the value of characterising volcanoes buried in sedimentary basins for improved understanding of the interplay between volcanism and plate tectonics in Zealandia and further afield.

KEYWORDS:

• Buried volcanoes
• seismic reflection
• Te Riu-a-Māui/Zealandia
• rift volcanism
• intraplate volcanism
• subduction-related volcanism
• plate tectonics

Introduction

Over the last 150 years, geological mapping and studies of igneous rocks from onshore New Zealand, New Caledonia and several surrounding small islands have provided an enormous amount of information about volcanic activity in Te Riu-a-Māui/Zealandia region (e.g. Hochstetter 1867; Marshall 1932; Mortimer and Scott 2020). More recently, onshore information has been supplemented by oceanographic data from dredging the sea floor above submarine seamounts, ridges and plateaus across Zealandia (e.g. Timm et al. 2010; Bache, Sutherland, et al. 2014; Hoernle et al. 2020). These complementary datasets are the basis for current geological models describing the syn-and post Gondwana break-up volcanism in Zealandia (e.g. Hoernle et al. 2006; Mortimer et al. 2018; Van der Meer et al. 2018). However, to date, few studies have had incorporated detailed information from volcanoes buried within Zealandia sedimentary basins (e.g. Herzer 1995; Uruski 2019; Barrier 2019).

From coastal exposures, geologists have long suspected that large eroded volcanoes could extend beneath the New Zealand continental shelf and possibly beyond. It was not until the 1960s that igneous rocks were first documented within sedimentary basins in offshore New Zealand. Initially, recognition of these buried igneous systems was inferred by remote sensing techniques such as seismic reflection, gravimetric and magnetometric surveys acquired during petroleum exploration campaigns (e.g. Hatherton 1968; Knox 1982; Hayward 1987). Oil and gas discoveries within the onshore Taranaki Basin stimulated further offshore exploration, including the acquisition of new seismic surveys and drillholes (Figure 1). Many of these offshore drillholes encountered volcanic and plutonic rocks of various ages confirming an igneous origin of the anomalies recognised from geophysical data (e.g. Milne 1975; Field et al. 1989; Bergman et al. 1992).

Figure 1. Map showing the seismic dataset, main regional physiography, sedimentary basins, and location of offshore volcanoes presented in this study. Basins highlighted in darker yellow are discussed in this paper. Numbers and abbreviations are: (1) Vulcan-Romney volcanic zone; (2) Tikati volcanic complex; (3) Matuku intrusions and small vents; (4) Kaiwero volcanic field; (5) Northland-Mohakatino volcanic belts; (6) Aotea volcanic complex; (7) West Ngatutura volcanic field; (8) Sloop volcanic complex; (9) Barque volcanic complex; (10) Tapuku East volcanic complex; (11) East Waiareka-Deborah volcanic field; (12) Papatowai volcanic field; (13) Maahunui volcanic field; (14) Waka intrusions and small vents; (15) Toroa volcanic field; (TVZ) Taupo Volcanic Zone; (T. Kings R.) Three Kings Ridge. Volcanoes numbered in grey circles are only shown in the supplementary material of this paper. Background image shows a Digital Terrain Model from NIWA (New Zealand Institute of Water and Atmosphere).

In the past few decades, an increasing quality and coverage of seismic surveys onshore and offshore New Zealand have enabled us to apply a systematic methodology for establishing the locations, timing and morphologies of a variety of igneous-dominated structures belonging to diverse magmatic systems (i.e. rift, intraplate, and subduction-related; Figure 2). This paper builds on and extends existing knowledge mainly from Field et al. (1989), Herzer (1995), King and Thrasher (1996), Giba et al. (2013), Seebeck et al. (2014), Bischoff (2019), Barrier (2019) and Uruski (2019), proposing a systematic classification of the best understood examples of volcanoes buried offshore New Zealand. We focus mainly on describing volcanoes buried in the Reinga-Northland, Taranaki, Deepwater Taranaki, Canterbury and Great South basins, where seismic reflection and drillhole data are present in great quantity (Figure 1). Additional volcanoes are likely buried in other parts of Zealandia (e.g. Norfolk Ridge, Lord Howe Rise, and west of the Puysegur Trench), but because of limited seismic and drillhole coverage, these areas were excluded from this study. Tables 1 and 2 show the nomenclature and main references for the volcanoes presented here. Additional information is available in Supplementary Materials 1–9.

Figure 2. Simplified Late Cretaceous-Cenozoic volcano-stratigraphic framework and main tectonic events offshore New Zealand. Symbols in black indicate the volcanic type and are: composite cone (CC), volcanic complex (VC), volcanic field (VF), caldera (Ca), shield (Sh), intrusions and small-volume vents (IV). Colours in the bars correspond to the active phase of the volcanoes and their tectonic setting of eruptions and are: syn-rift (dark green), immediate post-rift (green), scattered intraplate (blue), subduction-related (red). Numbers in red are the estimated volume of erupted magma in km³. Vertical lines represent uncertainties in the age of the volcanoes.

Table 1. Nomenclature and main references of the volcanoes buried in the Taranaki, Deepwater Taranaki and Reinga-Northland basins.

This study		Uruski (2019)	Zhu et al. (2017)	King and Thrasher (1996)	Herzer (1995)	Brodie (1965)
Vulcan-Romney volcanic zone	Chronos composite volcano	–	Coopworth volcano	–	–	–
	Rhea composite volcano	Vulcan-Hestia volcanoes	Hestia-Vesta volcanoes	–	–	–
	Coopworth volcanoes	–	Coopworth volcano	–	–	–
	Romney volcanic field	Romney Volcano	Romney volcanic complex	–	–	–
	Hades caldera	Vulcan-Hestia volcanoes	Hestia-Vesta volcanoes	–	–	–
	Vulcan composite volcano			–	–	–
	Hestia composite volcano			–	–	–
Tikati volcanic complex		Tikati volcanic field	Corriedale-East v.f. and Corriedale v.c.	–	–	–
Matuku intrusions and small vents		–	–	–	–	–
Kaiwero volcanic field		Paleogene mounds	–	–	–	–
Northland-Mohakatino volcanic belts		–	–	Mohakatino Volcanic Centre	Northland Volcanic Belt	–
Aotea volcanic complex		Aotea Seamount	–	–	–	Aotea Seamount
West Ngatutura volcanic field		Pliocene volcanic field	–	–	–	–

Notes: Some of these volcanoes have been previously referred to as petroleum exploration targets, which are not discussed in detail in this paper. Cretaceous volcanic rocks in the area of the Matuku-3D are reported in Higgs and Strogen (2014).

Table 2. Nomenclature and references of the volcanoes buried in the Canterbury and Great South basins.

This study	Barrier (2019)	Bischoff et al. (2019)	Blanke (2012)	Beckman (2012)	Field et al. (1989)
Sloop volcanic complex	Sloop volcanic complex	–	–	Sloop volcano	Galleon volcanics
Barque volcanic complex	Barque volcanic complex	–	–	–	Galleon volcanics
Tapuku East volcanic complex	Tapuku East volcanic complex	–	–	Tapuku East volcanics	–
Papatowai volcanic field	–	–	–	–	–
East Waiareka-Deborah volcanic field	Paleocene to Oligocene volcanoes	–	–	–	Endeavour volcanics
Maahunui volcanic field	Maahunui volcanic field	Maahunui volcanic field	–	–	Volcanics of Banks Peninsula
Waka intrusions and small vents	Igneous intrusions	–	Saucer-shaped sills	–	–
Toroa volcanic field	–	–	–	–	–

Note: At the region of the Papatowai volcanic field, Waight et al. (2019) mention possible Cenozoic volcanic rocks based in interpretation of aeromagnetic data.

The results of this study may have an application to those undertaking detailed investigation of volcanoes or volcanic fields buried in sedimentary basins, for improved understanding of regional volcanism across Zealandia and Gondwana, and for development of models relating volcanism and tectonics.

Geological Background

Zealandia is a region of 4.9 million km² largely (ca 95%) submerged in the Pacific Ocean and Tasman Sea (Mortimer and Campbell 2017). It consists of Cambrian-Early Cretaceous basement overlain by Late Cretaceous-Cenozoic sedimentary and igneous rocks (Laird and Bradshaw 2004; Mortimer et al. 2018). For most of the Mesozoic, Zealandia formed the active convergent south-eastern margin of the Gondwana supercontinent, which separated from Australia and Antarctica in the Late Cretaceous (e.g. Laird 1993; Sutherland et al. 2001; Mortimer 2004). An initial phase of crustal extension (ca 105–83 Ma) leading to continental break-up gave rise to several rift basins throughout Zealandia. These basins were subsequently filled with Late Cretaceous and younger strata up to 10 km thick. In and around the New Zealand landmass, these basins were affected by Eocene to Recent plate boundary deformation associated with the Hikurangi and Puysegur subduction zones, and dextral strike-slip transpression along the Alpine Fault (e.g. Walcott 1978; King et al. 1999; Nicol et al. 2007; Figures 1 and 2).

Most sedimentary basins within Zealandia have experienced recurrent and widespread igneous activity (e.g. Field et al. 1989; King and Thrasher 1996; Sahoo et al. 2015). Late Cretaceous-Cenozoic magmatism formed numerous volcanic zones, complexes, fields and centres. These volcanoes are part of the Zealandia Megasequence and are grouped into the Rūaumoko Volcanic Region, broadly classified as having an intraplate (Horomaka Supersuite) or a subduction-related (Whakaari Supersuite) origin (Mortimer et al. 2014).

Intraplate volcanoes are predominantly basaltic to trachytic and phonolitic in composition, typically forming clusters of scattered small-volume volcanoes (<1 km³), and less common large polygenetic volcanic systems not associated with plate boundaries or hotspot tracks. These include those of Waiareka-Deborah and Auckland volcanic fields (e.g. Coombs et al. 1986; Kereszturi and Németh 2016; Hopkins et al. 2020), as well as the large volcanic complexes that shaped the Otago and Banks peninsulas (e.g. Sewell 1988; Weaver and Smith 1989; Scott et al. 2020). In contrast, the convergent Australian-Pacific plate boundary formed linear, mainly calc-alkaline basalt, andesite, dacite and rhyolite-dominated volcanic arcs extending from oceanic crust north of Zealandia onto continental crust towards the North Island of New Zealand (e.g. Cole and Lewis 1981; King 2000; Sutherland et al. 2017). The arcs broadly show the migration of eruptive centres to the SE, formerly extending along the Three Kings Ridge to Northland region (ca 25–17 Ma), and from the Lau-Colville Ridge across Coromandel Peninsula to the Mohakatino belt (ca 16–4 Ma). The youngest modern arc (ca 2–0 Ma) extends southwards from the Tonga-Kermadec trend into the Bay of Plenty and Taupo Volcanic Zone (e.g. Herzer 1995; Mortimer et al. 2010; Seebeck et al. 2014; Figures 1 and 2).

Data and methods

Data used in this paper has been mainly sourced from the 2018 New Zealand Petroleum and Minerals Exploration Data Pack, which includes a large collection of reports, maps, drillhole records, 2D and 3D seismic surveys. In this study, buried volcanoes were identified using a combination of seismic reflection and drillhole data interpretation. The spacing of the seismic lines is typically <2 km, with 3D datasets having line spacing of ≤12.5 m (Figure 1) and vertical resolution is usually a few tens of metres. Drillhole data commonly include wireline-logs together with lithological, geochemical, geochronological, and biostratigraphic information from cutting samples and sometimes from drill cores.

The age of the volcanic rocks in the subsurface is rarely determined by radiometric dating and is primarily achieved by mapping chronostratigraphic surfaces that correlate seismic anomalies possibly of igneous origin with biostratigraphic markers identified in the drillholes. Chronostratigraphic mapping follows seismic and sequence stratigraphic principles such as stratal reflection relationships and depositional trends within seismic facies (e.g. Mitchum and Vail 1977; Hunt and Tucker 1992; Catuneanu 2006), giving time resolution in the order of few (<5 Myr), similar to the precision of radiometric ages. Interpretation of the environment in which the eruptions occurred was determined by plotting the location of volcanoes on paleogeographic maps of similar age (Arnot et al. 2016), and/or by calibration with paleoenvironmental data obtained from microfossils from drillholes across the studied areas.

Detailed characterisation of the buried volcanoes is based on criteria such as the geometry, internal and external configuration of seismic reflections, deformation of enclosing strata, and stratal relationships within the seismic images (Mitchum and Vail 1977), together with analysis of the volcanic and sub-volcanic architecture (e.g. Bischoff et al. 2017, Bischoff, Nicol, Cole, et al. 2019, Bischoff, Nicol, Barrier, et al. 2019, Bischoff, Rossetti, et al. 2019). Optimal characterisation can be obtained by applying a range of techniques such as seismic attribute analysis, opacity rendering of the seismic signal, decomposition and merging of an array of frequencies of the seismic wavelength spectrum, and 3D mapping of geobodies with the similar seismic response (e.g. Chopra and Marfurt 2005; Planke et al. 2017; Marfurt 2018).

The height of the buried volcanoes is estimated from their present-day morphology (i.e. after erosion and compaction during burial) by recording the distance between the pre- and post-eruptive surfaces in two-way-travel-time (TWT) in seconds and multiplying it by the estimated acoustic velocity of the volcanic rocks within the seismic anomaly. Massive and unaltered basalts typically have acoustic velocities between 5000 and 7000 m/s, while velocities of non-welded stratified pyroclastic rocks are commonly around 3000 m/s (e.g. Planke et al. 2000; Holford et al. 2012; Klarner and Klarner 2012). These values were calibrated with data available in the literature (e.g. Heap and Kennedy 2016; Millett et al. 2016; Cant et al. 2018), and when possible, with velocities of volcanic rocks recorded in wireline-logs from drillholes penetrating the buried volcanoes (e.g. Milne 1975; Field et al. 1989; Rad 2015). Using our analysis, each buried volcanic edifice was approximated as a cone, a trapezoidal prism, or as a spherical cap to roughly estimate the volume of erupted material. As explained in the literature (e.g. Magee et al. 2019; Reynolds et al. 2016; Bischoff, Nicol, Barrier, et al. 2019), these estimations are only approximate, but variations in the acoustic velocities of volcanic rocks do not affect the first-order interpretations of volcano morphologies.

Northern Zealandia

Vulcan–Romney volcanic zone – Late Cretaceous

Vulcan–Romney volcanic zone comprises seven large Late Cretaceous volcanoes, as well as several small vents and numerous intrusive bodies presently buried beneath ca 4000 m of sedimentary strata in the Deepwater Taranaki Basin (Figures 1–3). This group of volcanoes has been named in petroleum reports and the literature after gods and goddesses of Greek mythology, as well as after sheep breeds common in New Zealand. We follow this nomenclature where possible, and extend it to describe volcanoes not previously named (Table 1).

Figure 3. (a) Regional map of the post-eruptive surface of the Vulcan-Romney volcanic zone (ca 68 Ma). Numbers are: (1) Chronos composite volcano; (2) Rhea composite volcano; (3) Coopworth volcanoes; (4) Romney volcanic field; (5) Hades caldera; (6) Vulcan composite volcano; and (7) Hestia composite volcano. (b) 2D cross-section across the Romney volcanic field. Vertical black arrows indicate the location of eruptive vents. Blue lines correspond to fault planes and red lines to the location of possible conduits. (c) Plan view map of Coopworth volcanoes and Romney volcanic field. Dots are located at the position of eruptive vents. Dashed black lines mark the terminal front of lava-flow deposits and the boundaries of the volcanic fields. (d) Detailed structural map of the post-eruptive surface of the Hades caldera at ca 74 Ma.

Late Cretaceous volcanic activity in the Deepwater Taranaki Basin was initially inferred by the interpretation of seismic reflection lines (McAlpine 1999; Uruski and Warburton 2010). In 2014, the Romney-1 drillhole penetrated a sequence of volcanic rocks interbedded with coal measures of Haumurian age (83.6–66 Ma), as well as sandstones and mudstones of the Rakopi Formation, confirming a volcanic origin of the anomalies observed in seismic imagery. These volcanic rocks are overlain by shallow marine sandstones and siltstones of the latest Cretaceous North Cape Formation, indicating that eruptions occurred near paleo-shoreline (Schiøler et al. 2014; Uruski 2019). Drill cuttings were initially identified as moderately- to highly-altered tuffs and epiclastic rocks of basaltic composition (Rad 2015), but our petrographic, SEM and EDS analysis indicates that both mafic and felsic (possibly alkaline) rocks are present in the Romney-1 drillhole (Supplementary Material 1).

Vulcan-Romney volcanoes erupted in two phases (Zhu et al. 2017). Eruptions in the Chronos volcano mark the onset of volcanic activity in the area at ca 85 Ma. The present-day morphology of this volcano is semi-conical, 4.5 km wide across the base of the cone and ca 900 m high. Rhea erupted shortly after Chronos, and also forms a semi-conical edifice at least 7 km in width and ca 1400 m high (Figures 3 and 4). The large volume of each edifice (>20 km³) suggests that they were constructed by multiple eruptive cycles, rather than a single volcanic episode. Both volcanoes are aligned in an NW orientation parallel with the contemporaneous trend of the Zealandia Rift (Strogen et al. 2017).

Figure 4. Uninterpreted (a) and interpreted (b) 2D seismic section showing the morphology and stratigraphic relationships of some of the large volcanoes of the Vulcan-Romney volcanic zone. Po is the post-eruptive surface of each volcano, while Pr is the pre-eruptive surface, these surfaces are the upper and lower bounds of each coloured polygon. Numbers indicate the stratigraphic order relative to the onset of volcanic activity in each volcano. Note that some volcanoes interfinger (e.g. Vulcan and Hestia) indicating that volcanic activity was contemporaneous, while in others (i.e. Coopworth and Hades) the Pr and Po amalgamate with increasing distance from the eruptive centre, indicating that eruptions were relatively short-lived compared to the rate of sedimentary and volcanic processes that control the local basin infill. Blue lines correspond to fault planes and red lines indicate possible magma conduits. The location of this line is shown in Figure 3.

The onset of the second eruptive phase is marked by eruptions in the Coopworth volcanoes at ca 74 Ma (Figures 3 and 4), which comprise two main small-volume cones and minor scattered vents. The northern volcano is smaller (1000 m across and 160 m high) than the southern (1800 m across and 200 m high). The small relief edifices are associated with large peripheral aprons, suggesting a morphology observed in fields of scattered spatter cones and extensive lava fields, a characteristic observed in basaltic monogenetic volcanic fields (Németh 2010).

Romney volcanic field erupted shortly after Coopworth volcanoes and comprises a large shield volcano, as well as numerous (possibly hundreds) scattered small-volume vents and several shallow intrusions typically of saucer-shaped morphology (Figures 3(c) and 5). The large shield volcano is 15 km in width and ca 400 m high as preserved in the basin. Its flanks contain many parasitic vents preferentially orientated parallel with normal faults (Figure 5(b)), which globally is commonly observed in volcanoes associated with rifting and extension (e.g. Galland et al. 2014; Burchardt et al. 2018; Vries and Vries 2018). The extremities of the shield volcano show high-amplitude reflections with ‘toe-like’ terminations (Figure 5(c)), a morphology typically associated with both Pahoehoe and a′ā lava flows sourced in Hawaiian eruptions (Gregg 2017). In contrast to the shield volcano, the scattered small-volume vents tend to show cone-type morphology with extensive flat-lying aprons of high-amplitude reflections that possibly correspond to spatter cones and related lava fields (Figure 3). These vents are preferentially aligned with NW-SE to NE-SW normal faults (Figure 3(c)). In the western part of the Romney field, there is a narrow NE trending structure that resembles a fissure vent (Figure 5(a)), which is also located along a normal fault, reinforcing the interpretation of volcanic activity simultaneously with extension, as commonly described in rift volcanoes (Sigmundsson et al. 2018; Neal et al. 2019).

Figure 5. (a) Plan view map of the Romney volcanic field revealed by spectral decomposition at the 4100 ms time-slice. (b) Opacity rendering of the Romney-3D showing the morphology of a double-summit shield-like volcano composed by several eruptive vents. (c) Time-slice section at 4200 milliseconds coupled with a 2D cross-section across the shield volcano shown in (b). Note the ‘bird-foot’ morphology highlighting the termination of lava flows, while the concentric structures mark the location of conduits. S1 and S2 are mapped summits of the shield volcanoes. WCTRT is the West Coast-Taranaki rift trend. Note the structural control of faults in the location of eruptive vents.

Volcanic activity in Hades caldera occurred simultaneously or immediately after eruptions in Romney volcanic field, at around 73 Ma. Hades has a semi-circular structure 10 km across and ca 800 m high with a central depression 3.5 km wide and 1 km deep bounded by ring faults (Figures 3(d) and 4). Its bedded circular flanks suggest that material was spread relatively uniformly and deposited in tabular layers around the vent, as is common with high energy pyroclastic eruptions (Kereszturi and Németh 2013), while a central depression and ring faults are frequently observed in explosive and non-explosive caldera volcanoes (Cole et al. 2005; Martí et al. 2008). The moderate amplitude and bedded reflections on Hades flanks suggest deposition of fragmented material more likely sourced by explosive eruptions.

Vulcan and Hestia are two large long-lived composite volcanoes that interfinger with each other and with basin strata deposited over at least 8 Myr (Figure 3(a)). Vulcan is 25 km in width and ca 1500 high, while Hestia is 22 km wide and ca 1000 m high. Volcanic activity started in Vulcan at ca 74 Ma and ended while Hestia was still active, which is confirmed by the western flank of Hestia overlapping the eastern flank of Vulcan (Figure 4). Eruptions ceased in Hestia at around 68 Ma, marking the end of volcanism in the Vulcan-Romney volcanic zone. The magma composition and eruptive mechanisms that constructed Vulcan and Hestia remain poorly calibrated in absence of drillholes, but possibly include pyroclastic and lava-flow deposits and volcanogenic sedimentary rocks, as common of large long-lived composite volcanoes (Manville et al. 2009; de Silva and Lindsay 2015). Both volcanoes are located along N-S trending faults parallel with the trend of the West Coast-Taranaki rift (Strogen et al. 2017).

Kaiwero volcanic field – Eocene

The Kaiwero volcanic field is a cluster of >50 scattered Eocene small-volume volcanoes and several shallow (<1 km) intrusions currently buried by 300–1500 m of sedimentary strata offshore western North Island, New Zealand (Figure 6). The name is derived from the Māori for the large submarine plateau (Challenger Plateau) adjacent to the area where the volcanoes are located (Figure 1 and Table 1).

Figure 6. (a) Map showing the distribution of volcanoes in the Kaiwero volcanic field and available 2D seismic grid. The coloured bars show the location and present burial depth of the volcanoes in TWT. The area highlighted in pink corresponds to the limit of occurrence of intrusive and extrusive bodies of Eocene age. (b) 2D seismic section (line MER10-004) showing a small cone-type volcano located at the tip of a saucer-shaped intrusion. Both the pre-eruptive surface and forced folding above the intrusions indicate an Eocene age. Red lines represent the location of possible magma conduits. (c) 2D seismic section (line DTB01-20) showing a small cone-type volcano and an intrusion, both of Eocene age. Numbers are: (1) intrusions; (2) jacked-up dome; (3) conduits; (4) seismic artefact (loss of reflectivity); (5) cone-type volcano.

Volcanism occurred over an area of ca 17,000 km², extending from the NE boundary of the Challenger Plateau towards the SW Deepwater Taranaki Basin. Volcanic rocks of Late Eocene age are reported in western Challenger Plateau (Nelson et al. 1986; Uruski 2019), and several drillholes in the western Taranaki Basin (e.g. Amokura-1, Pateke-1 and Kiwi-1) penetrate a unit rick in bentonite clays at the top of the Turi Formation, referred to in petroleum exploration reports as ‘the ash unit’.

The volcanoes erupted onto a low-gradient submarine sea floor dipping N to NE, in paleo-water-depths ranging from 50 to 1500 m (Figure 6). Most volcanoes are cone-type, <1 km wide and <200 m high. Some edifices appear to be formed by overlapped cones up to 5 km wide and ca 300 high. A single small (<1 km across) crater-type volcano was observed in the central part of the field (line DTB01-31). The volcanoes are often located above the tips of saucer-shaped intrusions or associated with high-amplitude reflections that cross-cut the Cretaceous to Paleocene sedimentary sequence.

The association of vents with large aprons of high-amplitude reflections suggests eruptions of low viscosity lavas, which together with the internal mound and rubble morphology of the vents, may indicate that in some cases they correspond to submarine pillow mounds. The volcanoes that show bedded flanks may have been formed by submarine pyroclastic eruptions (e.g. submarine Strombolian and eruption-fed density currents), which is consistent with the widespread occurrence of the ‘ash unit' in drillholes over 100 km to the SE. The scattered distribution of small-volume volcanoes gives the Kaiwero volcanic field a characteristic observed in monogenetic volcanic fields (Cas et al. 1989; Németh 2010; Kereszturi and Németh 2013). This volcanic field does not show a relationship with plate boundaries or pre-existing crustal structures and it is likely that to have an intraplate origin consistent with the Horomaka Supersuite (Mortimer et al. 2014).

Northland-Mohakatino volcanic belts – Early Miocene to Recent

The volcanoes of the Northland-Mohakatino volcanic belts have been identified by geophysical tomography since the 1960s (Hatherton 1968; Table 1). Several drillholes (beginning with e.g. Mangaa-1 in 1968, and including Kora-1–4, Albacore-1, and Tangaroa-1) drilled into volcanic rocks of Miocene age, confirming a volcanic origin of the geophysical anomalies. Volcanoes in these belts are considered to be of subduction-related origin (Whakaari Supersuite; Mortimer et al. 2018).

These volcanoes form a geographically continuous chain of more than 70 large basaltic-andesitic calc-alkaline edifices (Bergman et al. 1992), hundreds of parasitic and satellite vents, as well as numerous intrusive bodies (Figures 1 and 7). The age of the volcanoes range from Early Miocene (ca 25 Ma) in the north, to the active Mount Taranaki in the south, forming a roughly north–south trending belt oblique to the modern trend of the Taupo Rift and to the strike of the subducting Pacific plate. Most volcanoes are deeply buried (>1000 m) offshore of the west coast of the North Island and correlate in age and lithologies with volcanic rocks outcropping in the Northland and Coromandel Peninsulas, and with modern volcanoes of the Taranaki Peninsula and Taupo Volcanic Zone (e.g. Herzer 1995; Hayward et al. 2001; Seebeck et al. 2014).

Figure 7. Simplified map showing the distribution of Miocene-Quaternary subduction-related volcanoes onshore and offshore the North Island of New Zealand. Coloured polygons show the location and age of volcanoes mapped in seismic data, while dots correspond to dredged samples of volcanic origin or the location active volcanoes shown in the Global Volcanism Programme webmap. White dashed lines show the trends of subduction-related volcanoes by age groups. Note a progressive younging in the age of the volcanoes from NW to SE. Aotea volcanic complex has a representative dredge sample comprising a basalt with an Ar/Ar date of 22.5 Ma (Mortimer et al. 2018), indicative of intraplate origin. See discussion and Supplementary Material 5 for more details. Mapping of the offshore volcanoes was combined with information from Herzer (1995), Hayward et al. (2001), Mortimer et al. (2010), Giba et al. (2013), Seebeck et al. (2014) and GNS Geological Map (Heron 2014). Background image shows a bathymetric map from NIWA (New Zealand Institute of Water and Atmosphere).

The morphologies of volcanic edifices include large stratocones (e.g. Kora), shields (e.g. Waipoua), and volcanic complexes with several overlapping edifices (e.g. within the Parihaka and Nimitz 3D seismic volumes). Many of these volcanoes were long-lived systems active for >8 Myr (e.g. Kora, Manukau, Te Kumi; Giba et al. 2013). Typically, in the north, the volcanoes are wider and show low-angle flanks (e.g. Manukau, 75 km wide and ca 1600 m high), while smaller and steeper volcanoes predominantly occur in the south (e.g. Kora, 15 km wide and ca 1200 m high), which could reflect changes in the geometry of the subducting slab (Seebeck et al. 2014). Most volcanoes have erupted in a submarine setting, comprising a complex variety of lithofacies such as tuffs to tuff-breccias, and pillow-lavas commonly interbedded with volcanogenic sandstones and conglomerates (Bergman et al. 1992; Bear and Cas 2013; Shumaker et al. 2018; Figure 8(c,d)).

Figure 8. Oblique 3D seismic visualisation and volcanogenic rocks of the Kora Volcano, a long-lived Miocene stratocone of the Mohakatino belt. (a and b) Main chronostratigraphic surfaces rock units and structures coupled with geobody mapping of Kora plumbing system. Note that intrusions were preferentially emplaced into pre-existing Cretaceous-Paleocene grabens and along normal faults. (c) Photograph of a core sample from the drillhole Kora-1a at the depth of 1783 m showing a moderately sorted lapilli-tuff. (d) A polymictic volcanic conglomerate from 1908 m in Kora-1a. The location of the Kora Volcano is shown in Figure 7. Modified from Bischoff et al. (2017) and Bischoff and Nicol (2019).

The plumbing systems of the Northland-Mohakatino belts emplaced hundreds of km³ of magma into pre-existing Cretaceous grabens and associated normal fault zones, which forms a large batholithic mass of igneous rocks N-S trending and extending semi-continuously for >500 km (Figure 8). Observations in the Kora and Nimitz 3D datasets indicate that most of these intrusions occur in association with large polygenetic volcanoes. The intrusions show seismically detectable sheet-like shapes such as dikes and sills, and relatively small (<7 km across) swarms of hybrid bodies with a variety of geometries such as saucer- and stock-like morphologies that cross-cut each other and the sedimentary host rocks (e.g. Infante-Paez and Marfurt 2017; Morley 2018; Kumar et al. 2018). In detail, these bodies more commonly occur as interconnected networks of magmatic intrusions rather than forming large plutonic rock masses. Dikes and sills of polygenetic volcanoes usually form along, or branching from, pre-existing Cretaceous faults, commonly converging towards a relatively stationary central vent, while small parasitic/satellite volcanoes are typically, but not exclusively, located above the tips of peripheral shallow (<2 km) saucer-shaped intrusions (Figure 8(a,b); Bischoff et al. 2017).

West Ngatutura volcanic field – Pliocene

The West Ngatutura volcanic field is a Pliocene cluster of at least 40 scattered small-volume cone-type volcanoes varying in size from <1 km across and <400 m high, distributed over an area of ca 2100 km². Presently, these volcanoes are buried by 400–800 m of sedimentary strata offshore of the Waikato and Auckland coasts (Figures 1 and 9). We refer to this cluster of volcanoes as the western equivalent of the onshore Ngatutura volcanic field (Table 1).

Figure 9. (a) Map showing the location of the Pliocene West Ngatutura volcanic field, regional fault structures and possibly correlative onshore volcanic fields. The coloured bar shows the isochron map between the pre- and post-eruptive surfaces, revealing the location and morphology of the volcanoes. Onshore volcanic rocks are from GNS Geological Map (Heron 2014) and faults are from Giba et al. (2013). (b) Interpreted 2D seismic section (line OA07-025) showing the morphology of some volcanoes and intrusions in the West Ngatutura volcanic field. Numbers are: (1) feeder dikes; (2) saucer-shaped intrusion; (3) jacked-up dome; (4) seismic artefact (multiple); (5) seismic artefact (hyperbola); (6) seismic artefact (pull-up velocity); (7) small-volume cone-type volcano; (8) lava-flows; (9) vents.

The volcanoes erupted within the lower part of the Pliocene sequence of the Taranaki Basin (ca 3.5 Ma) in paleo-water-depths around 2000 m. There is no evidence of volcanic rocks of this age from the nearest drillholes. However, mound- and cone seismic anomalies are underlain by pull-up of seismic velocities artefacts, indicating that rocks within the anomalies have a much higher acoustic impedance than surrounding sedimentary strata, supporting the interpretation of an igneous origin (Uruski 2019; Figure 9(b)). This interpretation is reinforced by numerous high-amplitude reflections cross-cutting host strata located below the pre-eruptive surface of these volcanoes, as is typical of igneous intrusions. In addition, doming of reflections overlying cone-type volcanoes is often observed, where volcanic rocks have compacted less than surrounding sedimentary strata (e.g. Planke et al. 1999; Holford et al. 2012; Bischoff et al. 2017).

The random distribution of the volcanoes is typical of the intraplate Horomaka Supersuite (Mortimer et al. 2014). However, a NE alignment of vents is observed in the east part of the field, which is located above a Cretaceous fault of same direction (Figure 9(b)), and parallel with the concurrent opening of the Taranaki Northern Graben (King and Thrasher 1996), situated 20 km east of the field. Such alignment of vents is common for the Neogene intraplate volcanoes erupted in the North Island (Le Corvec et al. 2013).

Other volcanoes buried in the northern Zealandia

Other Late Cretaceous (and possible older) volcanoes identified offshore the North Island are Tikati volcanic complex (Uruski 2019; Supplementary Material 3), and intrusions and small vents in the area of the Matuku-3D dataset (Supplementary Material 4). The north-western limit of the Deepwater Taranaki Basin contains a large (ca 100 km) W–E elongated seamount oblique to the NW orientation of the New Caledonia Trough (Aotea Seamount, Brodie 1965). Its morphology in seismic imagery shows a main volcanic body with multiple edifices and parasitic and satellite vents, consistent with it being a volcanic complex (Supplementary Material 5). The volcanoes of Aotea erupted immediately above the base of the Miocene chronostratigraphic surface indicating an age of ca 23 Ma, which is supported by Ar–Ar dating of ca 22.5 Ma, while a dredge sample returned an ultra-alkaline chemical signature, suggesting an intraplate origin (Mortimer et al. 2018).

Southern Zealandia

Late Cretaceous volcanoes

Numerous Late Cretaceous volcanic centres are mapped in areas of coeval continental rifting in the Canterbury and Great South basins (Field et al. 1989; Bischoff et al. 2016; Barrier 2019; Tables 1–4). Many of these volcanoes occur immediately above the basement of the host basin, which, due to the similarity in the seismic signal of volcanic and basement rocks, makes their characterisation difficult (Barrier 2019). Similar seismic facies are identified SE of Banks Peninsula (e.g. line IP256-98-001) where volcanic rocks of the Mount Somers Volcanic Group were intersected by the Leeston-1 and JD George-1 drillholes proving a Late Cretaceous volcanic origin for these seismic facies in at least some cases. Large Late Cretaceous volcanic complexes within the Canterbury and Great South basins are typically imaged with good resolution in seismic data. Examples of these volcanic complexes are Sloop (Supplementary Material 6), Barque (Figure 10), and Tapuku East (Supplementary Material 7).

Figure 10. (a) Spectral decomposition of the Endurance-3D seismic cube with a depth-slice window at 3000 m. (b) 2D seismic section across the Late Cretaceous volcanoes of the Barque volcanic complex. This volcanic complex comprises two large edifices (ca 17 km across and ca 1000 m high each), as well as several parasitic and satellite vents and numerous intrusive bodies (Beggs 2016). Note that most eruptive centers are located at, or aligned with, NE-SW trending normal faults, which are parallel to the trend of extension of the Zealandia rift in the area (Barrier 2019). Figures 1 and 10 show the location of the volcanic complex.

Table 3. Main characteristics of the volcanoes buried in the Taranaki, Deepwater Taranaki and Reinga-Northland basins.

Volcano	Age	Tectonic setting	Type	Environ. eruption	Dominant comp.
Chronos composite volcano	lC ∼85 Ma	Zealandia rift	Composite cone	Paralic	Unknown
Rhea composite volcano	lC ∼84 Ma
Coopworth volcanoes	lC ∼74 Ma	West Coast-Taranaki rift	Cluster of small-volume cone-type volcanoes		Basaltic *
Romney volcanic field	lC ∼74–73 Ma		Cluster of small-volume cone-type volcanoes fissures and crater-rows, and a large shield volcano		Basaltic to intermediate. Rhyolitic? #
Hades caldera	lC ∼73 Ma		Caldera		Unknown
Vulcan composite volcano	lC ∼73–70 Ma		Composite cone
Hestia composite volcano	lC ∼72–68 Ma
Tikati volcanic complex	lC ∼70 Ma		Cluster of small-volume cone-type volcanoes and one large multi-vent complex	Shallow-submarine
Matuku intrusions and small vents	lC ∼68 Ma		Cluster of intrusions and relative small-volume vents	Paralic	Basaltic #
Kaiwero volcanic field	E ∼55 Ma	Intraplate	Cluster of small-volume cone-type volcanoes	Shallow to deep-submarine	Basaltic *
Northland-Mohakatino volcanic belts	eM-Q ∼23–0 Ma	Intra-arc	Belt of large composite and shield volcanoes	Deep-submarine to subaerial	Basaltic-Andesitic #
Aotea volcanic complex	eM ∼23 Ma	Intraplate	Composite multi-vent massif	Deep-submarine	Basaltic #
West Ngatutura volcanic field	Pli ∼3.5 Ma	Intraplate	Cluster of small-volume cone-type volcanoes		Basaltic *

Notes: Symbol (*) refers to composition estimated from seismic geomorphology analysis and (#) from integration of seismic and drillhole data. Abbreviations are: Late Cretaceous (lC), Eocene (E), Early Miocene-Quaternary (eM-Q), Middle Miocene (mM), Pliocene (P).

Table 4. Main characteristics of the volcanoes buried in the buried in the Canterbury and Great South basins.

Volcano	Age	Tectonic setting	Type	Environ. eruption	Dominant composition
Sloop volcanic complex	lC ∼83–70 Ma	Rift and Post-rift	Composite multi-vent massif	Paralic	Unknown
Barque volcanic complex	lC ∼79–65 Ma	Post-rift	SW edifice: Strato-shield, central caldera and several overlapping small vents. NE edifice: stratocone		Probably Basaltic *
Tapuku East volcanic complex	lC ∼83–70 Ma	Rift and Post-rift	Composite multi-vent massif		Unknown
East Waiareka-Deborah volcanic field	P-O	Intraplate	Cluster of small-volume cone- and crater-type volcanoes	Shallow to deep-submarine	Basaltic alkaline #
Papatowai volcanic field	O or eM		Cluster of small-volume cone-type volcanoes and pillow mounds?		Unknown
Maahunui volcanic field	mM 12.7–11.5 Ma		Cluster of small-volume cone- and crater-type volcanoes	Deep-submarine	Basaltic alkaline #
Waka intrusions and small vents	M ∼18 Ma		Cluster of intrusions and relative small-volume vents		Probably Basaltic *
Toroa volcanic field	Ple ?		Cluster of small-volume cone- type volcanoes	Shallow to deep-submarine	Unknown

Notes: Symbol (*) refers to composition estimated from seismic geomorphology analysis and (#) from integration of seismic and drillhole data. Abbreviations are: Late Cretaceous (lC), Paleocene-Oligocene (P-O), Oligocene (O), Early Miocene (eM), Middle Miocene (mM), Miocene (M), Pleistocene (Ple).

East Waiareka-Deborah volcanic field – Paleocene to Oligocene

A cluster of >50 randomly distributed small-volume edifices is buried by up to 3 km within Paleocene-Oligocene strata of the Canterbury Basin, covering an area of ca 3,500 km² (Figure 11). These edifices have been interpreted as an extension of the volcanoes of the Waiareka-Deborah field, onshore in the region of Oamaru (e.g. Field et al. 1989; Barrier et al. 2017; Fensom 2020). There is presently no evidence of Paleocene volcanism in inland eastern Otago, thus, we refer to the offshore volcanoes as the East Waiareka-Deborah volcanic field to differentiate them from the onshore counterpart (Table 1).

Figure 11. (a) Map showing the location and age of volcanoes onshore and offshore eastern South Island (Horomaka Supersuite). Onshore volcanic rocks are from GNS Geological Map (Heron 2014). (b,c) 2D sections showing the age and morphology of small-volume cone-type volcanoes buried in the Canterbury Basin. Numbers are: (1) intrusions; (2) conduits. (d) An outcrop within the Kakanui Head, a volcano of the Waiareka-Deborah volcanic field showing an angular contact between amalgamated and poorly-sorted pyroclastic beds inward-dipping towards the central crater of the edifice, while well-sorted and tabular beds dip outward.

Some volcanoes in this field were intersected by the Endeavour-1, Cutter-1, Galleon-1, Clipper-1, Oamaru-1 and Oamaru-2 drillholes (Field et al. 1989), confirming an igneous origin for the seismic anomalies. Volcanic rocks recovered from these drillholes comprise basaltic tuffs up to 124 m thick interbedded with marine mudstones, sandstones and limestones, indicating that eruptions occurred in a shallow (<500 m) marine environment (Schiøler et al. 2011), which is in accordance with interpretations of onshore volcanics (Coombs et al. 1986; Corcoran and Moore 2008). Samples from the Endeavour-1 drillhole contain glassy basaltic tuffs eventually altered to palagonite (Wilding and Sweetman 1971), a characteristic of the rocks exposed onshore (Cas et al. 1989; Moorhouse et al. 2015; Fensom 2020).

The volcanoes are typically <1 km wide and <400 m high. The flanks of well-preserved cones usually show bedded seismic facies that dip outward from a central zone of chaotic and/or inward-dipping reflections of moderate amplitude (Figure 11(b,c)), suggesting accumulation of pyroclastic material (Bischoff, Nicol, Barrier, et al. 2019). This interpretation is reinforced by the tuffaceous facies described in the drillholes (Field et al. 1989), and by the comparable morphology of volcanoes in seismic imagery and onshore exposures (Figure 11(d); Fensom 2020). Some minor high-amplitude reflections located next to the vents may indicate the presence of pillow-lava deposits.

Disrupted seismic facies commonly occur below the vents and likely indicate the presence of magma conduits (Figure 11(b,c)). Large (>3 km) intrusive bodies are rarely observed below this volcanic field. This contrasts with other monogenetic volcanic fields presented in this study, suggesting that the East Waiareka-Deborah volcanoes are mainly fed directly from a magma source at least deeper than the seismic resolution of our dataset (i.e. >5 km), which is in agreement with petrological models for the volcanoes onshore (Coombs et al. 1986). Based on the random distribution of the volcanoes, distance from plate boundaries and correlation with the onshore counterparts, the East Waiareka-Deborah volcanic field likely has an intraplate origin (Horomaka Supersuite).

Papatowai volcanic field – Oligocene or Early Miocene

The Papatowai volcanic field comprises at least 30 scattered small-volume cone-type volcanoes located offshore of the Clutha district, South Island, New Zealand (Figure 12). We refer to this cluster of volcanoes using the name of the town and river adjacent to the area where they are located (Table 1). There is no physical evidence of volcanic rocks in this area, however, the presence of pull-up velocity artefacts underlying the volcanoes, as well as high-amplitude reflections with saucer-shaped geometry cross-cutting host pre-eruptive strata indicate a volcanic origin (Figure 12). In addition, the interpretation of aeromagnetic data suggests the presence of Cenozoic volcanic rocks in the region (Waight et al. 2019).

Figure 12. (a) TWT map on the proxy post-eruptive surface of small-volume cone-type volcanoes of the Papatowai volcanic field. (b and c) 2D sections (lines OMV08-042 and Husky-H-104) showing the morphology of some of the volcanoes and part of their plumbing systems. Numbers are: (1) feeder dikes, (2) intrusion; (3) jacked-up dome; (4) seismic artefact (pull-up velocity); (5) pillow mounds? (6) small-volume cone-type volcano; (7) pre-eruptive surface; (8) post-eruptive surface.

Volcanism in this field occurred over an area of ca 1600 km². The volcanoes erupted in paleo-water-depths <500 m. It is not possible to precisely determine the age of these volcanoes based exclusively on seismic lines because they are buried within a thin condensed section of sedimentary rocks of Oligocene to Early Miocene age, with epoch boundaries below the seismic resolution of the dataset at that location.

As is the case in the East Waiareka-Deborah and Kaiwero volcanic fields, the flanks of the Papatowai cones usually show bedded seismic facies outward dipping from a central zone of chaotic and/or inward-dipping reflections of mod-amplitude (Figure 12(b,c)), suggesting pyroclastic eruptions. In addition, most cones have peripheral high-amplitude reflections, which could correspond to pillow-lavas or other forms of dense submarine lava-flow deposits. The Papatowai volcanoes likely have an intraplate origin and belong to the Horomaka Supersuite.

Other volcanoes buried in the southern Zealandia

Offshore southeast of Banks Peninsula, the Maahunui volcanic field comprises a cluster of Middle Miocene volcanoes and shallow (<1 km) intrusions currently buried by ca 1000 m of sedimentary strata in the offshore Canterbury Basin (Supplementary Material 8; Bischoff, Nicol, Cole, et al. 2019, Bischoff, Nicol, Barrier, et al. 2019, Bischoff, Rossetti, et al. 2019). The Waka-3D seismic survey located offshore Dunedin contains several large (ca 5 km across) saucer-shaped intrusions, mainly aligned with pre-existing Cretaceous faults (Blanke 2012; Figure 13). Above these intrusions, a number of volcanic vents are located onto the Miocene paleo-sea floor (Tables 2 and 4), having been erupted at a water depth of >500 m. The lateral extent of this volcanic field is unclear due to the limited resolution of the seismic data outside the area of the 3D survey, but they correlate in age and morphology with monogenetic vents erupted in the vicinity of the Dunedin volcano onshore (Figure 13; Scott et al. 2020). Another field of scattered small-volume volcanoes and shallow intrusions probably of Pleistocene age occurs ca 100 km SE of Stewart Island, near the location of the Toroa-1 drillhole (Supplementary Material 9; Tables 2 and 4). Maahunui, Waka and Toroa volcanic fields are all likely intraplate in origin (Horomaka Supersuite).

Figure 13. Examples of intrusions and vents of the Miocene Waka volcanics. (a) Amplitude opacity rendered seismic volume (Waka-3D) revealing a number of intrusions emplaced into Paleocene sedimentary strata and vents erupted onto the Miocene paleo-sea-bed. Note the spatial relationship between the intrusive bodies and vents. (b) Geobody extraction of a saucer-shaped intrusion showing its geometry, relationship with enclosing strata, conduits, and co-genetic eruptive vents. Modified from Bischoff and Nicol (2019).

Discussion

Insights from volcanoes buried in sedimentary basins can help to fill important gaps in understanding patterns of volcanism across Zealandia. However, vast areas are still poorly imaged (e.g. West Coast, Aotea, Bellona, Emerald, Northeast Slope basins, and towards east of the Canterbury and Great South basins), or only covered by sparse 2D seismic grids that cannot give insights in the same resolution as 3D datasets (e.g. Reinga-Northland Basin).

Regions of high erosion or those affected by tectonic processes associated with the Hikurangi and Puysegur subduction zones, and with uplift of the Southern Alps (e.g. Ballance 1993; Lebrun et al. 2003; Stagpoole and Nicol et al. 2007) generally have poor scope for burial and preservation of volcanoes. The following sections discuss the first-order regional features in areas where preservation was enhanced by low deformation, little or no erosion, and the presence of seismic reflection data is sufficient to image volcanoes and volcanic fields (Figures 1 and 14).

Figure 14. Simplified geological evolution of the New Zealand region showing four main tectonic-magmatic phases. (a) The first phase (ca 100–85 Ma) is characterised by widespread intracontinental break-up (Zealandia rifts) and simultaneous volcanism broadly associated with lithosphere thinning and rifting. (b) The second phase (ca 85–66 Ma) corresponds to the separation of Zealandia from eastern Australia and Antarctica, and volcanism in western New Zealand coeval with the West Coast-Taranaki rift, and volcanic activity immediately after cessation of rifting in southern Zealandia. (c) During the third phase (ca 66–30 Ma), diffuse intraplate volcanism mainly formed fields of small-volume volcanoes. (d) The fourth phase (ca 30–0 Ma) is characterised by both subduction-related volcanism and semi-continuous diffuse intraplate magmatic activity. Maps are compiled and updated from Herzer (1995), Hayward et al. (2001), Mortimer et al. (2010), Giba et al. (2013), Seebeck et al. (2014), Bache, Mortimer, et al. (2014), Mortimer et al. (2018), Bischoff, Nicol, Barrier, et al. (2019), Waight et al. (2019), Uruski (2019), Barrier (2019), Global Volcanism Programme, and GNS Geological Map (Heron 2014).

Late Cretaceous syn- and immediate post-rift volcanism

Late Cretaceous volcanic activity typically formed large (20–1400 km³) composite cones and volcanic complexes that were active for ca 1–15 Myr, and less commonly fields of scattered small-volume (<1 km³) volcanoes erupted into subaerial to shallow marine environments (Figure 2). These volcanoes primarily erupted under the influence of a tensional crustal stress field (i.e. σ₁ vertical, σ₂ and σ₃ horizontal) associated with the break-up and separation of Zealandia from eastern Gondwana (Figures 2 and 14(a,b)). For example, in the Taranaki and Deepwater Taranaki basins, two eruptive phases of volcanism occurred in Late Cretaceous and were at least in part concurrent with two rifting events (i.e. Zealandia and West Coast-Taranaki rifts; Strogen et al. 2017). During the first eruptive phase, ca 85–84 Ma and possibly older, Chronos, Rhea and at least two additional unnamed volcanoes (Uruski 2019) erupted along the NW-SE rift system of the New Caledonia Trough (Figures 3 and 14(a)). The precise timing of cessation of this eruptive phase is difficult to determine based on the available data, but it was contemporaneous with, or followed by, deposition of the lowermost Taranaki Delta around 83 Ma (Schiøler et al. 2014). In northern Zealandia, the first rift phase shut down at ca 83 Ma, perhaps due to the onset of seafloor spreading in the Tasman Sea and the associated cessation of normal faulting in the interior of Zealandia (Gaina et al. 1998; Strogen et al. 2017; Barrier, Nicol, et al. 2020). Relaxation or reduction of the tensional stresses across Zealandia with the advent of sea floor spreading may have caused the cessation of this first eruptive phase.

The second eruptive phase (ca 74–68 Ma) formed a broad N-S alignment of volcanoes along the extension of the West Coast-Taranaki rift (Figures 2 and 14(b)). These include Coopworth, Romney, Hades, Vulcan, Hestia (Figures 3 and 4), Tikati (Supplementary Material 3), and volcanoes within the Matuku-3D dataset (Supplementary Material 4). We observe that many, if not most, volcanoes of the second eruptive phase are typically located along concurrent N-S trending normal faults (Figure 3). This is also documented in Zhu et al. (2017) and Uruski (2019) and suggests that the second eruptive phase is linked with extension concurrent with West Coast-Taranaki rifting.

Uruski (2019) argues that crustal extension does not control the location of volcanoes in Deepwater Taranaki Basin, since many volcanoes erupted above local basement highs with magma migrating through thicker crust. However, the locus of eruptions along continental rifts, such as those of Late Cretaceous age in Zealandia, is thought to be controlled by the interplay between the rate of extension of the crust and the presence of pre-existing structures (Corti 2009; Vries and Vries 2018). Rifts associated with high rates of extension tend to form volcanoes at the centres of the rift axis, where intense crustal thinning occurs (e.g. Afar Rift; Illsley-Kemp et al. 2017), while rifts with low extension rates typically have volcanoes located along the rift margins and on local basement highs (e.g. Limagne Rift; Vries and Vries 2018). Volcanoes erupted at both rift axis and margins are buried in offshore New Zealand. We infer that the locations of these volcanoes probably reflect the interplay of tectonic stress during volcanic activity and the presence of pre-existing crustal structures (Figure 5(a)).

A relationship between extension and the location of eruptive centres is also observed in southern Zealandia. In the Canterbury Basin, 70% of the syn-rift volcanoes are within 5 km of a rift fault (Barrier 2019). This author also suggest that elongated features formed along rift faults may correspond to volcanoes erupted in fissures, typical of extension-related volcanism elsewhere (e.g. Corti 2009; Sigmundsson et al. 2018; Neal et al. 2019). Large (>500 km³) polygenetic volcanoes erupted post Zealandia break-up (e.g. Barque, Sloop and Tapuku) are also aligned with both NE-SW and NW-SE trending rift faults (Figure 14(b); Barrier 2019; Figure 10). Similarly, volcanoes of Late Cretaceous age in the Great South Basin follow the NE-SW trend of rifting (Figure 14(a,b)). In addition, areas of high-positive magnetic anomalies across Zealandia are interpreted to be caused by Late Cretaceous mafic rift-related igneous rocks (Waight et al. 2019; Mortimer and Scott 2020).

Collectively, evidence compiled from previous work and new data presented in this paper indicates that during the separation of Zealandia from south-eastern Gondwana, Late Cretaceous magmatism and continental rifting were closely linked in time and space (Figure 14(a,b)). In a tectonic sense, we consider that these rift-related volcanoes should be distinguished from the intraplate Horomaka Supersuite (Mortimer et al. 2014). We propose here a new stratigraphic unit, Te Kiekie Supersuite, to describe Late Cretaceous volcanoes that form within coeval intracontinental rifting of Zealandia. Te Kiekie is the Māori name for the Mount Somers, inland Canterbury Basin, which represents volcanic activity contemporaneous with Zealandia rifting (e.g. Laird and Bradshaw 2004; Mortimer et al. 2014; Van der Meer et al. 2018). This classification could open up scope for future research for identifying petrological changes and differences between Late Cretaceous and Cenozoic intraplate volcanoes of the Horomaka Supersuite.

Cenozoic diffuse intraplate volcanism

Offshore buried volcanoes erupted throughout the Cenozoic and remote from the Australian-Pacific plate boundary (>200 km) predominately form clusters of scattered small-volume (<1 km³) vents (Figure 2), as characteristic of monogenetic volcanic fields (Németh and Kereszturi 2015). The random distribution and absence of obvious structural control on the location of these buried volcanoes suggest that they have an intraplate origin, which is supported by geochemistry studies on their onshore counterpart volcanic fields, such as the Auckland and Waiareka-Deborah (Scott et al. 2020; Hopkins et al. 2020).

In detail, alignment of vents and intrusions does occur, but at a much more local scale than volcanoes of Late Cretaceous age (Figures 9(a) and 13). These alignments are observed in three Neogene volcanic fields (West Ngatutura, Waka and Maahunui) that erupted a large distance (<300 km) from the primary plate boundary structure (Figures 1, 2 and 14(c,d)). In all three cases, the vents and intrusions are aligned parallel to the trend of pre-existing faults, which is commonly also observed for Neogene monogenetic volcanic fields in the North Island (Le Corvec et al. 2013). In contrast, during a period of tectonic quiescence across most New Zealand in the Paleogene, neither onshore nor offshore monogenetic fields show any clear relationship with pre-existing crustal structures.

After ca 40 Myr of predominantly monogenetic small-volume intraplate volcanic activity in Zealandia, the large polygenetic volcanoes of Banks Peninsula, Otago Peninsula, and Aotea Seamount erupted concurrently with Neogene tectonic activity in the Hikurangi and Puysegur subduction zones and the transpressional Alpine Fault. Banks Peninsula erupted at the intersection of WNW-ESE and NE-SW trending basement structures of Cretaceous age (i.e. Chatham Rise and Endeavour High; Field et al. 1989; Barrier et al. 2020). Eruptive centres formed along normal faults of similar orientation, presumably due to reactivation concurrently with volcanic activity (Hampton and Cole 2009). Otago Peninsula erupted near the limit of the Rakaia and Caples terranes, and is also aligned with the Titri fault and Taieri graben (Coombs et al. 1960; Deckert et al. 2002). Aotea volcanic complex erupted onto the West Norfolk Ridge, with normal faults along the boundary with the New Caledonia Trough (Supplementary Material 5).

The influence of tectonics on fault displacement is reported to extend far from plate boundaries. For example, in the intra-cratonic Paraná Basin, southern Brazil, extension and reactivation of normal faults along the boundaries of crustal blocks are linked with the timing of terrain accretion on the western Gondwana plate boundary, which was located at least 1500 km from the basin (Milani and Ramos 1998; Holz et al. 2006). There is a general agreement that magma erupted remote from plate boundaries have to rise through thicker (>10 km) crust, facilitated mostly by the existence of an extensional stress field in the lithosphere, and by the interaction of magma with pre-existing crustal structures (e.g. Shaw 1980; Le Corvec et al. 2013; Vries and Vries 2018).

The spatial associations between major structures and Neogene volcanoes across Zealandia may suggest that the stress field imposed at plate boundaries could be at least in part responsible for fault reactivation, extension and eruptions in an intraplate setting some distance from the primary plate boundary structures.

Cenozoic subduction-related volcanism

Zealandia arc-related volcanoes mainly comprise large deep-submarine to subaerial composite, shield and caldera volcanoes (Ballance et al. 1985; Cole 1986; King and Thrasher 1996; Figure 2). The relative locations of the arcs are complex and have been interpreted to reflect progressive rollback and/or steepening of the slab and southward migration of crustal extension of the upper plate during the Miocene to Recent (Kamp 1984; Mortimer et al. 2010; Giba et al. 2010; Figure 14(d)). Volcanoes associated with subduction along the Hikurangi margin define a broad NE–SW trending zone that was oriented approximately parallel to the strike of the subducting Pacific plate since at least from the last 16 Ma (Seebeck et al. 2014). Until about 12 Ma, significant volcanism did not extend south of about 38^o S latitude, migrating ca 60 km further south during 12–8 Ma, with voluminous eruptions then in the area of the Parihaka-3D survey up to about 4 Ma (Hayward et al. 2001; Giba et al. 2013; Figure 7). After 4 Ma, volcanism migrated further SE to its present location at the Taupo-Kermadec-Tonga trend (Cole and Lewis 1981; Figure 7).

The active volcanoes of the Central North Island are now situated about 80–100 km above the Pacific plate subducting slab, with magma interpreted to be produced by slab dehydration and partial melting of the asthenosphere (Seebeck et al. 2014; Illsley-Kemp et al. 2019; Figure 7). Southwest of the Central North Island and south of the Mt Taranaki, seismicity indicates that the subducting slab reaches depths of >100 km, however, there is no evidence of volcanic activity in seismic reflection surveys and drillholes on this area (Figure 1). A similar southward termination also occurred for the Miocene Mohakatino volcanoes, although this termination was northwest of the contemporary termination of the Central North Island volcanoes (Figure 14(d)).

During both time intervals volcanism was absent from regions primarily accommodating crustal contraction, where near-horizontal σ1 stress may generally increase fault-normal stresses and reduce fracture apertures, both of which would act to reduce the utility of crustal weakness in the ascent of magma. In areas where crustal extension rate is sufficiently low that it is not manifest as normal faulting, the rise of magma through the crust may be facilitated by pre-existing Cretaceous faults (Giba et al. 2013). Conversely, volcanism is also absent in some areas where normal faulting and crustal extension has occurred. At the present time, the southwards transition from crustal tension to crustal compression in the southern Taranaki Basin occurs midway between North and South Islands (e.g. Nicol et al. 2007; Rajabi et al. 2016; Massiot et al. 2019), and yet there is no volcanism south of Mt Taranaki. The absence of volcanism in this case may reflect the presence of thickened continental crust due to late Cenozoic thrusting and shortening, and this thickening may inhibit the upward flow of magma to the ground surface. Collectively, this evidences suggest that the interplay of stress fields, pre-existing structures and crustal thickness may play important role on the location of the Cenozoic subduction-related volcanoes of the Norther Zealandia.

Conclusions

Interpretation of a large seismic reflection dataset and petroleum exploration drillhole from offshore New Zealand revealed over 400 Late Cretaceous-Cenozoic volcanoes buried within Te Riu-a-Māui/Zealandia sedimentary basins. In this paper, we propose a systematic classification of the best understood examples of these volcanoes into three large-scale categories (i.e. rift, intraplate, and subduction-related). These categories are sub-divided into informal volcanic zones, fields, complexes, and individual volcanic edifices, according to the location, distribution, age, size and morphology of the volcanoes. Results indicate that the location of eruptive centres was primarily controlled by the stress regime operating at each tectonic setting, with the local influence of pre-existing crustal structures and types of host rocks. Late Cretaceous volcanoes mainly comprise large (>20 km³) long-lived (>1 Myr) composite cones and volcanic complexes that erupted concurrently in time, space and orientation of continental rifts formed during break-up of Zealandia from the south-eastern Gondwana. Cenozoic intraplate volcanism was randomly distributed, and characterised by clusters of scattered small-volume (<1 km³) cones and craters associated with numerous shallow (<2 km) magmatic bodies, typically of saucer-shaped geometry. In detail, only Cenozoic intraplate volcanoes erupted contemporaneously with tectonic activity show a spatial relationship with pre-existing structures, suggesting that fault reactivation and volcanism in an intraplate setting could be influenced by the stress field imposed near (<400 km) plate boundaries. In contrast, subduction-related volcanism has formed a geographically continuous chain of over 70 large strato- and shield volcanoes from Early Miocene in the north to the active Mt Taranaki in the south. These volcanoes are the south-western extremity of multiple volcanic arcs associated with the Kermadec subduction zone, in which southward migration in the locus of eruptions is interpreted to reflect progressive rollback and/or steepening of the subducting Pacific Place beneath the Australian Plate. In detail, the location of arc-related eruptive centers was influenced by the interplay of stress field, thickness of the crust and presence of pre-existing tectonic structures. This paper showcases the utility of seismic reflection for determining the relationship between volcanic and tectonic activity. Studying the volcanoes buried in sedimentary basins can help advance understanding of syn- and post-Gondwana break-up magmatism, with global impacts to comprehend the interplay between volcanism and plate tectonics in Zealandia and further afield.

Acknowledgments

We would like to thank IHS Markit and Schlumberger for providing academic licence to use Kingdom and Petrel software. Thanks for the constructive reviews of Nick Mortimer and an anonymous reviewer, and to Marcos Rossetti for petrographic discussions. We are very grateful for all the support provided by the School of Earth and Environment of the University of Canterbury, particularly to Rob Spiers and John Southward.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The data that support the findings of this study were provided by the New Zealand Petroleum and Minerals (NZPAM) and are openly available in the 2018 Petroleum Exploration Data Pack and at https://data.nzpam.govt.nz.

Additional information

Funding

This study is part of a research programme funded by the 2017 Endeavour Fund – Smart Ideas from the Ministry of Business, Innovation and Employment of New Zealand (MBIE) [grant number UOCX1707].