The compositional diversity and temporal evolution of an active andesitic arc stratovolcano: Tongariro, Taupō Volcanic Zone, New Zealand

Pure, Leo R.; Wilson, Colin J. N.; Charlier, Bruce L. A.; Gamble, John A.; Townsend, Dougal B.; Leonard, Graham S.

doi:10.1007/s00410-023-02004-0

The compositional diversity and temporal evolution of an active andesitic arc stratovolcano: Tongariro, Taupō Volcanic Zone, New Zealand

Original Paper
Open access
Published: 25 April 2023

Volume 178, article number 30, (2023)
Cite this article

Download PDF

You have full access to this open access article

Contributions to Mineralogy and Petrology Aims and scope Submit manuscript

The compositional diversity and temporal evolution of an active andesitic arc stratovolcano: Tongariro, Taupō Volcanic Zone, New Zealand

Download PDF

1971 Accesses
1 Citation
2 Altmetric
Explore all metrics

Abstract

New geochemical data, including Sr–Nd–Pb isotopes for whole-rock and groundmass samples, are reported for edifice-forming eruptives at Tongariro volcano, New Zealand, which span its ~ 350 ka to late Holocene history. Tongariro eruptives are medium-K basaltic-andesites to dacites (53.0–66.2 wt% SiO₂) that evolved via assimilation-fractional crystallisation (AFC) processes partly or mostly in the uppermost 15 km of the crust. When ordered chronologically using a high-resolution ⁴⁰Ar/³⁹Ar-dated eruptive stratigraphy, the compositional data show systematic 10–130 kyr cycles. Mafic replenishment events inferred from MgO values occurred at ~ 230, ~ 151, ~ 88 and ~ 56 ka and in the late Holocene, with high-MgO flank vents erupting at ~ 160, ~ 117, ~ 35 and ~ 17.5 ka. Cycles in Sm/Nd, ⁸⁷Sr/⁸⁶Sr, ¹⁴³Nd/¹⁴⁴Nd and Pb isotopic ratios, which are decoupled from MgO, K₂O and Rb/Sr cycles, indicate periods of prolonged crustal residence of magmas from ~ 230 to ~ 100 ka and ~ 95 to ~ 30 ka. AFC modelling shows that intermediate and silicic melt compositions, with r-values between 0.1 and 1, are needed to reproduce Tongariro compositional arrays. AFC models also indicate that ~ 20% of the average Tongariro magma comprises assimilated (meta)sedimentary basement material. Locally, Tongariro and adjacent Ruapehu volcanoes attain their most crust-like ⁸⁷Sr/⁸⁶Sr and ¹⁴³Nd/¹⁴⁴Nd compositions at ~ 100 and ~ 30 ka, paralleling with zircon model-age crystallisation modes at the rhyolitic Taupō volcano ~ 50 km to the NNE. These coincidences suggest that the timing and tempo of magma assembly processes at all three volcanoes were contemporaneous and may have been tectonically influenced since at least 200 ka.

Deep crustal anatexis, magma mixing, and the generation of epizonal plutons in the Southern Rocky Mountains, Colorado

Article 11 January 2015

Generation of Neogene adakitic-like magmas in the Argentine Puna-Eastern Cordillera transition: the Huachichocana Subvolcanic Complex

Article 28 March 2023

Late Neoproterozoic to Carboniferous genesis of A-type magmas in Avalonia of northern Nova Scotia: repeated partial melting of anhydrous lower crust in contrasting tectonic environments

Article 12 July 2017

Introduction

Stratovolcanoes are iconic features of volcanism globally, mostly associated with arc volcanism at modern subduction zones (Davidson and de Silva 2000). Their eruptive products provide insights into magmatic processes associated with the creation and modification of the continental crust and their eruptions pose substantial hazards on and around the edifice. Stratovolcanoes are dominated by broadly intermediate (andesite/dacite) eruption products, generally inferred to be the end products of basalt magmas being injected into the crust. Basalt itself tends to erupt only in lesser amounts, either as a mixing end-member in hybrid intermediate lavas (e.g. Beier et al. 2017; Streck and Leeman 2018; Conway et al. 2020), or from peripheral vents that may be isolated (e.g. Izbekov et al. 2004) or form part of an adjacent basaltic field (e.g. Hildreth and Lanphere 1994; Bacon et al. 1997; Hildreth 2007). Rhyolite is commonly found in the glassy mesostasis of intermediate composition lavas and as melt inclusions in the phenocryst assemblages of andesites (e.g. Donoghue et al. 1995), but only rarely is erupted as discrete products (e.g. Hildreth and Fierstein 2012; Calvert et al. 2018).

To place any individual stratovolcano into its proper context, there is a need to integrate information from mapping and geochronology with geochemical studies, isotopic studies and any available geophysical information (e.g. Dungan et al. 2001; Conway et al. 2018). In general, studies of individual composite volcanoes over the past three decades have addressed one or more of the following.

(a) The origins of the volumetrically dominant andesite magmas. A widely held view is that andesites in general represent hybrid magmas, composed of mixtures of basaltic and dacitic end-members (Eichelberger et al. 2006; Humphreys et al. 2006). This view arises from the presence of mafic enclaves in many andesites and the ubiquitous compositional diversity of crystals and interstitial melts, with the spectrum of melt compositions noted for the apparent scarcity of andesitic liquids (e.g. Reubi and Blundy 2009; Melekhova et al. 2013 and for New Zealand, Donoghue et al. 1995; Gamble et al. 1999; Price et al. 2005, 2012).

(b) The respective amounts of mantle-derived versus crust-derived material in andesitic magmas. Arc magmas often show direct petrographic evidence in their xenolithic and crystalline cargoes for both mantle and crustal inputs during magma assembly (e.g. Rudnick et al. 1986; Gruender et al. 2010; Tollan et al. 2017). Where the isotopic characteristics of mantle and crustal end-members are known, or can be inferred, mixing and assimilation-fractional crystallisation (AFC) models can be used to estimate their relative proportions (DePaolo 1981; Bohrson and Spera 2001). The amount of molten crust incorporated into arc magmas also has implications for the heat balances involved with magma assembly, such as energy-balanced crustal fusion and magma crystallisation. For example, the operation of multiple petrogenetic processes at a volcano, such as in situ differentiation paired with mixing, may be implied where both AFC and mixing models are needed to explain compositional arrays (e.g. Sisson et al. 2014).

(c) The diversity versus uniformity of erupted compositions over time, and likely implications for the degree of organisation of the magmatic system. At one extreme, substantial cones may be constructed from a restricted range of compositions over thousands to tens of thousands of years (Villarrica: Lara and Clavero 2004; North Sister: Schmidt and Grunder 2009; Klyuchevskoy: Bergal-Kuvikas et al. 2017), implying the presence of an organised and persistent magma chamber or chambers. At the other extreme, small eruptions like at Ruapehu in 1995–1996, may include juvenile ejecta that span the total compositional range known in the lifetime of the volcano (Gamble et al. 1999; Nakagawa et al. 1999, 2002; Kilgour et al. 2013). Comparably large compositional diversity is also seen in the Holocene eruptive record at Tongariro (e.g. in MgO concentrations), which cover much of the diversity seen in the longer-term Pleistocene eruptive record (cf. Hobden et al. 1999, 2002; Pure et al. 2020). However, these compositional trends are rapid over tens or hundreds of years in the Holocene, compared with gradual, systematic trends in the Pleistocene. Unusually large eruptions at stratovolcanoes, such as those from Katmai (Hildreth and Fierstein 2012) and Mazama (Bacon and Druitt 1988), reflect the periodic evacuation of organised systems capable of building magma bodies in the 10–100 km³ range. However, the development of calderas at arc stratovolcanoes is uncommon (e.g. Hildreth 2007 for the Cascades) which implies that the development of large-scale organised magma systems necessary for such large eruptions is rare. The degree of organisation in an arc magma system, and its compositional evolution through time, has also been connected to thermal conditioning of the crust based on thermochemical models (e.g. Annen et al. 2006, 2015), but such models have not been tested with high-resolution time–composition datasets that span extended timescales (> 200 kyr).

(d) Temporal changes in eruption rates and erupted compositions between neighbouring volcanoes, with the implications these changes hold for the development of edifices and the long-term drivers of magmatic systems. In some examples, such changes have been attributed to the effects of ice loading and unloading on the edifice (e.g. Rawson et al. 2016). In other cases, however, temporal changes in compositions and apparent eruptive rates appear to be asynchronous with glacial/interglacial timings (e.g. Calvert et al. 2018; Pure et al. 2020) and instead reflect long-term fluctuations in the supply of mafic magmas into the system. Arising from the latter is the issue of whether the eruptive sequences recorded at a volcano reflect an overall maturation of the magmatic system(s), which may be expressed as unidirectional, acyclic trends in eruptive compositions (e.g. Annen et al. 2006). Such trends may arise when mafic magma intrusion rates into the crust are constant. Alternatively, the occurrence of cyclic variations in eruptive compositions (or eruptive rates) may indicate variable mafic intrusion rates into the crust and/or variable tectonic stress states (e.g. Pure et al. 2020).

This paper considers these topic areas through a case study of Tongariro volcano in the North Island of New Zealand (Fig. 1). Tongariro, along with its neighbour Ruapehu, represent two iconic stratovolcanoes that have been among the most active centres since European arrival in New Zealand (Townsend et al. 2017; Leonard et al. 2021). We present a high-resolution temporal record of magma diversity over Tongariro’s ~ 350 kyr lifespan with the aim of identifying and interpreting patterns in time–composition relationships. Geochemical data are arranged chronologically to examine relationships between major oxide, trace element and Sr–Nd–Pb isotopic data in individual samples with ⁴⁰Ar/³⁹Ar age determinations and also their assigned stratigraphic units (Pure et al. 2020). These investigations focus on whether time–composition trends demonstrate cyclicity, whether composition–volume relationships can provide new insights into petrogenetic processes, and to provide a new dataset for assessing petrogenetic models for arc volcanoes (cf. Hildreth and Moorbath 1988; Annen et al. 2006; Eichelberger et al. 2006).

Geological background

Tongariro volcano (Fig. 1) is situated in the andesite-dominated southern segment of the Taupō Volcanic Zone (TVZ) in the central North Island of New Zealand (Wilson et al. 1995; Leonard et al. 2021). The TVZ is part of an actively rifting arc that has been in its present locus for ~ 2 Myr, driven by westward subduction of the Pacific Plate beneath the Indo-Australian Plate. At the latitude of Tongariro, the plate convergence rate is ~ 45 mm/yr which decreases southwards to ~ 33 mm/yr in the South Island where the tectonic regime becomes transpressional (Nicol et al. 2007). On mainland North Island, the Taupō Rift is the expression of arc-related tectonic extension associated with the subduction zone, whereas the closely spatially overlapping TVZ denotes the area where related volcanism has occurred (Wilson et al. 1995). To the south of Ruapehu volcano, the rift terminates in a complex zone of variably oriented faulting (e.g. Villamor and Berryman 2006). Offshore to the north of New Zealand, the TVZ/rift arc continues into oceanic crust where it becomes the Kermadec arc and Havre Trough (Fig. 1).

Tongariro is a composite stratovolcano comprising many discrete vents (e.g. Pure et al. 2020). Previously, it has been defined as part of the Tongariro Volcanic Centre that included all Pleistocene–Holocene volcanoes south of Taupō volcano in the TVZ (i.e. including Ruapehu, Pihanga, Maungakau, Maungakatote, Pukeonake, Hauhungatahi, Ohakune Craters and Kakaramea: Cole 1978), which are labelled in Fig. 1. Some more recent studies have used the term ‘Tongariro Volcanic Centre’ to refer solely to the collection of vents on the Tongariro edifice and have treated each vent as a separate volcano (e.g. Shane et al. 2017). In contrast, we use the term ‘Tongariro volcano’ to describe all vents on the Tongariro edifice and the peripheral vent, Pukeonake, following the formal stratigraphic nomenclature proposed by Pure et al. (2020).

The total edifice volume of Tongariro is approximately 90 km³ relative to a pre-volcanic datum at 750 m a.s.l. (the highest point on the edifice is 2291 m a.s.l.: Pure et al. 2020). Approximately 21% (19 km³) of the edifice is represented by exposed eruptive material comprising predominantly lavas, with smaller volumes of pyroclastic flow deposits, agglutinates, breccias, hyaloclastites and intrusions including at least one prominent dike (Townsend et al. 2017; Cole et al. 2018; Pure et al. 2020). Pleistocene eruptive sequences at Tongariro have volumes ranging between 0.01 and 2 km³, with most formally defined stratigraphic members comprising 0.1–1 km³ of exposed and buried material that can be inferred with confidence (Pure et al. 2020). Eruptive sequences contain non-depositional unconformities: some valley floor areas also contain minor incisional disconformities (i.e. paraconformities) but these are not well characterised because of burial from successive units. Individual Holocene eruptive sequences are ≤ 0.53 km³ with many discrete units amounting to only 0.0003–0.0065 km³ (Hobden et al. 2002; Pure et al. 2020). A key difference between Pleistocene and Holocene eruptive records, in edifice materials, is the scarcity of pyroclastic flow deposits in the Pleistocene (Pure et al. 2020). However, there is a pumice-rich ignimbrite within the 214–207 ka Upper Tama Member, interposed between more voluminous lava sequences, and ~ 17.5 ka scoriaceous eruption deposits atop the Pukekaikiore landform. These pyroclastic deposits demonstrate that ‘Holocene’-style, small-volume eruptions occurred in the Pleistocene, even if they are poorly represented in the edifice record (Townsend et al. 2017; Pure et al. 2020).

In its tectonic setting, Tongariro sits in an active graben between uplifted eastern (Kaimanawa Mountains) and western shoulders, underlain by Mesozoic greywacke and which are locally covered by up to 200 m of Neogene sediments (Townsend et al. 2017). The Tongariro section of the TVZ arc, also known as the Tongariro Graben, is rifting at ~ 7 mm/yr with several recently active normal faults crossing the volcano (Gómez-Vasconcelos et al. 2017; Villamor et al. 2017; Pure 2020). In contrast, there is less evidence of faulting activity on Ruapehu and rates of extension diminish to the south, towards the rift termination (Fig. 1; Villamor and Berryman 2006; Villamor et al. 2007). This rifting is inferred to have lowered the central edifice areas, particularly of Tongariro, relative to the flanks (Gómez-Vasconcelos et al. 2017). Despite this subsidence, Tongariro and Ruapehu are the only stratovolcanoes in New Zealand that have repeatedly supported substantial glaciers (e.g. McArthur and Shepherd 1990) and deposits from three periods of major glacial advance are preserved in their 200–350 kyr histories (Conway et al. 2015; Eaves et al. 2016; Townsend et al. 2017).

The composite edifice of Tongariro is constructed from volcanic products erupted from multiple vents distributed over a ~ 12 × 5 km area (Fig. 2). Younger eruption centres on the edifice, such as those forming the cones of Ngāuruhoe and Red Crater (Hobden et al. 2002; Shane et al. 2017), are obvious from their prominence in the landscape and the presence of lava flows that can be traced directly back to their vent source. The oldest products that are directly (physically) linked with their vent sources are those from North Crater (unit ‘ari’ in Fig. 2) that erupted between 45 and 36 ka (Pure et al. 2020). Older (> 50 ka) vent locations are less clear because of burial by younger eruptives and/or sediments, or can only be inferred from material preserved elsewhere such as on the ring plain.

Although the provenance of totally buried edifice material is largely unknown, the visible edifice materials at Tongariro are more likely to form a relatively complete inventory of erupted compositions than at some other stratovolcanoes, at least since ~ 230 ka (Pure et al. 2020). This is because eruptions at Tongariro were distributed across multiple vent areas which minimised the amount of stockpiling and edifice growth within confined areas, and therefore large eruptions at Tongariro were less likely to entirely bury smaller, less dispersed eruptive units. The sampled edifice record (Fig. 2) also captures a relatively continuous eruptive history, with few gaps between the analytical errors of ⁴⁰Ar/³⁹Ar and K–Ar age determinations since ~ 230 ka (Hobden et al. 1996; Pure et al. 2020). Andesite boulders dated at 304 ± 11 ka (all age errors henceforth are 2 sd) that were ejected during an ~ 11 ka eruption from southern Tongariro, and a cryptic inlier of 512 ± 59 ka basaltic-andesite on northwestern Tongariro, suggest that buried materials at Tongariro pre-date 230 ka; therefore, the samples from the surficially exposed edifice record should be representative of compositional variation since ~ 230 ka (Heinrich et al. 2020; Pure et al. 2020).

Previous petrological studies of Tongariro eruptives have considered general aspects of the petrology of much of the edifice (Cole 1978; Hobden 1997), but specific links to the long-term evolution of the volcano have not been made (Hobden et al. 1999; Pure et al. 2020). Locally detailed studies of geochemical and isotopic characteristics have been undertaken on the products of the Ngāuruhoe cone (Rogan 1996; Davidson et al. 2007; Price et al. 2010; Cocker et al. 2021) but otherwise published isotopic data are lacking for the volcano. In addition, numerous studies have focused on the young (< 20 ka) history of eruptive activity, particularly on the growth of the Ngāuruhoe cone and regionally distributed pyroclastic products (Topping 1973; Nairn et al. 1998; Nakagawa et al. 1998; Hobden et al. 1999, 2002; Moebis et al. 2011; Arpa et al. 2017; Heinrich et al. 2020). In contrast, we focus on presenting a representative suite of data from the entire edifice record of the volcano to address the issues considered in the introduction. The chronology and names of units described here are summarised in supplementary file 4.

Analytical methods

Whole-rock analyses were made on fresh lava and pyroclastic samples (100–500 g) from the interiors of flows, and otherwise always on unaltered material, which were crushed with hardened or high-Cr steel crushers. Xenoliths encountered during the crushing process were removed and reserved for separate analysis (Pure et al. 2021). Microcrystalline groundmass material in the 250–350 micron size fraction was separated by standard magnetic and hand-picking separation methods from suitable whole-rock samples for ⁴⁰Ar/³⁹Ar age determinations (Pure et al. 2020). Strategically, aliquots of these same groundmass separates (that represent melt compositions) were reserved and also analysed for trace elements and isotopes for comparison with whole-rock samples (that comprise crystallised and quenched melts, crystal cargo and small country-rock xenoliths). The objective here was to ascertain the link between the whole-rock and the phenocrystic cargo. Where identifiable, crushed xenolith fragments were removed from devitrified groundmass material in the 250–350 micron size fraction. The xenolith-free groundmass separates (100–200 mg) were crushed with a clean agate mortar and pestle.

Major oxide concentrations of whole-rock samples were measured by X-ray fluorescence spectroscopy (XRF) using a Bruker AXS (WDS detector) S8 Tiger at the University of Waikato, Hamilton, New Zealand. Loss-on-ignition (LOI) mass changes were determined on 2 g of powdered rock after firing in alumina crucibles for 1 h at 1050 °C using a muffle furnace. Powdered rock (0.8 g) was fused with 8.0 g of BLiO 12:22 flux at 1050 °C for 20 min and moulded into glass discs. Analytical precision and accuracy were monitored with standards AGV-1, BCR-2, BHVO-2, OREAS, WS-E and an internal standard (andesite lava sample LP009B: Pure 2020). Analytical accuracy is within 5% (1 rsd) for each oxide concentration but generally ≤ 2.5% (1 rsd), as determined with rock standards compared with the recommended values of Jochum et al. (2016). Duplicate analyses of LP009B indicate an external precision (2 sd) better than 5% for all major oxides.

Trace element concentrations were determined via solution inductively coupled plasma mass spectrometry (ICP-MS) using a Thermo Scientific Element II at Victoria University of Wellington (VUW). Analyses were performed on 50–100 mg of powdered sample digested with concentrated HF and HNO₃ at a 20 HF:1 HNO₃ ratio, with the solution diluted appropriately in 1–3% HNO₃ for analysis. At the start of each analytical session, calibration curves (typical r² of 0.9994–1.0000) were determined with at least five multi-element standard solutions (0.2–20 ppb) prepared from certified 1000 ppm single-element solutions (Inorganic Ventures™). Trace element concentrations were calculated from drift-corrected and back-subtracted counting intensities. The results for AGV-2, BHVO-2 and WS-E were compared with, then, if accurate, normalised to the recommended values of these standards reported by Jochum et al. (2016). This procedure was necessary to quantify Lu concentrations in the first three (of ten) analytical sessions. Eu/Eu* (europium anomaly) values were calculated as Eu_N/(sqrt(Sm_N*Gd_N)) after normalisation with the CI chondrite values of McDonough and Sun (1995). Overall uncertainties were calculated from the contributions from the weighing reproducibility (0.04–0.20%), internal analytical precision (0.5–2.0%, 1 rsd) and external reproducibility (0.6–13.3%) that was monitored with internationally recognised standards AGV-2, BHVO-2 and WS-E against the recommended values reported by Jochum et al. (2016). When combined quadratically, the overall uncertainty in trace element concentrations for most elements is 0.5–9% (2 sd). External reproducibilities are given in supplementary file 1.

Sr–Nd–Pb isotopic compositions were measured by thermal ionisation mass spectrometry (TIMS) on a Thermo Finnegan Triton TIMS at VUW. Strontium, Nd and Pb separates were obtained following conventional techniques (e.g. Pin et al. 2014; supplementary file 2). All measurements were performed on outgassed, zone-refined Re filaments. Strontium samples were dissolved in 1 µL of concentrated HNO₃ and were loaded with a Ta₂F₅ activator onto single filaments, following Charlier et al. (2006). Nd samples were pre-dissolved in 0.01–1 M H₃PO₄ and loaded onto one side of a double filament assembly. Pb samples were pre-dissolved in silica gel and loaded onto single filaments. Pb isotopic compositions were determined with an in-house ²⁰⁷Pb-²⁰⁴Pb double-spike ‘Leeds DS’ with the following isotopic composition: 29.15% ²⁰⁴Pb, 6.26% ²⁰⁶Pb, 57.53% ²⁰⁷Pb, 7.06% ²⁰⁸Pb. Measurements of spiked and non-spiked samples were performed on entirely separate TIMS componentry (i.e. filament holders and slit plates), and the spiked and non-spiked data were reduced together using a modified Newton–Raphson inversion to determine the natural Pb isotopic composition and a 4000-iteration Monte-Carlo simulation to determine the uncertainty. Further details are in supplementary file 2.

TIMS analyses of recognised international solution standards (± 2 sd) gave NBS987 ⁸⁷Sr/⁸⁶Sr = 0.710269 ± 9 (n = 21); JNdi-1 ¹⁴³Nd/¹⁴⁴Nd = 0.512101 ± 25 (n = 25); NBS981 ²⁰⁶Pb/²⁰⁴Pb = 16.9422 ± 26, ²⁰⁷Pb/²⁰⁴Pb = 15.5002 ± 38, ²⁰⁸Pb/²⁰⁴Pb = 36.7278 ± 89, ²⁰⁷Pb/²⁰⁶Pb = 0.91489 ± 10 and ²⁰⁸Pb/²⁰⁶Pb = 2.16782 ± 25 (n = 46). All Sr–Nd-Pb isotopic data reported or plotted in this study were normalised to the recommended values of NBS987 ⁸⁷Sr/⁸⁶Sr = 0.710248 (Thirlwall 1991); JNdi-1 ¹⁴³Nd/¹⁴⁴Nd = 0.512115 (Tanaka et al. 2000) which equates to La Jolla ¹⁴³Nd/¹⁴⁴Nd = 0.511858 (Lugmair and Carlson 1978); NBS981 ²⁰⁶Pb/²⁰⁴Pb = 16.9409, ²⁰⁷Pb/²⁰⁴Pb = 15.4956, ²⁰⁸Pb/²⁰⁴Pb = 36.7228, ²⁰⁷Pb/²⁰⁶Pb = 0.91469, ²⁰⁸Pb/²⁰⁶Pb = 2.1677 (Thirlwall 2000). Unprocessed TIMS data and standard data are in supplementary files 5–7. After normalisation to standard values, AGV-2 gave averages of ⁸⁷Sr/⁸⁶Sr = 0.703979 ± 8, ¹⁴³Nd/¹⁴⁴Nd = 0.512782 ± 17, ²⁰⁶Pb/²⁰⁴Pb = 18.90766 ± 30, ²⁰⁷Pb/²⁰⁴Pb = 15.6112 ± 43, ²⁰⁸Pb/²⁰⁴Pb = 38.5669 ± 92, ²⁰⁷Pb/²⁰⁶Pb = 0.82571 ± 12 and ²⁰⁸Pb/²⁰⁶Pb = 2.03986 ± 21; BHVO-2 gave ⁸⁷Sr/⁸⁶Sr = 0.703483 ± 9 and ¹⁴³Nd/¹⁴⁴Nd = 0.512986 ± 9 (BHVO-2 not analysed for Pb isotopic compositions). For ⁸⁷Sr/⁸⁶Sr and ¹⁴³Nd/¹⁴⁴Nd, all individual AGV-2 and BHVO-2 values are within analytical error of those reported by Raczek et al. (2003), and for Pb isotope ratios all AGV-2 analyses are within error of recommended ‘leached’ values reported by Todd et al. (2015), after normalisation using the scheme above. Except where otherwise noted, all uncertainties reported for geochemical data are the quadratically combined within-run precision (2 se) and external reproducibilities (2 sd), including for the USGS standards above.

Results

Petrographic characteristics

The petrographic characteristics of Tongariro edifice-forming materials are reported in Pure et al. (2020) and representative features are shown in Fig. 3. In general, most whole-rock samples comprise plagioclase, orthopyroxene, clinopyroxene, olivine and Fe–Ti oxide phenocrysts within a glassy to holocrystalline matrix, which is sometimes flow-banded (compare points 4 and 5 in Fig. 3a). Silicate minerals, especially plagioclase, display a variety of internal textures and compositional zoning features, often with overgrowth rims, that indicate multiple generations or populations of the same mineral in individual samples (compare points 1 and 2 in Fig. 3a; see also Fig. 3d). Pyroxenes commonly occur as individual crystals but also as glomerocrysts, sometimes with plagioclase, Fe–Ti oxides and/or apatite (Fig. 3a, e). Crystals in glomerocrysts possess similar zoning and internal textures to co-occurring phenocrysts, which suggest a cogenetic origin. Crystals in glomerocrysts differ from the highly recrystallised, anhedral grains in metasedimentary xenoliths (Graham 1987; Pure et al. 2021). Olivine phenocrysts are uncommon but often contain Cr-spinels, and the host olivine is usually jacketed by an orthopyroxene reaction rim (Fig. 3b, c), as has been previously reported (e.g. Cole 1978). Fe–Ti oxides are present in ~ 90% of lavas (Hobden 1997) although highly vesicular, glassy and pyroclastic materials typically have lower abundances of Fe–Ti oxides owing to secondary crystallisation of these oxides in samples with completely devitrified groundmasses (Pure et al. 2020). Trace amounts of apatite (≤ 0.5 vol%) occur in dacites and high-silica andesites, but zircon, biotite and alkali feldspar are absent throughout. Hornblende is only significant in 349–293 ka Tupuna Formation andesites where it is the most abundant ferromagnesian mineral (Pure et al. 2020). Groundmass materials analysed are generally > 90% (micro)crystalline and comprise predominantly plagioclase (euhedral, subhedral and lath-like) with lesser amounts of pyroxene generally as orthopyroxene (subhedral and stubby) and equant, anhedral Fe–Ti oxides generally as magnetite (Hobden 1997). Groundmass grains are generally 1–20 µm but sometimes up to 50 µm in completely devitrified samples.

Geochemistry

Representative geochemical analyses for major oxides and trace elements are contained in Table 1 (whole-rock samples) and Table 2 (groundmass separates). Strontium, Nd and Pb isotopic compositions of whole-rock samples and groundmass separates are presented in Table 3. The full geochemical and isotopic dataset is presented in Figs. 4, 5, 6, 7, 8, 9, 10, 11, 12 and 13, and is provided in supplementary files 4–7. Radiogenic isotope data on samples from the Te Pupu Formation (Ngāuruhoe cone), Red Crater Formation and the Lower Tama Member (not represented in samples of this study) are available in Price et al. (2010), Shane et al. (2017) and Cocker et al. (2021).

Table 1 Representative major oxide and trace element concentrations of Tongariro whole-rock samples

Full size table

Table 2 Representative major oxide and trace element compositions of Tongariro groundmass separates

Full size table

Table 3 Sr–Nd–Pb isotopic compositions of Tongariro groundmass separates and whole-rock samples

Full size table

Major oxides

Edifice-forming materials at Tongariro are medium-K andesites and less commonly basaltic-andesites and dacites (Fig. 4; terminology following Gill 1981). SiO₂ is strongly correlated with K₂O, CaO and Fe₂O₃* (representing all Fe as ferric hereafter), but SiO₂ vs. Na₂O and vs. MgO show greater scatter. The compositional fields for Tongariro and Ruapehu closely overlap for these major oxides, except for SiO₂ vs. MgO data wherein Tongariro samples have generally lower MgO at a given SiO₂ concentration. Like other andesites globally (e.g. Gill 1981), Tongariro Fe₂O₃*, MgO and CaO concentrations decrease with increasing SiO₂ whereas K₂O and Na₂O increase.

Tongariro SiO₂ vs. MgO data define two curvilinear arrays for high-MgO and normal-MgO andesites (Fig. 4), which compare with a more prominent high-MgO array in Ruapehu samples (Conway et al. 2018). The Tongariro high-MgO array begins at ~ 9 wt% MgO and ~ 55 wt% SiO₂ and converges towards the normal-MgO array at ~ 60 wt% SiO₂ and ~ 4 wt% MgO. Most of the Tongariro samples that follow the high-MgO andesite array were erupted from flank vents (e.g. the 40–30 ka Pukeonake Formation), but also include upper edifice material (e.g. the 152–150 ka Waiaruhairiki and 130–96 ka Mangahouhouiti members) that contain minor proportions (≤ 5 vol%) of olivine (Pure et al. 2020). There is no significant relationship between whole-rock MgO concentrations and the relative abundance of pyroxenes because orthopyroxene occurs both more and less abundantly than clinopyroxene in samples across all MgO concentrations (Pure et al. 2020).

Scattered Al₂O₃ concentrations in Tongariro basaltic-andesites (14–18 wt%) converge to 16–17 wt% in dacites, in similar fashion to the corresponding Ruapehu lavas (Fig. 4b) (Graham and Hackett 1987; Price et al. 2012; Conway et al. 2018). At Tongariro, high TiO₂ concentrations (~ 1 wt%) are more common in mafic samples (~ 55 wt% SiO₂) than in high-silica andesites, but TiO₂ is uncorrelated with SiO₂ (Fig. 4a). In contrast, Ruapehu samples only reach ~ 0.8 wt% TiO₂ for equivalent SiO₂ concentrations (~ 55 wt%).

Trace elements

Trace element concentrations shown in Fig. 5 have Fe₂O₃* on the X-axis because SiO₂ concentrations could not be determined on HF-digested groundmass separates (analysed by solution ICP-MS), and because SiO₂ and Fe₂O₃* are demonstrably strongly correlated in the whole-rock data (Fig. 4c). Rb, Ba and Zr increase (Fig. 5) as whole-rock and groundmass compositions become more evolved (i.e. decreasing Fe₂O₃* which equates to increasing SiO₂: Fig. 4c). Concentrations of P₂O₅ and Y are relatively constant across the range of Fe₂O₃* values, while Sr, Sc and Ni decrease as Fe₂O₃* decreases (or SiO₂ increases).

Whole-rock samples display light rare earth element (REE) enrichment relative to heavy REEs, with (La/Yb)_N ratios of 4.5–10.2, averaging 6.5 (Fig. 6). Negative Eu anomalies are greater in samples with high REE concentrations, which coincides with small negative Ce anomalies in the most REE-rich samples. These REE data are compared with three regional basalts that have been widely used in geochemical modelling (Price et al. 2005, 2012; Conway et al. 2018): Kakuki and Waimarino from the central TVZ and a Havre Trough basalt from the offshore oceanic back-arc north of the North Island (Gamble et al. 1996; Barker et al. 2015; Waight et al. 2017). REE concentrations for Tongariro are similar to the Kakuki and Waimarino basalts but are different from the Havre Trough basalt, which has low light-REE concentrations relative to heavy REEs. In contrast, REE concentrations in Tongariro groundmass separates span a wider range than whole-rock samples (Fig. 7). Negative Ce anomalies are also more pronounced in REE-rich groundmass separates (Fig. 7) than in corresponding whole-rock samples (Fig. 6a).

Sr–Nd–Pb isotopes

Whole-rock and groundmass Sr, Nd and Pb isotopic compositions are given in Table 3 and are presented in Fig. 8. A comparison of isotopic ratios in the whole-rock vs. groundmass components from the same samples is shown in Fig. 9. For each isotopic system, Tongariro whole-rock and groundmass compositions form curvilinear arrays that scatter between regional basalts and metasedimentary basement rocks that likely comprise the Kaweka and Waipapa terranes (Fig. 8; Pure et al. 2021). Notably, Tongariro whole-rock and groundmass ¹⁴³Nd/¹⁴⁴Nd ratios are within analytical error of each other and of the 1:1 equal ratios line, except for the sample with the highest ¹⁴³Nd/¹⁴⁴Nd ratios (Fig. 9b). The tie lines in Fig. 8a show that ¹⁴³Nd/¹⁴⁴Nd ratios in groundmass separates are sometimes more crust-like than whole-rock samples, but other pairs show the reverse relationship. For ⁸⁷Sr/⁸⁶Sr and Pb isotope ratios, whole rock–groundmass pairs commonly differ beyond analytical error, and there is no systematic affinity for more crustal or mantle-like isotopic compositions in the groundmass versus whole-rock samples (Figs. 8 and 9).

Discussion

Magma assembly processes at Tongariro

The origins of volumetrically dominant andesitic magmas at arc stratovolcanoes have been debated using global datasets of rock and melt inclusion compositions (e.g. Reubi and Blundy 2009; Lee and Bachmann 2014). Andesites often contain physical evidence for being hybrid magmas in the form of mafic enclaves and multiple generations of crystals with disequilibrium textures (e.g. Gill 1981; Eichelberger et al. 2006; Humphreys et al. 2006; Streck and Leeman 2018; Conway et al. 2020). It is still debated, however, where and when andesite magmas are generated in the crust, in the mantle, or across the crust–mantle boundary and what their relationships are to less common, cogenetic arc basalts and rhyolites (e.g. Annen et al. 2006; Hildreth 2007; Reubi and Blundy 2009; Lee and Bachmann 2014; Blundy 2022). To address this issue, Tongariro compositional data are examined to identify which processes are implicated in generating these intermediate magmas and to estimate how much crust they have assimilated.

Crystallisation differentiation

The mafic enclaves, crystal clots (containing plagioclase, pyroxene, Fe–Ti oxides and apatite) and country-rock xenoliths in Tongariro andesites indicate that recharge, crystallisation-differentiation and assimilation processes contributed to magma assembly (Fig. 3; Pure et al. 2021). Of these processes, crystallisation differentiation is considered first using major oxide and trace element data in Figs. 5 and 10, which are interpreted noting the preference of certain elements to partition into certain minerals (i.e. following the Kd values compiled by Gill 1981 and Rollinson 1993, and references therein).

Ni vs. Fe₂O₃* (Fig. 5h) and Ni vs. MgO (Fig. 10a) arrays provide evidence for olivine fractionation in basalts and basaltic-andesite magmas above 8 wt% MgO. This is supported by the occurrence of olivine in primitive Tongariro eruptives (~ 8 wt% MgO) relative to its scarcity in eruptives with < 8 wt% MgO (Pure et al. 2020). Invariably, however, olivine crystals are jacketed by orthopyroxene overgrowths, indicating that they are not in equilibrium with the host melt. Below 8 wt% MgO, pyroxene fractionation likely drives further MgO decreases that occur concurrently with decreasing Sc, from ~ 35 to 15 ppm. Other compositional data (e.g. Y/Co vs. MgO, not shown) provide no indication that clinopyroxene and orthopyroxene fractionation occur at different stages of magma evolution. This is also supported by Pure et al. (2020) who report that MgO concentrations are not correlated with orthopyroxene/clinopyroxene ratios in Tongariro whole-rock samples.

Rb/Sr vs. CaO/Al₂O₃ data (Fig. 10c) suggest that plagioclase fractionation commenced at ~ 0.55 CaO/Al₂O₃ (which corresponds to ~ 7 wt% MgO: see Fig. 10d), and scattered Rb/Sr ratios at andesitic SiO₂ concentrations suggest that plagioclase may have been recycled during magma assembly (Fig. 10e). However, Rb/Sr vs. Eu/Eu* relationships indicate that groundmasses generally have lower Eu/Eu* (i.e. a stronger negative Eu anomaly) and higher Rb/Sr than whole rocks (Fig. 10e). Outside of analytical uncertainty, groundmass–whole-rock tie lines show that Eu/Eu* values in the groundmass are generally lower (indicating plagioclase fractionation relative to the whole rock) in the same samples. Plagioclase fractionation is also indicated by groundmass Sr concentrations that are on average ~ 40 ppm lower than in whole rocks (Fig. 5c). In whole rocks, stronger negative Eu anomalies generally occur in samples with elevated Rb/Sr ratios (Fig. 10f) that equate to higher SiO₂ concentrations (Fig. 10e), which suggests that plagioclase fractionation (as opposed to accumulation) is the more dominant process in Tongariro magmas.

Y vs. Fe₂O₃* and P₂O₅ vs. Fe₂O₃* groundmass–whole-rock tie lines in Fig. 5 support the idea that apatite fractionation and accumulation occurred in Tongariro magmas, with little overall net change in P₂O₅ concentrations as magmas evolved. Increased Rb, Ba and Zr in the groundmass, relative to whole-rock concentrations (Fig. 5a, b, f), supports the observation that alkali feldspar, biotite and zircon did not crystallise as phenocryst phases in Tongariro magmas, which is shown by their complete absence in all lava and pyroclast samples from the combined collections of Hobden (1997) and this study. Although geochemical data do not clearly demonstrate Fe–Ti oxide fractionation, it probably occurred because Fe–Ti oxides are abundant in crystal clots with pyroxene, plagioclase and apatite that were likely removed from Tongariro magmas via crystallisation differentiation (Fig. 3). Based on the interpretations above, we adopt the following proportions for the fractionating mineral assemblages used to model the chemical evolution from basaltic-andesites to dacites: olivine (4%), plagioclase (32 wt%), orthopyroxene (32 wt%) and clinopyroxene (32 wt%).

Assimilation and recharge

Sr–Nd–Pb isotopic data indicate that both crustal and mantle-derived materials contributed to the generation of Tongariro magmas. This is supported by whole-rock and groundmass ⁸⁷Sr/⁸⁶Sr and ¹⁴³Nd/¹⁴⁴Nd ratios that form arrays between regional basalts and basement metasediments (Fig. 8), with the latter inferred to comprise the Jurassic Waipapa and Kaweka terranes (Pure et al. 2021). Bulk magmas also appear to have co-evolved with their co-existing melts as shown by similar (or indistinguishable) isotopic compositions in the whole-rock and groundmass components of each sample, which is observed across the full compositional ranges examined (Fig. 9a–d). Regression lines between whole-rock and groundmass concentrations are within error of 1:1 equal composition lines for Sr, Nd and Pb isotopic ratios (95% confidence). This result requires that either Sr–Nd–Pb isotopic compositions were initially set in magmas before they crystallised phenocrysts (with negligible crystal fractionation or subsequent crustal assimilation), or that evolving magmas progressively fractionated crystals whilst concurrently assimilating country-rock melts and restites that continued to modify Sr–Nd–Pb isotopic compositions.

Relevant to the latter interpretation is that metasedimentary basement xenoliths in Tongariro eruptives have been interpreted as restites comprising ~ 50% of their former masses, with the remaining ~ 50% assumed to be assimilated as melts (Graham 1987; Pure et al. 2021). The metasedimentary restites form up to 6%, but generally only 1–2%, of total whole-rock volumes of eruptives (Hobden 1997). Additionally, xenocrysts derived from disaggregated country-rock xenoliths also likely occur cryptically in Tongariro eruptives, which would add to the proportions of incorporated country-rock material just mentioned. Consequently, the latter interpretation (i.e. concurrent crystal fractionation and crustal assimilation) is favoured on petrographic evidence (restitic metasedimentary xenoliths) and on trace element data that imply the fractionation of olivine, clinopyroxene, orthopyroxene (Ni, MgO, Sc, CaO/Al₂O₃ relationships: Fig. 10a–d) and plagioclase (Rb/Sr, SiO₂ and Eu/Eu* relationships: Fig. 10e, f).

Genetic relationships between melts and co-existing crystal cargoes are also revealed by comparing whole-rock and groundmass trace element concentrations from the same samples (n = 34). On average, Sr concentrations in whole-rock samples are greater by 13 ± 20% (± 2 sd; -34 to + 113 ppm total range) whereas Nd concentrations are on average greater by 13 ± 20% in groundmass separates (± 2 sd; -0.8 to + 6.5 ppm total range). The opposing preferences for Sr in the whole-rock component (i.e. in crystals, especially plagioclase) vs. Nd in melts appears to have little effect on the general covariance of Sr and Nd isotope ratios in groundmass samples and their whole-rock hosts (Fig. 9a, b). Th and Zr relationships in paired whole-rock and groundmass samples also show statistically significant enrichments of these elements in the groundmass (Fig. 9e, f). These results support the inference that the isotopic compositions of Tongariro magmas were continually modified as they evolved, which parallels the general correlations between Ca concentrations and ⁸⁷Sr/⁸⁶Sr ratios in plagioclase at Tongariro (Shane et al. 2019). Collectively, these findings challenge the idea that compositional diversity and mineralogical diversity are developed separately in arc magmas via decoupled processes, as has been suggested in popular petrogenetic models (e.g. Annen et al. 2006; Blundy 2022).

Assimilation with fractional crystallisation

Chemical and thermodynamic AFC models are often applied to compositional datasets to estimate the proportions of mantle-derived and crust-derived materials in magmas under scenarios of country-rock fusion and concurrent fractional crystallisation (e.g. DePaolo 1981; Bohrson and Spera 2001). In their common application, the ratio of the mass assimilation rate (M_a) against the mass crystallisation rate (M_c), known as the r-value (= M_a/M_c), is usually predetermined and limited to less than 1. Whether such assumptions are appropriate in continental arc settings is questionable because mixing and in situ crystallisation differentiation (without crystal fractionation) are common magma assembly processes at some arc volcanoes. An example is at Mount Rainier in the Cascades where evolved in situ melts were mixed into less evolved magmas that ascended through the system, which have compositions that are predicted by AFC evolution at r-values greater than 1 (Sisson et al. 2014). In such cases, crystallisation and country-rock fusion may be thermodynamically balanced across the broader open system, but this may not be reflected in chemical and isotopic arrays. In the context of Tongariro (and Ruapehu), the presence of restitic metasedimentary basement xenoliths in eruptives indicates that r-values determined from compositional arrays will not necessarily reflect thermodynamically balanced crystallisation and country-rock fusion. This is because country-rock fragments were physically incorporated into magmas in full but did not melt completely (Pure et al. 2021). Accordingly, our estimates of the proportion of assimilated crust in Tongariro samples is pursued from a chemical and isotopic perspective, and not from a thermodynamic perspective.

Additionally, although the percentage of assimilated crust in an arc magma can be estimated from δ¹⁸O data, there are few studies that report oxygen isotopic compositions for Tongariro eruptives, appropriate mafic end-member magmas and assimilated country rocks. Hobden (1997) reports δ¹⁸O for pyroxene grains, which may include up to ~ 15% of xenocrystic pyroxene from disaggregated xenolithic restites based on modal proportions of xenoliths in Tongariro eruptives (< 2%: Hobden 1997) and pyroxene in metasedimentary xenoliths (< 30%: Pure et al. 2021). Beier et al. (2017) report δ¹⁸O for mantle-like olivines in the Pukeonake Formation basaltic-andesites of Tongariro, but how these values transpose to basaltic arc magma compositions is unclear because the groundmass glass has mixed with more evolved melts. Other crystals in Pukeonake Formation basaltic-andesites were likely derived or crystallised from magmas that were more evolved than basalts. Regarding crustal assimilants, Graham et al. (1990) report δ¹⁸O for xenoliths in Ruapehu eruptives, which are similar to values reported by McCulloch et al. (1994) for the Waipapa and Kaweka terranes that comprise the upper crust (Price et al. 2015; Pure et al. 2021). Estimating the percentage of assimilated crust in mafic magmas that replenish the Tongariro system would therefore require arbitrary choices of mineral-melt fractionation correction values. Notwithstanding these limitations, mass balance with δ¹⁸O values reported by Graham et al. (1990), Hobden (1997) and Beier et al. (2017) would likely yield around ~ 10–25% of assimilated upper crustal metasedimentary rocks in Tongariro magmas, provided that there is no contribution from lower crustal contaminants that have largely unknown lithologies and compositions.

Geochemical modelling of magma evolution

Tongariro whole-rock and groundmass compositions are compared with AFC and mixing model results in Fig. 11. The models also use a fractionating mineral assemblage of olivine (4%), plagioclase (32 wt%), orthopyroxene (32 wt%) and clinopyroxene (32 wt%), following the discussions above. A mix of 50% average Kaweka terrane and 50% average Waipapa terrane is adopted for the bulk composition of the crustal contaminant, using the values of Price et al. (2015) and following Pure et al. (2021). The most primitive (53.1 wt% SiO₂) eruptive sampled from the Tongariro edifice is adopted as the starting composition for the models, which is the Te Rongo (TR) basaltic-andesite (Pure et al. 2020). This end-member composition is denoted by the yellow squares in Figs. 4, 5, 6, 7, 8 and 10.

Although the Tongariro magma system was likely replenished by magmas more primitive than the TR basaltic-andesite, such as the Kakuki and Waimarino arc basalts, neither of these basalts individually represent a plausible basaltic end-member across key trace elements (e.g. Sr, Rb, Nd) and isotope ratios (Figs. 5, 6, 7, 8, 9 and 10). For example, the Kakuki basalt may be an appropriate end-member for Sr and Rb (Fig. 10e, g), however, its Nd concentration is higher than the TR basaltic-andesite (Fig. 10h). Additionally, the Waimarino basalt could be considered an appropriate end-member from Nd concentrations (Fig. 10h), but its ¹⁴³Nd/¹⁴⁴Nd ratio is more crust-like than many higher-SiO₂ Tongariro andesites (Fig. 8a). The Waimarino basalt also has an elevated ⁸⁷Sr/⁸⁶Sr ratio (0.7044) that is more crust-like than other regional basalts (Fig. 8a: Gamble et al. 1993).

Whole-rock and groundmass samples generally plot between mixing curves and low r-value (= 0.1) AFC curves (Fig. 11); however, both intermediate and silicic melt Kd values are needed to model the entire Tongariro compositional array. Mixing and fractional crystallisation pathways for SiO₂ concentrations and ⁸⁷Sr/⁸⁶Sr ratios show that varied amounts of crystal fractionation and assimilation are needed to explain the compositional diversity of Tongariro samples because Tongariro data do not follow the ‘crystal fractionation only’ trajectory (Fig. 11a). Trial and error plotting shows that AFC curves adopt shapes and positions similar to mixing curves when r-values approach 1 (not shown).

Mixing curves and AFC curves follow similar trajectories in ⁸⁷Sr/⁸⁶Sr vs. ¹⁴³Nd/¹⁴⁴Nd space (Fig. 11b). At r-values of 0.3, AFC models predict that magmas will attain evolved ⁸⁷Sr/⁸⁶Sr (= 0.7062) and ¹⁴³Nd/¹⁴⁴Nd (= 0.51263) ratios with 10–50% remaining melt, or by mixing subequal proportions of TR basaltic-andesite and metasedimentary basement rocks. Variations in whole-rock Nd concentrations and ¹⁴³Nd/¹⁴⁴Nd ratios cannot be reproduced by a single AFC or mixing model, and require models that use both silicic and intermediate melt Kd values and r-values between 0.1 and 1 (Fig. 11c). Other modelling results for ⁸⁷Sr/⁸⁶Sr, La/Yb, Rb/Sr, Sc and Zr further support the inference that Tongariro magmas evolved along AFC pathways with r-values of 0.1 to 1 in the presence of silicic and intermediate composition melts (Fig. 11d–f).

Using an r-value of 0.5 for a typical AFC pathway of a Tongariro magma that evolved from a TR basaltic-andesite composition onwards, the percentage of remaining melt may be between ~ 20 and 100%, depending on the sample (Fig. 11). In a simple hypothetical case where only mantle-derived material is fractionated (and all assimilated materials remain in the evolving magma), the percentage of assimilated crust acquired in this stage of magma evolution is approximated by ~ 0.5 (average r-value) multiplied by ~ 40% crystallisation (~ 0–80% total range) of the original magma. This indicates that magmas evolving from a TR basaltic-andesite starting composition were contaminated with ~ 20% of assimilated metasedimentary basement rocks.

In some cases, r-values calculated from compositional arrays may not represent thermodynamically balanced processes because assimilation can occur without total melting of wall rocks and because conductive heat exchange can occur in magmas without mixing. On the former point, detailed comparisons of xenoliths and likely protolithic materials at Tongariro suggest that approximately 50% of each assimilated (meta)sedimentary xenolith was melted when incorporated in its host magma (Pure et al. 2021). The other ~ 50% of chemically incorporated material comprises disaggregated, restitic, country-rock xenoliths and xenocrysts that are seen petrographically in erupted materials. With this constraint, the AFC modelling results in Fig. 11 show that magmas probably evolved at r-values of 0.1 to 1, or thermodynamically balanced r-values of ~ 0.05–0.5 because of the incomplete melting of incorporated country rock. Conversely, assimilation that occurs at peripheral areas of a magma may be fuelled by heat conducted from inner areas of the magma, which may not be followed by mixing and homogenisation. In such cases, incorporated wall rock melts and restites may be concentrated in peripheral areas of a magma body that could ascend and erupt independently, and be falsely interpreted as the products of closed-system AFC evolution at r-values that exceed 1.

Collectively, the AFC modelling results (and highly approximate δ¹⁸O mass balance estimates) above indicate that the average Tongariro magma (basaltic-andesite to dacite) contains ~ 20% of assimilated metasedimentary basement rocks. While this estimate provides some constraints on the heat budget of the Tongariro magmatic system, the unerupted proportion of each magma that remains as cumulates or mush is difficult to estimate because the r-values in the early stages of magma evolution, at basaltic compositions, are poorly constrained. Moreover, although geophysical surveys (in particular using magnetotelluric methods) indicate that magmatic bodies containing up to ~ 45% melt exist down to ~ 12 km depth beneath the Tongariro edifice (Hill et al. 2015; Miller et al. 2018), the volume of intruded but unerupted magma is also largely unconstrained.

Time–composition relationships and the organisation of the magmatic system

The sequential evolution of magma compositions at a volcano reflects the timing and tempo of petrogenetic processes and the degree of organisation of the magmatic system. It is generally accepted that geochemical data should be ordered sequentially when making petrogenetic interpretations because compositions rarely evolve unidirectionally over a volcano’s entire lifespan (e.g. Eichelberger et al. 2006). To address this issue, Tongariro compositional data are integrated with a high-resolution edifice construction record, from ~ 350 ka to present (Pure et al. 2020).

Time–composition data for MgO and K₂O are shown in Fig. 12 for individual samples (points) and stratigraphic unit ranges (crosses). The MgO data display irregular cycles that have been interpreted as mafic replenishment events that commenced at ~ 230, ~ 151, ~ 88 and ~ 56 ka and in the late Holocene, with high-MgO flank vents erupting at ~ 160, ~ 117, ~ 35 and ~ 17.5 ka (Pure et al. 2020). These events are each followed by 10–70 kyr intervals where MgO concentrations gradually decrease from ~ 9–5 wt% to ~ 4–2 wt%. These intervals are inferred to reflect prolonged crystal fractionation because the opposite pattern occurs in K₂O vs. time data; that is, K₂O increases as MgO decreases. Time–composition data prior to 250 ka have a lower temporal resolution and are not interpreted.

In contrast, time–composition trends in Sm/Nd and Sr–Nd–Pb isotopic compositions, which are shown for samples (points) and stratigraphic unit ranges (crosses) in Fig. 12, occur on 30–130 kyr cycles and do not align with the MgO and K₂O trends. This result does not appear to be an artefact of lower temporal resolution in the Sr–Nd–Pb isotopic dataset. For example, the Waiaruhairiki Member and its sample (LP245) define an MgO maximum (5.5 wt%) at ~ 150 ka (Fig. 12a) but also occur in the middle of the cycle to increasingly crust-like ⁸⁷Sr/⁸⁶Sr (0.7054), ¹⁴³Nd/¹⁴⁴Nd (0.51272) and ²⁰⁶Pb/²⁰⁴Pb (18.848) ratios at ~ 150 ka (Fig. 13d–f). In general, the evolution toward more crustal ⁸⁷Sr/⁸⁶Sr, ¹⁴³Nd/¹⁴⁴Nd, Sm/Nd and ²⁰⁶Pb/²⁰⁴Pb ratios (shown by the green lines in Fig. 13; Price et al. 2015; Pure et al. 2021), is interpreted to show periods of prolonged crustal residence and crustal assimilation from ~ 230 to ~ 100 ka and ~ 95 to ~ 30 ka. The timings of these cycles, relative to the MgO vs. time trends, indicate that the episodes of mafic replenishment and prolonged crustal residence (with associated crustal assimilation) are partially decoupled at Tongariro.

Such systematic time–composition trends indicate a moderate degree of organisation in Tongariro’s magma system, at least beneath the central edifice area which is the inferred vent area for most eruptive units represented in the long-term record (Pure et al. 2020). The gradual and cyclic shifts in the concentrations of compatible elements (e.g. MgO in olivine and pyroxenes), incompatible elements (e.g. K₂O) and radiogenic isotopes show that magma bodies of at least 1 km³ (i.e. the erupted component) both evolved and intermittently erupted over ≥ 10 kyr periods. Notably, the relatively low rate of compositional variation in the long-term eruptive record since ~ 350 ka (Fig. 12) contrasts with the rapid compositional change observed in small-volume (< 1 km³) Holocene eruptives (e.g. Hobden et al. 1999; 2002). For example, the entire range of ⁸⁷Sr/⁸⁶Sr ratios over the long-term record occurs within eruptives from the Ngāuruhoe cone at Tongariro alone, which have ages of ≤ 7 ka and individual flow volumes of < 0.1 km³, but more often < 0.01 km³ (Hobden et al. 1999, 2002; Pure et al. 2020).

Similar composition vs. volume relationships occur at Ruapehu (Gamble et al. 1999; Conway et al. 2018), and it appears that the smaller the eruptive-volume window that is examined, the larger the compositional diversity. The modern situation at Tongariro and Ruapehu may in fact represent the smaller-scale ‘background’ activity that has been present through much of the lifetime of both volcanoes but remains generally unrecorded in the long-term eruptive record (e.g. Gamble et al. 1999). With the caveat that erupted volumes may not correlate with magma batch volume, because magma bodies feeding eruptions may only be partially evacuated, we conclude that there is a lower chance of observing rapid, compositional heterogeneity in the edifice record because smaller-volume eruptives are less likely to be preserved or sampled.

The observations above can be reconciled in a model of the Tongariro magma system wherein magma batches of diverse volumes and compositions coexist in the crust. The rapid compositional variation observed in smaller-volume eruptives, such as from the Ngāuruhoe cone, may arise when ascending magmas avoid or only briefly transit through longer-lived (and potentially larger and more integrated) parts of the magma system. Concurrently, other magmas may instead stall and homogenise to produce the gradual, systematic compositional trends that are recorded periodically when the longer-lived system is tapped during eruptions (e.g. Figure 12). This process has been inferred elsewhere, such as at Mount St. Helens, where antecrystic > 100 ka zircons occur in < 100 ka eruptives, despite thermal perturbations to 820–940 °C having occurred at ~ 100 ka (above zircon saturation temperatures of 740–780 °C), which would have dissolved zircons in ~ 100 years (Claiborne et al. 2010). Reconciling these observations requires brief (< 100 year) interactions between the hot, mafic magmas that ascended through the cooler, zircon-saturated magma system. Similarly, ascending mafic magmas at Tongariro appear to either recharge the longer-lived system where they hybridise and evolve or, instead, they may erupt peripherally after brief, limited interactions with the longer-lived, predominantly andesitic system.

Although the repeated 10–130 kyr compositional cycles demonstrate a degree of organisation in Tongariro’s magmatic system (Fig. 12), their cyclicity suggests that any long-term, thermal conditioning of the crust occurs at non-constant rates. This observation contrasts with simpler cases of uniform heating that are adopted by thermochemical models of crustal evolution in arc settings (Smith et al. 2003; Annen et al. 2006). The time–composition trends at Tongariro show that magmas were produced by a moderately organised magmatic system (i.e. showing gradual compositional changes), but that each time–composition cycle was of a different duration. Any compositional trends in magma compositions induced by thermal conditioning of the crust, that arise from repeated intrusions of mantle-derived magmas (Annen et al. 2006), appear to be undetectable or over-printed by stronger time–composition variations that were induced by cyclic petrogenetic processes. We propose that the cyclic compositional trends at Tongariro (Fig. 12) likely arose from open-system processes that include temporally variable mafic intrusion rates and crustal residence durations (when AFC evolution occurred) that can both be modulated by tectonic processes.

Other modelling and across-arc studies have indicated that most compositional diversity in arc magmas is acquired in the deep crust, with petrographic diversity being developed later (Hildreth and Moorbath 1988; Annen et al. 2015). The Tongariro dataset presented here is, however, difficult to fully reconcile with that view. Partial conflict arises because (i) several (if not tens of) percent of total whole-rock material comprises assimilated crust in the form of restitic metasedimentary xenoliths and melts (Fig. 3; Hobden 1997; Graham 1987; Pure et al. 2021); (ii) sample petrography (Fig. 3) and trace element and major oxide relationships indicate that olivine, clinopyroxene, orthopyroxene and plagioclase fractionated from magmas (Figs. 5 and 10); (iii) correlated Ca concentrations and ⁸⁷Sr^/86Sr ratios in plagioclase (Shane et al. 2019) and similar (but often indistinguishable) Sr–Nd–Pb isotopic compositions in crystal cargoes and co-existing melts (Fig. 9) suggest that crystal growth and fractionation occurred concurrently with crustal assimilation; (iv) the compositions of restitic metasedimentary xenoliths at Tongariro are similar to the compositions of the Kaweka and Waipapa terranes that are the inferred sources of the xenoliths (Graham 1987; Hobden 1997; Price et al. 2015; Pure et al. 2021); (v) the Kaweka and Waipapa terranes are inferred to comprise the uppermost 15 km of the crust beneath the Tongariro–Ruapehu area which indicates that assimilation occurred in the upper crust (Harrison and White 2006; Behr et al. 2011; Stern and Benson 2011); and (vi) Tongariro Sr–Nd–Pb isotopic compositions in whole-rock and groundmass samples plot between a mafic basaltic-andesite end-member and the Sr–Nd–Pb isotopic compositions of Kaweka and Waipapa terrane reference materials which shows that these terranes are plausible contaminants of evolving Tongariro magmas (Fig. 8). In contrast with the findings of Hildreth and Moorbath (1988) and Annen et al. (2015), we conclude that crystal fractionation and upper crustal assimilation at Tongariro occurred concurrently and in sufficient amounts to explain the full range of Sr–Nd–Pb isotopic compositions in the combined datasets of Hobden (1997) and this study.

We note that the assimilation of the lower crust into Tongariro magmas cannot be discounted because the composition and lithology of lower crustal materials are largely unknown, so their influence on Tongariro magma compositions cannot be accurately assessed. Geobarometry estimates from a limited number of Tongariro samples suggest that some crystal growth (and presumably crustal assimilation) occurred in lower crustal areas at depths of < 25 km or < 7 kb (Hobden 1997; Arpa et al. 2017). However, representative geobarometry estimates for the greater Tongariro eruptive history are lacking. Therefore, on the available evidence, it is most likely that the radiogenic isotopic compositions of Tongariro eruptives were acquired from assimilating subequal (if not predominant) amounts of upper crustal rocks relative to lower crustal rocks.

Long-term drivers and aspects of magmatic systems at Tongariro and neighbouring volcanoes

Tongariro and neighbouring Ruapehu are predominantly andesitic composite stratovolcanoes comprising several cones and vent systems (Cole 1978; Hobden et al. 1996; Gamble et al. 1999, 2003; Conway et al. 2016; Pure et al. 2020). The two volcanoes are separated by ~ 20 km along the NNE–SSW-trending axis of the TVZ arc and have independent magmatic roots (Leonard et al. 2021). Tongariro’s closest neighbouring volcano to the north that was active during the Holocene is Taupō volcano, which has been a source of voluminous rhyolitic eruptions, distinguishing it from the andesitic southern TVZ volcanoes (Wilson et al. 1995).

Compared with Tongariro, Ruapehu has a potentially more organised magma system that is reflected in its more systematic time–composition trends in major oxide, trace element and radiogenic isotopic compositions in eruptives from ~ 230 ka to present (Price et al. 2012; Conway et al. 2018). Like Tongariro, the compositions of historic and small-volume Ruapehu eruptives also span the entire range seen in the long-term edifice record at Ruapehu since ~ 230 ka (Gamble et al. 1999). Overall, the compositional records suggest that magmatic systems at Tongariro and Ruapehu comprise multiple magma bodies with a range of sizes, compositions and longevities that coexist at any one time (Nakagawa et al. 1999, 2002; Gamble et al. 2003; Kilgour et al. 2013).

Time–composition patterns at Tongariro vs. Ruapehu

A brief comparison of time–composition trends at Tongariro and Ruapehu, as inferred from their erupted products, is presented in Leonard et al. (2021). Here, we expand on this comparison with a compilation of Rb/Sr, ⁸⁷Sr/⁸⁶Sr and ¹⁴³Nd/¹⁴⁴Nd vs. age data from both volcanoes (Fig. 13; Hobden 1997; Gamble et al. 2003; Price et al. 2012; Conway et al. 2018; Pure et al. 2020). In general, time–composition trends at Tongariro and Ruapehu are similar and occur contemporaneously for ⁸⁷Sr/⁸⁶Sr and ¹⁴³Nd/¹⁴⁴Nd ratios but are less coordinated for MgO and Rb/Sr. Temporal MgO and Rb/Sr trends in ~ 200–100 ka eruptives from Tongariro and Ruapehu do not overlap. MgO concentrations at Ruapehu decrease from ~ 8 to ~ 2 wt% between ~ 45 and 40 ka and are mirrored by increased Rb/Sr ratios from ~ 0.2 to ~ 0.6 (Gamble et al. 2003; Price et al. 2012; Conway et al. 2016, 2018). Covariance in MgO values and Rb/Sr ratios at Tongariro is also observed, as shown by the interpreted grey trend lines (Fig. 13a, b), but the Rb/Sr variations are smaller than those for Ruapehu and do not increase significantly in the ~ 45–40 ka period. Rb/Sr ratios in the Ruapehu 200–150 ka Te Herenga and 166–80 ka Wahianoa formations have little overlap with Tongariro Rb/Sr ratios of ~ 0.1–0.4 in the similarly aged 290–189 ka Haumata, 189–130 ka Mangahouhounui and 133–52 ka Taiko formation eruptives (Tongariro stratigraphic definitions in supplementary file 4, from Pure et al. 2020; Ruapehu definitions from Gamble et al. 2003; Price et al. 2012; Conway et al. 2016, 2018; Townsend et al. 2017). At the two volcanoes, broadly contemporaneous satellite vent eruptions occurred at 40–30 ka (Pukeonake Formation on western Tongariro: Fig. 1) and at 35.4 ± 0.6 ka (Ohakune Formation southwest of Ruapehu: Fig. 1), both of which comprise eruptives with elevated MgO concentrations of ~ 6–11 wt% (Froggatt and Lowe 1990; Beier et al. 2017; Pure et al. 2020; Svoboda et al. 2022).

Tongariro and Ruapehu display contemporaneous time–composition cycles in ⁸⁷Sr/⁸⁶Sr and ¹⁴³Nd/¹⁴⁴Nd ratios (Fig. 13c, d). Ruapehu’s 200–150 ka Te Herenga Formation has ⁸⁷Sr/⁸⁶Sr ratios of ~ 0.7048–0.7053 that are interposed between the ~ 230–190 ka Tongariro Haumata Formation ⁸⁷Sr/⁸⁶Sr ratios of ~ 0.7047–0.7052 and 189–130 ka Tongariro Mangahouhounui Formation ⁸⁷Sr/⁸⁶Sr ratios of ~ 0.7051–0.7055 (Price et al. 2012; Conway et al. 2016). Overlapping ⁸⁷Sr/⁸⁶Sr ratios centred at ~ 0.7055 are observed for the similar-age 166–80 ka Ruapehu Wahianoa Formation and Tongariro’s 189–130 ka Mangahouhounui and 133–52 ka Taiko formations. ⁸⁷Sr/⁸⁶Sr ratios in Mangawhero Formation eruptives at Ruapehu increase from ~ 0.7050 to 0.7062 between ~ 45 and 40 ka, and then steadily decline to ~ 0.7052 for Whakapapa Formation eruptives into the Holocene (Gamble et al. 2003; Price et al. 2012; Conway et al. 2016, 2018). This sequence matches the pattern at Tongariro wherein ⁸⁷Sr/⁸⁶Sr ratios increase from ~ 0.7047 in ~ 56 ka Te Tatau Member eruptives to ~ 0.7057 in ~ 36–23 ka Mokomoko Formation eruptives (North Crater and Blue Lake vents), notwithstanding the high-MgO 40–30 ka Pukeonake Formation satellite cone (⁸⁷Sr/⁸⁶Sr ~ 0.7047). Subsequently Tongariro ⁸⁷Sr/⁸⁶Sr ratios decrease to ~ 0.7051 in 11.0–1.8 ka Te Ahititi Member eruptives (Red Crater Formation), not including the ~ 17.5 ka Makahikatoa Formation flank vent (⁸⁷Sr/⁸⁶Sr ~ 0.7044: Hobden 1997). Late Holocene-aged, small-volume eruptives from the Ngāuruhoe vent at Tongariro show large ⁸⁷Sr/⁸⁶Sr diversity and are discussed above (Hobden et al. 1999; 2002). Similar compositional diversity is also seen in small-volume eruptives at Ruapehu in the late Holocene (Gamble et al. 1999; Nakagawa et al. 1999).

Time vs. ¹⁴³Nd/¹⁴⁴Nd trends at Tongariro and Ruapehu display simultaneous excursions to more crustal values for the ~ 230–80 ka and 50 ka to Holocene periods (Fig. 13d). Tongariro ¹⁴³Nd/¹⁴⁴Nd ratios decrease from ~ 0.51285 to ~ 0.51265 between ~ 230 and 100 ka which is matched by a similar decrease to more crustal ¹⁴³Nd/¹⁴⁴Nd ratios at Ruapehu from ~ 0.51290 in the 200–150 ka Te Herenga Formation to ~ 0.51270–0.51280 in the 166–80 ka Wahianoa Formation. Ruapehu Mangawhero Formation eruptives have initially high ¹⁴³Nd/¹⁴⁴Nd ratios of ~ 0.51268–0.51275 at ~ 45–40 ka that become more crust-like (~ 0.51264–0.51273) at ~ 25–15 ka, closely overlapping with Tongariro ¹⁴³Nd/¹⁴⁴Nd ratios for the same period. Subsequently, Tongariro and Ruapehu ¹⁴³Nd/¹⁴⁴Nd ratios increase to more mantle-like values into the Holocene, clustering around ~ 0.5127–0.5128.

Together, the extensive geochemical and geochronological datasets that cover most of Tongariro’s and Ruapehu’s edifices show broadly contemporaneous variation in MgO, Rb/Sr, ⁸⁷Sr/⁸⁶Sr and ¹⁴³Nd/¹⁴⁴Nd over their long-term eruptive records (Fig. 13). Given that the magmatic roots at Tongariro and Ruapehu are physically disconnected (Leonard et al. 2021), the contemporaneous compositional cycles at the two volcanoes suggest that the timings of mafic replenishment events and intervals of protracted magma residence in the crust (which appear to be decoupled at Tongariro: see Fig. 12 and associated discussion) are both controlled by a shared external process, most likely to be tectonic, that can cause shifts in crustal stress states.

At one extreme, synchronised eruptive activity in the ~ 11 ka Pahoka–Mangamate eruptions involved several explosive eruptions from compositionally distinct magmas at Tongariro and Ruapehu and their satellite vent systems, where the clustered timing of the eruptions has been attributed to tectonic processes (Nairn et al. 1998; Nakagawa et al. 1998; Heinrich et al. 2020). This inference has been recently disputed, however, because tephras from these eruptions contain crystals with similar reverse zoning patterns, amongst other compositional similarities (Heinrich et al. 2022). These features were interpreted by Heinrich et al. (2022) as showing that the entire Tongariro magmatic system was interconnected and partially homogenised. However, our view is that these similarities in reverse zoning patterns could also be achieved by contemporaneous mafic replenishment without the requirement for lateral interconnectivity between shallow crustal magmas. Such contemporaneous mafic replenishment into compositionally distinctive but isolated magma reservoirs at Tongariro has been reported for the Te Maari, Red Crater and Te Pupu (Ngāuruhoe) formations, without lateral connectivity between their shallow crustal magma bodies (Shane et al. 2017; Pure et al. 2020).

Contemporaneous magmatic events at Tongariro, Ruapehu and Taupō volcanoes

Evidence for tectonically synchronised volcanic activity between Taupō volcano and southern TVZ andesite volcanoes was first proposed by Kohn and Topping (1978). Based largely on interfingered tephra deposits, they reported elevated transition metal concentrations in titanomagnetite crystals that distinguished the voluminous ~ 12–11 ka tephras erupted at Tongariro and Taupō from subsequent eruptives from both volcanoes. These contemporaneous shifts in tephra compositions suggested that magmatic variations between separate volcanoes (and unconnected magma systems) could be synchronised by an external, common influence. Subsequent studies by Hodder (1983) and Nairn et al. (1998) have further supported this hypothesis. Here, we propose that contemporaneous fluctuations in magmatic activity at Tongariro, Ruapehu and Taupō are a long-term feature over their > 200 kyr histories.

In Fig. 13, the time–composition data from Tongariro and Ruapehu are compared with zircon crystallisation model ages (obtained by U–Th disequilibrium methods) from rhyolitic eruptives at Taupō (Charlier et al. 2005; Wilson and Charlier 2009). At Tongariro, minima in ¹⁴³Nd/¹⁴⁴Nd ratios and maxima in ⁸⁷Sr/⁸⁶Sr ratios show that magmas attained their most crustal isotopic compositions at ~ 100 ka and ~ 30 ka, broadly coinciding with the ~ 95 ka and ~ 35 ka peaks in zircon model ages at Taupō (Fig. 13c–e). Sr and Nd isotope ratios at Ruapehu also reach their most crust-like values in the ~ 40–20 ka interval, hence broadly coinciding with the ~ 35 ka zircon model-age peak at Taupō. This contemporaneity is interpreted to show that magmas erupted from the three volcanoes resided in the crust for prolonged periods where they progressively underwent country-rock assimilation (Tongariro and Ruapehu) and cooling (all three volcanoes) from ~ 230 to ~ 100 ka and from ~ 95 to ~ 30 ka.

The idea that magmas experienced prolonged residence in the crust within the ~ 95–30 ka period is further supported by Ruapehu U–Th isotope data reported by Hughes (1999). The sole Ruapehu sample erupted within the ~ 95–30 ka period that was analysed has a U–Th disequilibrium model age of 81 +13/-12 ka, which is within error of the eruption age of 74 ± 37 ka determined via ⁴⁰Ar/³⁹Ar techniques (Gamble et al. 2003). These overlapping ages imply a relatively short residence time within the crust that is likely to be ~ 10–20 kyr or less, given the U–Th and ⁴⁰Ar/³⁹Ar age uncertainties. Younger Ruapehu eruptives with ⁴⁰Ar/³⁹Ar age determinations between ~ 22 and 0 ka have U–Th disequilibria model ages of 72 +23/-19 ka and 62 +18/-15 ka that imply longer crustal times of ~ 40–50 kyr (Hughes 1999; Gamble et al. 2003; Conway et al. 2016). These results indicate that the magmas feeding these eruptives were formed at a similar time but some erupted ~ 50 kyr apart, provided that the bulk of the crystal population was not inherited and of substantially older age. On this evidence, crustal residence was potentially short (i.e. < 10 kyr) for Ruapehu magmas erupted between ~ 95–30 ka relative to longer crustal residence times (i.e. ~ 40–50 kyr) for younger (< 30 ka) eruptives.

Evidence for influxes of hot, mafic magma into the Taupō system between ~ 85 and 45 ka and after ~ 35 ka, that coincide with inferred mafic replenishment events at Tongariro and Ruapehu (Fig. 13), is provided by zircon crystallisation model ages and zircon saturation modelling. Zircon U–Th model ages from late Pleistocene Taupō eruptives have two peak crystallisation modes at 95–86 ka and 41–34 ka (Charlier et al. 2005; Wilson and Charlier 2009). Additionally, zircon saturation modelling for these Taupō eruptives by Charlier et al. (2005, using the model of Watson and Harrison 1983) shows that temperature was a predominant factor controlling the crystallisation and/or survival of zircons in Taupō magmas. We therefore propose that zircon crystallisation was suppressed in the periods between ~ 85 and 45 ka, and after ~ 35 ka, by temporarily elevated magmatic temperatures (Fig. 13e). The reduction in the number of zircon model ages in these periods, together with the time vs. MgO data, suggests that mafic replenishment events at ~ 88 and ~ 56 ka (Tongariro) and ~ 45 ka (Ruapehu) may have also occurred at Taupō (Fig. 13a, e). Shifts to more crust-like compositions at ~ 30 ka for Tongariro and Ruapehu also broadly coincide with the 41–34 ka zircon age peak at Taupō, and subsequent shifts to less crust-like ⁸⁷Sr/⁸⁶Sr and ¹⁴³Nd/¹⁴⁴Nd ratios at Tongariro and Ruapehu after ~ 30 ka align with a reduction in zircon model ages at Taupō (Fig. 13c–e). In general, the temporal alignment of peak zircon age abundances at Taupō and more crust-like eruptive compositions at Tongariro and Ruapehu, and vice versa, can be explained by tectonically modulated, contemporaneous mafic recharge events at all three volcanoes, with interposed periods of reduced mafic input wherein magmas cool and evolve to more crust-like compositions.

However, the cooler, more crust-like Tongariro and Ruapehu magmas erupted at ~ 40–30 ka temporally coincide with the eruption of mafic magmas from flank vents, which indicates further complexity exists than given in the interpretation above. These mafic flank vent eruptives are the high-MgO 40–30 ka Pukeonake Formation and ~ 17.5 ka Makahikatoa Formation satellite cones on western Tongariro and the 35 ka Ohakune Formation southwest of Ruapehu (see Ohakune Craters in Fig. 1). The ages of these mafic vents bracket and coincide with the 41–34 ka zircon age peak at Taupō and the excursion to more crust-like compositions at Tongariro and Ruapehu (Fig. 13) (Froggatt and Lowe 1990; Conway et al. 2016; Townsend et al. 2017; Pure et al. 2020; Svoboda et al. 2022). It is unclear, however, if the timing of mafic flank vent eruptions should be rapidly followed by reductions in crust-like character of central edifice magma systems (i.e. within ~ 1 kyr). Potentially, mafic flank vent eruptions could be sourced from magmas in the deeper crust that bypassed the predominantly andesitic central edifice magmatic systems at Tongariro and Ruapehu when they ascended (e.g. Price et al. 2012), as is suggested by the volume–composition relationships discussed above. If ascending mafic magmas largely avoid the cooler, more evolved central edifice magmatic systems then there would be reduced heating and mafic replenishment of these central systems. This would effectively preserve the cooler, more crust-like character of magmas in the Tongariro and Ruapehu central edifice systems. Collectively, the zircon age dataset from Taupō (Charlier et al. 2005; Wilson and Charlier 2009) and the time–composition datasets of Tongariro and Ruapehu (Gamble et al. 2003; Price et al. 2012; Conway et al. 2016, 2018; Pure et al. 2020) support a model wherein tectonic processes modulate the timing of certain magma assembly processes at the three volcanoes, subject to minor deviations from the general patterns described above, such as mafic magmas erupting from flank vents instead of perturbing centralised magmatic systems.

A role for the tectonic modulation of magmatic activity in the crust is supported by measurements of crustal stress before, during and after recent volcanic eruptions at Ruapehu (Gerst and Savage 2004). Elevated slip rates on the Rangipo Fault in the late Pleistocene have also been associated with volcanic–tectonic interactions at Ruapehu (Fig. 1: Villamor et al. 2007). Additionally, the action of tectonics on regulating eruption rates and timings have been inferred in many other studies elsewhere in the TVZ, the western United States, the Chilean Andes and in Japan (Lanphere and Sisson 2003; Allan et al. 2012; Geshi and Oikawa 2016; Myers et al. 2016; Rawson et al. 2016). Interrelated behaviour between rifting and magmatism is also known in the central TVZ where rifting is aided by thermally assisted crustal weakening, and vice versa, which can result in positive feedbacks (Rowland et al. 2010; Allan et al. 2012; Ellis et al. 2014). At Tongariro specifically, rifting and tectonic processes have been used to explain the distribution of vents at Tongariro and the timing of eruptions (Gómez-Vasconcelos et al. 2017; Heinrich et al. 2020; Pure et al. 2020). On these grounds, we propose that in the central and southern TVZ, the duration of magma residence in the crust and the timing of mafic magma replenishment events are likely to be regulated by tectonic processes, such as episodic variations in crustal stress regimes. On current data (Fig. 13), such processes appear to be capable of synchronising the most volumetrically significant magma-forming and magma evolution events at adjacent arc volcanoes.

Conclusions

Chemical and isotopic data for Tongariro whole-rock and groundmass samples are reported for ~ 220 new samples that add to an existing database of ~ 370 samples (Cole 1979; Hobden 1997). We considered these data both with and independently of their stratigraphic context, which is provided by the high-resolution eruptive stratigraphy of Pure et al. (2020), and draw the following conclusions.

(1)
Tongariro magmas were produced when mantle-derived basalts intruded the crust, of which the uppermost ~ 15 km (at least) comprises (meta)sedimentary basement rocks of the Jurassic Kaweka and Waipapa terranes. These basaltic magmas progressively evolved to basaltic-andesites through to dacites via AFC and associated mixing and mingling processes, whilst partially retaining their crystal cargoes as they evolved (i.e. partial in situ crystallisation differentiation). From the most mafic end-member composition observed in the edifice record (the Te Rongo basaltic-andesite, a.k.a. the Red Crater ‘basalt’: Gamble et al. 1993), magmas crystallised and fractionated olivine (~ 4%), clinopyroxene (~ 32%), orthopyroxene (~ 32%) and plagioclase (~ 32%), as indicated by SiO₂, MgO, CaO/Al₂O₃, P₂O₅, Ni, Sc, Rb/Sr, Eu/Eu*, Y and Sr arrays (Fig. 10). Petrographic observations of ~ 700 whole-rock samples (e.g. Figure 3), and variations of Rb, Ba and Zr concentrations with indices of fractionation, indicate that no zircon, alkali feldspar or biotite crystallised in Tongariro magmas (Fig. 5).
(2)
AFC modelling results, and highly approximate δ¹⁸O mass balance estimates, indicate that typical Tongariro magmas contain ~ 20% of assimilated basement metasediments, derived in subequal proportions from the Kaweka and Waipapa terranes. Assimilated materials were variably melted and occur as both restitic xenoliths (including xenocrysts from disaggregated xenoliths) and incorporated melts. AFC modelling indicates that Tongariro magmas contained melts ranging from intermediate to silicic compositions, which evolved with r-values of 0.1–1 (Fig. 11).
(3)
A large proportion of Tongariro’s compositional diversity in major oxides, trace elements and Sr–Nd–Pb isotopes was acquired partly or mostly in the upper 15 km of the 30–40 km thick crust, via AFC processes.
(4)
When ordered sequentially, both individual samples with ⁴⁰Ar/³⁹Ar age determinations (whole-rock and groundmass samples) and entire stratigraphic units show systematic time-compositional variation on 10–130 kyr cycles. Following interpretations by Pure et al. (2020), increases in MgO at ~ 230, ~ 151, ~ 88, ~ 56 ka and in the late Holocene, and high-MgO flank vents that erupted at ~ 160, ~ 117, ~ 35 and ~ 17.5 ka, represent mafic replenishment events (Fig. 12). These shifts in MgO concentrations are mirrored by decreases in incompatible element markers (e.g. K₂O, Rb/Sr). Other time–composition cycles in Sm/Nd, ⁸⁷Sr/⁸⁶Sr, ¹⁴³Nd/¹⁴⁴Nd and ²⁰⁶Pb/²⁰⁴Pb ratios, which are decoupled from MgO cycles, indicate periods of prolonged crustal residence and contamination from ~ 230 to ~ 100 ka and ~ 95 to ~ 30 ka.
(5)
The range of compositions in eruptives from ~ 350 ka to the Holocene is comparable to the range of compositions from the late Holocene alone, such as for ⁸⁷Sr/⁸⁶Sr (but not for all major oxides or trace elements). However, the volumes represented by the late Holocene eruptive record are generally far smaller (typically < 0.01 km³ per lava flow relative to ≥ 1 km³ for units in the longer-term eruptive record). In contrast to the longer-term record, the late Holocene record is also typified by relatively unsystematic time–composition trends (Hobden et al. 1999, 2002). On this basis, we propose that the smaller-volume eruptives are less likely to be representative of the degree of organisation in their source volcano’s magmatic system.
(6)
There is a repeated temporal alignment of compositional cycles and evidence for contemporaneous magmatic cyclicity and/or eruptive activity at Tongariro, Ruapehu and Taupō (Fig. 13). This alignment suggests that an external process (i.e. tectonics) has regulated the assembly and eruption of the most significant and voluminous magma bodies at each of these volcanoes since at least 200 ka. Cycles of prolonged crustal residence and assimilation in magmas at these volcanoes occurred from ~ 230 to ~ 100 ka and ~ 95 to 30 ka, which are partially decoupled from cyclic mafic replenishment events (see point 4, above).

Data availability

The authors confirm that the data supporting the findings of this study are available within the article [and/or] its supplementary materials.

References

Allan ASR, Wilson CJN, Millet M-A, Wysoczanski RJ (2012) The invisible hand: tectonic triggering and modulation of a rhyolitic supereruption. Geology 40:563–566. https://doi.org/10.1130/G32969.1
Article Google Scholar
Annen C, Blundy JD, Sparks RSJ (2006) The genesis of intermediate and silicic magmas in deep crustal hot zones. J Petrol 47:505–539. https://doi.org/10.1093/petrology/egi084
Article Google Scholar
Annen C, Blundy JD, Leuthold J, Sparks RSJ (2015) Construction and evolution of igneous bodies: towards an integrated perspective of crustal magmatism. Lithos 230:206–221. https://doi.org/10.1016/j.lithos.2015.05.008
Article Google Scholar
Arpa MC, Zellmer GF, Christenson B, Lube G, Shellnutt G (2017) Variable magma reservoir depths for Tongariro Volcanic Complex eruptive deposits from 10,000 years to present. Bull Volcanol 79:56. https://doi.org/10.1007/s00445-017-1137-5
Article Google Scholar
Bacon CR, Druitt TH (1988) Compositional evolution of the zoned calcalkaline magma chamber of Mount Mazama, Crater Lake, Oregon. Contrib Mineral Petrol 98:224–256. https://doi.org/10.1007/BF00402114
Article Google Scholar
Bacon CR, Bruggman PE, Christiansen RL, Clynne MA, Donnelly-Nolan JM, Hildreth W (1997) Primitive magmas at five Cascade volcanic fields: melts from hot, heterogeneous sub-arc mantle. Can Mineral 35:397–423
Google Scholar
Barker SJ, Wilson CJN, Allan ASR, Schipper CI (2015) Fine-scale temporal recovery, reconstruction and evolution of a post-supereruption magmatic system. Contrib Mineral Petrol 170:5. https://doi.org/10.1007/s00410-015-1155-2
Article Google Scholar
Behr Y, Townend J, Bannister S, Savage MK (2011) Crustal shear wave tomography of the Taupo Volcanic Zone, New Zealand, via ambient noise correlation between multiple three-component networks. Geochem Geophys Geosyst 13:Q03015. https://doi.org/10.1029/2010GC003385
Article Google Scholar
Beier C, Haase KM, Brandl PA, Krumm SH (2017) Primitive andesites from the Taupo Volcanic Zone formed by magma mixing. Contrib Mineral Petrol 172:33. https://doi.org/10.1007/s00410-017-1354-0
Article Google Scholar
Bergal-Kuvikas O, Nakagawa M, Kuritani T, Muravyev Y, Malik N, Klimenko E, Amma-Miyasaka M, Matsumoto A, Shimada S (2017) A petrological and geochemical study on time-series samples from Klyuchevskoy volcano. Kamchatka Arc Contrib Mineral Petrol 172:35. https://doi.org/10.1007/s00410-017-1347-z
Article Google Scholar
Blundy J (2022) Chemical differentiation by mineralogical buffering in crustal hot zones. J Petrol 63:egac054. https://doi.org/10.1093/petrology/egac054
Article Google Scholar
Bohrson W, Spera F (2001) Energy-constrained open-system magmatic processes I: general model and energy-constrained assimilation and fractional crystallization (EC-AFC) formulation. J Petrol 42:999–1018. https://doi.org/10.1093/petrology/42.5.999
Article Google Scholar
Calvert AT, Fierstein J, Hildreth W (2018) Eruptive history of Middle Sister, Oregon Cascades, USA—Product of a late Pleistocene eruptive episode. Geosphere 14:2118–2139. https://doi.org/10.1130/GES01638.1
Article Google Scholar
Charlier BLA, Wilson CJN, Lowenstern JB, Blake S, van Calsteren PW, Davidson JP (2005) Magma generation at a large, hyperactive silicic volcano (Taupo, New Zealand) revealed by U-Th and U-Pb systematics in zircons. J Petrol 46:3–32. https://doi.org/10.1093/petrology/egh060
Article Google Scholar
Charlier BLA, Ginibre C, Morgan D, Nowell GM, Pearson DG, Davidson JP, Ottley CJ (2006) Methods for the microsampling and high-precision analysis of strontium and rubidium isotopes at single crystal scale for petrological and geochronological applications. Chem Geol 232:114–133. https://doi.org/10.1016/j.chemgeo.2006.02.015
Article Google Scholar
Claiborne LL, Miller CF, Flanagan DM, Clynne MA, Wooden JL (2010) Zircon reveals protracted magma storage and recycling beneath Mount St. Helens Geology 38:1011–1014. https://doi.org/10.1130/G31285.1
Article Google Scholar
Cocker K, Shane P, Stirling CH, Reid MR (2021) Constraints on assembly of Tongariro and Ruapehu andesite magmas based on Sr-isotope compositions of plagioclase and groundmass. Lithos 400:106406. https://doi.org/10.1016/j.lithos.2021.106406
Article Google Scholar
Cole JW (1978) Andesites of the Tongariro volcanic centre, North Island, New Zealand. J Volcanol Geotherm Res 3:121–153. https://doi.org/10.1016/0377-0273(78)90007-0
Article Google Scholar
Cole JW (1979) Chemical analyses of lavas and ignimbrites of the Taupo Volcanic Zone. Victoria University of Wellington, Publication, Department of Geology p 13
Google Scholar
Cole RP, White JDL, Conway CE, Leonard GS, Townsend DB, Pure LR (2018) The glaciovolcanic evolution of an andesitic edifice, South Crater, Tongariro volcano, New Zealand. J Volcanol Geotherm Res 352:55–77. https://doi.org/10.1016/j.jvolgeores.2017.12.003
Article Google Scholar
Conway CE, Townsend DB, Leonard GS, Wilson CJN, Calvert AT, Gamble JA (2015) Lava-ice interaction on a large composite volcano: a case study from Ruapehu. New Zealand Bull Volcanol 77:21. https://doi.org/10.1007/s00445-015-0906-2
Article Google Scholar
Conway CE, Leonard GS, Townsend DB, Calvert AT, Wilson CJN, Gamble JA, Eaves SR (2016) A high-resolution ⁴⁰Ar/³⁹Ar lava chronology and edifice construction history for Ruapehu volcano, New Zealand. J Volcanol Geotherm Res 327:152–179. https://doi.org/10.1016/j.jvolgeores.2016.07.006
Article Google Scholar
Conway CE, Gamble JA, Wilson CJN, Leonard GS, Townsend DB, Calvert AT (2018) New petrological, geochemical and geochronological perspectives on andesite-dacite magma genesis at Ruapehu volcano, New Zealand. Am Mineral 103:565–581. https://doi.org/10.2138/am-2018-6199
Article Google Scholar
Conway CE, Chamberlain KJ, Harigane Y, Morgan DJ, Wilson CJN (2020) Rapid assembly of high-Mg andesites and dacites by magma mixing at a continental arc stratovolcano. Geology 48:1033–1037. https://doi.org/10.1130/G47614.1
Article Google Scholar
Davidson J, de Silva S (2000) Composite volcanoes. In: Sigurdsson H et al (eds) Encyclopedia of Volcanoes. Academic Press, San Diego, pp 663–681
Google Scholar
Davidson JP, Morgan DJ, Charlier BLA, Harlou R, Hora JM (2007) Microsampling and isotopic analysis of igneous rocks: implications for the study of magmatic systems. Annu Rev Earth Planet Sci 35:273–311. https://doi.org/10.1146/annurev.earth.35.031306.140211
Article Google Scholar
DePaolo DJ (1981) Trace element and isotopic effects of combined wallrock assimilation and fractional crystallization. Earth Planet Sci Lett 53:189–202. https://doi.org/10.1016/0012-821X(81)90153-9
Article Google Scholar
Donoghue SL, Gamble JA, Palmer A, Stewart RB (1995) Magma mingling in Pourahu Member andesitic pyroclastic flows, Ruapehu Volcano, New Zealand. J Volcanol Geotherm Res 68:177–192. https://doi.org/10.1016/0377-0273(95)00012-J
Article Google Scholar
Dungan MA, Wulff AA, Thompson RA (2001) Eruptive stratigraphy of the Tatara-San Pedro Complex, 36ºS, Southern Volcanic Zone, Chilean Andes: reconstruction method and implications for magma evolution at long-lived arc volcanic centers. J Petrol 42:555–626. https://doi.org/10.1093/petrology/42.3.555
Article Google Scholar
Eaves SR, Mackintosh AN, Winckler G, Schaefer JM, Alloway BV, Townsend DB (2016) A cosmogenic ³He chronology of late Quaternary glacier fluctuations in North Island, New Zealand (39° S). Quat Sci Rev 132:40–56. https://doi.org/10.1016/j.quascirev.2015.11.004
Article Google Scholar
Eichelberger JC, Izbekov PE, Browne BL (2006) Bulk chemical trends at arc volcanoes are not liquid lines of descent. Lithos 87:135–154. https://doi.org/10.1016/j.lithos.2005.05.006
Article Google Scholar
Ellis S, Heise W, Kissling W, Villamor P, Schreurs G (2014) The effect of crustal melt on rift dynamics: case study of the Taupo Volcanic Zone. NZ J Geol Geophys 57:453–458. https://doi.org/10.1080/00288306.2014.972961
Article Google Scholar
Ersoy Y, Helvacı C (2010) FC–AFC–FCA and mixing modeler: a Microsoft® Excel^© spreadsheet program for modeling geochemical differentiation of magma by crystal fractionation, crustal assimilation and mixing. Comp Geosci 36:383–390. https://doi.org/10.1016/j.cageo.2009.06.007
Article Google Scholar
Froggatt PC, Lowe DJ (1990) A review of late Quaternary silicic and some other tephra formations from New Zealand: their stratigraphy, nomenclature, distribution, volume, and age. NZ J Geol Geophys 33:89–109. https://doi.org/10.1080/00288306.1990.10427576
Article Google Scholar
Gamble JA, Smith IEM, Graham IJ, Kokelaar BP, Cole JW, Houghton BF, Wilson CJN (1990) The petrology, phase relations and tectonic setting of basalts from the Taupo Volcanic Zone, New Zealand and the Kermadec island arc - Havre Trough, SW Pacific. J Volcanol Geotherm Res 43:253–270. https://doi.org/10.1016/0377-0273(90)90055-K
Article Google Scholar
Gamble JA, Smith IEM, McCulloch MT, Graham IJ, Kokelaar BP (1993) The geochemistry and petrogenesis of basalts from the Taupo Volcanic Zone and Kermadec Island Arc, S.W Pacific. J Volcanol Geotherm Res 54:265–290. https://doi.org/10.1016/0377-0273(93)90067-2
Article Google Scholar
Gamble JA, Woodhead JD, Wright IC, Smith IEM (1996) Basalt and sediment geochemistry and magma petrogenesis in a transect from oceanic island arc to rifted continental margin arc: the Kermadec-Hikurangi margin, SW Pacific. J Petrol 37:1523–1546. https://doi.org/10.1093/petrology/37.6.1523
Article Google Scholar
Gamble JA, Wood CP, Price RC, Smith IEM, Stewart RB, Waight T (1999) A fifty year perspective of magmatic evolution on Ruapehu Volcano, New Zealand: verification of open system behaviour in an arc volcano. Earth Planet Sci Lett 170:301–314. https://doi.org/10.1016/S0012-821X(99)00106-5
Article Google Scholar
Gamble JA, Price RC, Smith IEM, McIntosh WC, Dunbar NW (2003) ⁴⁰Ar/³⁹Ar geochronology of magmatic activity, magma flux and hazards at Ruapehu volcano, Taupo Volcanic Zone, New Zealand. J Volcanol Geotherm Res 120:271–287. https://doi.org/10.1016/S0377-0273(02)00407-9
Article Google Scholar
Gerst A, Savage MK (2004) Seismic anisotropy beneath Ruapehu volcano: a possible eruption forecasting tool. Science 306:1543–1547. https://doi.org/10.1126/science.1103445
Article Google Scholar
Geshi N, Oikawa T (2016) Orientation of the eruption fissures controlled by a shallow magma chamber in Miyakejima. Front Earth Sci 4:99. https://doi.org/10.3389/feart.2016.00099
Article Google Scholar
Gill J (1981) Orogenic andesites and plate tectonics. Springer, New York, p 390
Book Google Scholar
Gómez-Vasconcelos MG, Villamor P, Cronin S, Procter J, Palmer A, Townsend D, Leonard G (2017) Crustal extension in the Tongariro graben, New Zealand: insights into volcano-tectonic interactions and active deformation in a young continental rift. Geol Soc Am Bull 129:1085–1099. https://doi.org/10.1130/B31657.1
Article Google Scholar
Graham IJ (1987) Petrography and origin of metasedimentary xenoliths in lavas from Tongariro Volcanic Centre. NZ J Geol Geophys 30:139–157. https://doi.org/10.1080/00288306.1987.10422179
Article Google Scholar
Graham IJ, Hackett WR (1987) Petrology of calc-alkaline lavas from Ruapehu volcano and related vents, Taupo Volcanic Zone, New Zealand. J Petrol 28:531–567. https://doi.org/10.1093/petrology/28.3.531
Article Google Scholar
Graham IJ, Blattner P, McCulloch MT (1990) Meta-igneous granulite xenoliths from Mount Ruapehu, New Zealand: fragments of altered oceanic crust? Contrib Mineral Petrol 105:650–661. https://doi.org/10.1007/BF00306531
Article Google Scholar
Gruender K, Stewart RB, Foley S (2010) Xenoliths from the sub-volcanic lithosphere of Mt. Taranaki, New Zealand. J Volcanol Geotherm Res 190:192–202. https://doi.org/10.1016/j.jvolgeores.2009.09.014
Article Google Scholar
Harrison A, White RS (2006) Lithospheric structure of an active backarc basin: the Taupo Volcanic Zone, New Zealand. Geophys J Int 167:968–990. https://doi.org/10.1111/j.1365-246X.2006.03166.x
Article Google Scholar
Heinrich M, Cronin SJ, Pardo N (2020) Understanding multi-vent Plinian eruptions at Mt. Tongariro volcanic complex, New Zealand. Bull Volcanol 82:30. https://doi.org/10.1007/s00445-020-1369-7
Article Google Scholar
Heinrich M, Cronin SJ, Hansteen TH (2022) Magmatic drivers of a 200-year-long high-magnitude explosive flare-up from Mt. Tongariro, New Zealand. J Volcanol Geotherm Res 427:107569. https://doi.org/10.1016/j.jvolgeores.2022.107569
Article Google Scholar
Hildreth W (2007) Quaternary magmatism in the Cascades: geologic perspectives. US Geol Surv Prof Pap 1744:1–125. https://doi.org/10.3133/pp1744
Article Google Scholar
Hildreth W, Fierstein J (2012) Eruptive history of Mount Katmai, Alaska. Geosphere 8:1527–1567. https://doi.org/10.1130/GES00817.1
Article Google Scholar
Hildreth W, Lanphere MA (1994) Potassium-argon geochronology of a basalt-andesite-dacite arc system: the Mount Adams volcanic field, Cascade Range of southern Washington. Geol Soc Am Bull 106:1413–1429. https://doi.org/10.1130/0016-7606(1994)106%3c1413:PAGOAB%3e2.3.CO;2
Article Google Scholar
Hildreth W, Moorbath S (1988) Crustal contributions to arc magmatism in the Andes of central Chile. Contrib Mineral Petrol 98:455–489. https://doi.org/10.1007/BF00372365
Article Google Scholar
Hill GJ, Bibby HM, Ogawa Y, Wallin EL, Bennie SL, Caldwell TG, Keys H, Bertrand EA, Heise W (2015) Structure of the Tongariro volcanic system: insights from magnetotelluric imaging. Earth Planet Sci Lett 432:115–125. https://doi.org/10.1016/j.epsl.2015.10.003
Article Google Scholar
Hobden BJ (1997) Modelling magmatic trends in space and time: eruptive and magmatic history of the Tongariro volcanic complex. PhD thesis, University of Canterbury, Christchurch, p 540
Google Scholar
Hobden BJ, Houghton BF, Lanphere MA, Nairn IA (1996) Growth of the Tongariro volcanic complex: new evidence from K-Ar age determinations. NZ J Geol Geophys 39:151–154. https://doi.org/10.1080/00288306.1996.9514701
Article Google Scholar
Hobden BJ, Houghton BF, Davidson JP, Weaver SD (1999) Small and short-lived magma batches at composite volcanoes: time windows at Tongariro volcano, New Zealand. J Geol Soc Lond 156:865–868. https://doi.org/10.1144/gsjgs.156.5.0865
Article Google Scholar
Hobden BJ, Houghton BF, Nairn IA (2002) Growth of a young, frequently active composite cone: Ngauruhoe volcano, New Zealand. Bull Volcanol 64:392–409. https://doi.org/10.1007/s00445-002-0216-3
Article Google Scholar
Hodder APW (1983) Diapiric replenishment of magma chambers and triggering of volcanism in the Taupo Volcanic Zone, New Zealand. Chem Geol 38:275–285. https://doi.org/10.1016/0009-2541(83)90059-1
Article Google Scholar
Hughes RD (1999) The timescales of andesite generation at Mount Ruapehu. The Open University, Milton Keynes, p 317
Google Scholar
Humphreys MC, Blundy JD, Sparks RSJ (2006) Magma evolution and open-system processes at Shiveluch volcano: insights from phenocryst zoning. J Petrol 47:2303–2334. https://doi.org/10.1093/petrology/egl045
Article Google Scholar
Izbekov PE, Eichelberger JC, Ivanov BV (2004) The 1996 eruption of Karymsky volcano, Kamchatka: historical record of basaltic replenishment of an andesite reservoir. J Petrol 45:2325–2345. https://doi.org/10.1093/petrology/egh059
Article Google Scholar
Jochum KP, Weis U, Schwagger B, Stoll B, Wilson SA, Haug GH, Andreae MO, Enzweiler J (2016) Reference values following ISO guidelines for frequently requested rock reference materials. Geostand Geoanal Res 40:333–350. https://doi.org/10.1111/j.1751-908X.2015.00392.x
Article Google Scholar
Kilgour G, Blundy J, Cashman K, Mader HM (2013) Small volume andesite magmas and melt–mush interactions at Ruapehu, New Zealand: evidence from melt inclusions. Contrib Mineral Petrol 166:371–392. https://doi.org/10.1007/s00410-013-0880-7
Article Google Scholar
Kohn BP, Topping WW (1978) Time-space relationships between late Quaternary rhyolitic and andesitic volcanism in the southern Taupo Volcanic Zone, New Zealand. Geol Soc Am Bull 89:1265–1271. https://doi.org/10.1130/0016-7606(1978)89%3c1265:TRBLQR%3e2.0.CO;2
Article Google Scholar
Langridge RM, Ries WF, Litchfield NJ, Villamor P, Van Dissen RJ, Barrell DJ, Rattenbury MS, Heron DW, Haubrock S, Townsend DB, Lee JM (2016) The New Zealand active faults database. NZ J Geol Geophys 59:86–96. https://doi.org/10.1080/00288306.2015.1112818
Article Google Scholar
Lanphere MA, Sisson TW (2003) Episodic growth at Mt. Rainier, Washington: a product of tectonic throttling? Geol Soc Am Annual Meeting 35:644
Google Scholar
Lara LE, Clavero J (2004) Villarrica volcano (39.5°S), Southern Chile, Andes. Servicio Nacional De Geología y Minería Boletín 61:1–73
Google Scholar
Lee CTA, Bachmann O (2014) How important is the role of crystal fractionation in making intermediate magmas? Insights from Zr and P systematics. Earth Planet Sci Lett 393:266–274. https://doi.org/10.1016/j.epsl.2014.02.044
Article Google Scholar
Leonard GS, Cole RP, Christenson B, Conway CE, Cronin SJ, Gamble JA, Hurst T, Kennedy BM, Miller CA, Pure LR, Procter JN, Townsend DM, White JDL, Wilson CJN (2021) Ruapehu and Tongariro stratovolcanoes: a review of current understanding. NZ J Geol Geophys 64:389–420. https://doi.org/10.1080/00288306.2021.1909080
Article Google Scholar
Ludwig KR (2002) Isoplot/Ex version 2.49: a geochronological toolkit for microsoft excel. Berkeley Chronology Center, Berkeley
Google Scholar
Lugmair GW, Carlson RW (1978) The Sm-Nd history of KREEP. Proc Lunar Planet Sci Conf 9th:689–704
McArthur JL, Shepherd MJ (1990) Late Quaternary glaciation of Mount Ruapehu, North Island, New Zealand. J R Soc NZ 20:287–296. https://doi.org/10.1080/03036758.1990.10416823
Article Google Scholar
McCulloch MT, Kyser TK, Woodhead JD, Kinsley L (1994) Pb-Sr-Nd-O isotopic constraints on the origin of rhyolites from the Taupo Volcanic Zone of New Zealand: evidence for assimilation followed by fractionation from basalt. Contrib Mineral Petrol 115:303–312. https://doi.org/10.1007/BF00310769
Article Google Scholar
McDonough WF, Sun SS (1995) The composition of the Earth. Chem Geol 120:223–253. https://doi.org/10.1016/0009-2541(94)00140-4
Article Google Scholar
Melekhova E, Annen C, Blundy J (2013) Compositional gaps in igneous rock suites controlled by magma system heat and water content. Nat Geosci 6:385–390. https://doi.org/10.1038/ngeo1781
Article Google Scholar
Miller CA, Kang SG, Fournier D, Hill G (2018) Distribution of vapor and condensate in a hydrothermal system: insights from self-potential inversion at Mount Tongariro, New Zealand. Geophys Res Lett 45:8190–8198. https://doi.org/10.1029/2018GL078780
Article Google Scholar
Moebis A, Cronin SJ, Neall VE, Smith IE (2011) Unravelling a complex volcanic history from fine-grained, intricate Holocene ash sequences at the Tongariro Volcanic Centre, New Zealand. Quat Int 246:352–363. https://doi.org/10.1016/j.quaint.2011.05.035
Article Google Scholar
Myers ML, Wallace PJ, Wilson CJN, Morter BK, Swallow EJ (2016) Prolonged ascent and episodic venting of discrete magma batches at the onset of the Huckleberry Ridge supereruption, Yellowstone. Earth Planet Sci Lett 451:285–297. https://doi.org/10.1016/j.epsl.2016.07.023
Article Google Scholar
Nairn IA, Kobayashi T, Nakagawa M (1998) The ~10 ka multiple vent pyroclastic eruption sequence at Tongariro Volcanic Centre, Taupo Volcanic Zone, New Zealand: Part 1. Eruptive processes during regional extension. J Volcanol Geotherm Res 86:19–44. https://doi.org/10.1016/S0377-0273(98)00085-7
Article Google Scholar
Nakagawa M, Nairn IA, Kobayashi T (1998) The ~10 ka multiple vent pyroclastic eruption sequence at Tongariro Volcanic Centre, Taupo Volcanic Zone, New Zealand: Part 2. Petrological insights into magma storage and transport during regional extension. J Volcanol Geotherm Res 86:45–65. https://doi.org/10.1016/S0377-0273(98)00086-9
Article Google Scholar
Nakagawa M, Wada K, Thordarson T, Wood CP, Gamble JA (1999) Petrologic investigations of the 1995 and 1996 eruptions of Ruapehu volcano, New Zealand: formation of discrete and small magma pockets and their intermittent discharge. Bull Volcanol 61:15–31. https://doi.org/10.1007/s004450050259
Article Google Scholar
Nakagawa M, Wada K, Wood CP (2002) Mixed magmas, mush chambers and eruption triggers: evidence from zoned clinopyroxene phenocrysts in andesitic scoria from the 1995 eruptions of Ruapehu volcano, New Zealand. J Petrol 43:2279–2303. https://doi.org/10.1093/petrology/43.12.2279
Article Google Scholar
Nicol A, Mazengarb C, Chanier F, Rait G, Uruski C, Wallace L (2007) Tectonic evolution of the active Hikurangi subduction margin, New Zealand, since the Oligocene. Tectonics 26:TC4002. https://doi.org/10.1029/2006TC002090
Article Google Scholar
Pin C, Gannoun A, Dupont A (2014) Rapid, simultaneous separation of Sr, Pb and Nd by extraction chromatography prior to isotope ratios determination by TIMS and MC-ICPMS. J Anal Atom Spectrom 29:1858–1870. https://doi.org/10.1039/c4ja00169a
Article Google Scholar
Price RC, Gamble JA, Smith IEM, Eggins SM, Wright IC (2005) An integrated model for the temporal evolution of andesites and rhyolites in New Zealand’s North Island. J Volcanol Geotherm Res 140:1–24. https://doi.org/10.1016/j.jvolgeores.2004.07.013
Article Google Scholar
Price RC, Turner S, Cook C, Hobden B, Smith IEM, Gamble JA, Handley H, Maas R, Möbis A (2010) Crustal and mantle influences and U-Th-Ra disequilibrium in andesitic lavas of Ngauruhoe volcano, New Zealand. Chem Geol 277:355–373. https://doi.org/10.1016/j.chemgeo.2010.08.021
Article Google Scholar
Price RC, Gamble JA, Smith IEM, Maas R, Waight T, Stewart RB, Woodhead J (2012) The anatomy of an andesite volcano: a time-stratigraphic study of andesite petrogenesis and crustal evolution at Ruapehu Volcano, New Zealand. J Petrol 53:2139–2189. https://doi.org/10.1093/petrology/egs050
Article Google Scholar
Price RC, Mortimer N, Smith IEM, Maas R (2015) Whole-rock geochemical reference data for Torlesse and Waipapa terranes, North Island, New Zealand. NZ J Geol Geophys 58:213–228. https://doi.org/10.1080/00288306.2015.1026832
Article Google Scholar
Pure LR (2020) The volcanic and magmatic evolution of Tongariro volcano. PhD thesis, Victoria University of Wellington, Wellington, p 434
Google Scholar
Pure LR, Townsend DB, Leonard GS, Wilson CJN, Calvert AT, Cole RP, Conway CE, Gamble JA, Smith TB (2020) A high resolution ⁴⁰Ar/³⁹Ar lava chronology and edifice construction history for Tongariro volcano, New Zealand. J Volcanol Geotherm Res 403:106993. https://doi.org/10.1016/j.jvolgeores.2020.106993
Article Google Scholar
Pure LR, Charlier BLA, Wilson CJN, Gamble JA, Vry JK, Townsend DB, Leonard GS (2021) Chemical and isotopic changes induced by pyrometamorphism in metasedimentary xenoliths at Tongariro volcano, New Zealand. Lithos 400–401:106404. https://doi.org/10.1016/j.lithos.2021.106404
Article Google Scholar
Raczek I, Jochum KP, Hofmann AW (2003) Neodymium and strontium isotope data for USGS reference materials BCR-1, BCR-2, BHVO-1, BHVO-2, AGV-1, AGV-2, GSP-1, GSP-2 and eight MPI-DING reference glasses. Geostand Newsl 27:173–179. https://doi.org/10.1111/j.1751-908X.2003.tb00644.x
Article Google Scholar
Rawson H, Pyle DM, Mather TA, Smith VC, Fontijn K, Lachowycz SM, Naranjo JA (2016) The magmatic and eruptive response of arc volcanoes to deglaciation: insights from southern Chile. Geology 44:251–254. https://doi.org/10.1130/G37504.1
Article Google Scholar
Reubi O, Blundy J (2009) A dearth of intermediate melts at subduction zone volcanoes and the petrogenesis of arc andesites. Nature 461:1269–1273. https://doi.org/10.1038/nature08510
Article Google Scholar
Rogan W (1996) New insights on magmatic processes from trace element zonation in phenocrysts. The Open University, Milton Keynes, p 252
Google Scholar
Rollinson H (1993) Using geochemical data: evaluation, presentation, interpretation. Longman, Harlow, UK, p 352
Google Scholar
Rowland JV, Wilson CJN, Gravley DM (2010) Spatial and temporal variations in magma-assisted rifting, Taupo Volcanic Zone, New Zealand. J Volcanol Geotherm Res 190:89–108. https://doi.org/10.1016/j.jvolgeores.2009.05.004
Article Google Scholar
Rudnick RL, McDonough WF, McCulloch M, Taylor SR (1986) Lower crustal xenoliths from Queensland, Australia: evidence for deep crustal assimilation and fractionation of continental basalts. Geochim Cosmochim Acta 50:1099–1155. https://doi.org/10.1016/0016-7037(86)90391-1
Article Google Scholar
Schmidt ME, Grunder AL (2009) The evolution of North Sister: a volcano shaped by extension and ice in the central Oregon Cascade Arc. Geol Soc Am Bull 121:643–662. https://doi.org/10.1130/B26442.1
Article Google Scholar
Shane P, Maas R, Lindsay J (2017) History of Red Crater volcano, Tongariro Volcanic Centre (New Zealand): abrupt shift in magmatism following recharge and contrasting evolution between neighboring volcanoes. J Volcanol Geotherm Res 340:1–15. https://doi.org/10.1016/j.jvolgeores.2017.04.008
Article Google Scholar
Shane P, Cocker K, Coote A, Stirling CH, Reid MR (2019) The prevalence of plagioclase antecrysts and xenocrysts in andesite magma, exemplified by lavas of the Tongariro volcanic complex, New Zealand. Contrib Mineral Petrol 174:89. https://doi.org/10.1007/s00410-019-1626-y
Article Google Scholar
Sisson TW, Salters VJM, Larson PB (2014) Petrogenesis of Mount Rainier andesite: magma flux and geologic controls on the contrasting differentiation styles at stratovolcanoes of the southern Washington Cascades. Geol Soc Am Bull 126:122–144. https://doi.org/10.1130/B30852.1
Article Google Scholar
Smith IEM, Worthington TJ, Stewart RB, Price RC, Gamble JA (2003) Felsic volcanism in the Kermadec arc, SW Pacific: crustal recycling in an oceanic setting. Geol Soc Lond Spec Publ 219:99–118. https://doi.org/10.1144/GSL.SP.2003.219.01.05
Article Google Scholar
Stern T, Benson A (2011) Wide-angle seismic imaging beneath an andesitic arc: central North Island, New Zealand. J Geophys Res 116:B09308. https://doi.org/10.1029/2011JB008337
Article Google Scholar
Streck MJ, Leeman WP (2018) Petrology of “Mount Shasta” high-magnesian andesite (HMA): a product of multi-stage crustal assembly. Am Mineral 103:216–240. https://doi.org/10.2138/am-2018-6151
Article Google Scholar
Svoboda C, Rooney TO, Girard G, Deering C (2022) Transcrustal magmatic systems: evidence from andesites of the southern Taupo Volcanic Zone. J Geol Soc Lond 179:jgs2020–204. https://doi.org/10.1144/jgs2020-204
Article Google Scholar
Tanaka T, Togashi S, Kamioka H, Amakawa H, Kagami H, Hamamoto T, Yuhara M, Orihashi Y, Yoneda S, Shimizu H, Kunimaru T, Takahashi K, Yanagi T, Nakano T, Fujimaki H, Shinjo R, Asahara Y, Tanimizu M, Dragusanu C (2000) Jndi-1: a neodymium isotopic reference in consistency with LaJolla neodymium. Chem Geol 168:279–281. https://doi.org/10.1016/S0009-2541(00)00198-4
Article Google Scholar
Thirlwall MF (1991) Long-term reproducibility of multicollector Sr and Nd isotope ratio analysis. Chem Geol 94:85–104. https://doi.org/10.1016/S0009-2541(10)80021-X
Article Google Scholar
Thirlwall MF (2000) Inter-laboratory and other errors in Pb isotope analyses investigated using a ²⁰⁷Pb–²⁰⁴Pb double spike. Chem Geol 163:299–322. https://doi.org/10.1016/S0009-2541(99)00135-7
Article Google Scholar
Todd E, Stracke A, Scherer EE (2015) Effects of simple acid leaching of crushed and powdered geological materials on high-precision Pb isotope analyses. Geochem Geophys Geosyst 16:2276–2302. https://doi.org/10.1002/2015GC005804
Article Google Scholar
Tollan PME, Dale CW, Hermann J, Davidson JP, Arculus RJ (2017) Generation and modification of mantle wedge and lithosphere beneath the West Bismarck island arc: melting, metasomatism and thermal history of peridotite xenoliths from Ritter Island. J Petrol 58:1475–1510. https://doi.org/10.1093/petrology/egx062
Article Google Scholar
Topping WW (1973) Tephrostratigraphy and chronology of late Quaternary eruptives from the Tongariro Volcanic Centre, New Zealand. NZ J Geol Geophys 16:397–423. https://doi.org/10.1080/00288306.1973.10431368
Article Google Scholar
Townsend DB, Leonard GS, Conway CE, Eaves SR, Wilson CJN (2017) Geology of the Tongariro National Park area. Lower Hutt (NZ): GNS Science. 1 sheet + 109 pp, scale 1:60,000. (GNS Science geological map 4)
Villamor P, Berryman KR (2006) Late Quaternary geometry and kinematics of faults at the southern termination of the Taupo Volcanic Zone, New Zealand. NZ J Geol Geophys 49:1–21. https://doi.org/10.1080/00288306.2006.9515144
Article Google Scholar
Villamor P, Van Dissen R, Alloway BV, Palmer AS, Litchfield N (2007) The Rangipo fault, Taupo rift, New Zealand: an example of temporal slip-rate and single-event displacement variability in a volcanic environment. Geol Soc Am Bull 119:529–547. https://doi.org/10.1130/B26000.1
Article Google Scholar
Villamor P, Berryman KR, Ellis SM, Schreurs G, Wallace LM, Leonard GS, Langridge RM, Ries WF (2017) Rapid evolution of subduction-related continental intraarc rifts: the Taupo Rift, New Zealand. Tectonics 36:2250–2272. https://doi.org/10.1002/2017TC004715
Article Google Scholar
Waight TE, Troll VR, Gamble JA, Price RC, Chadwick JP (2017) Hf isotope evidence for variable slab input and crustal addition in basalts and andesites of the Taupo Volcanic Zone, New Zealand. Lithos 284:222–236. https://doi.org/10.1016/j.lithos.2017.04.009
Article Google Scholar
Watson EB, Harrison TM (1983) Zircon saturation revisited: temperature and composition effects in a variety of crustal magma types. Earth Planet Sci Lett 64:295–304. https://doi.org/10.1016/0012-821X(83)90211-X
Article Google Scholar
Wilson CJN, Charlier BLA (2009) Rapid rates of magma generation at contemporaneous magma systems, Taupo volcano, New Zealand: insights from U-Th model-age spectra in zircons. J Petrol 50:875–907. https://doi.org/10.1093/petrology/egp023
Article Google Scholar
Wilson CJN, Houghton BF, McWilliams MO, Lanphere MA, Weaver SD, Briggs RM (1995) Volcanic and structural evolution of Taupo Volcanic Zone, New Zealand: a review. J Volcanol Geotherm Res 68:1–28. https://doi.org/10.1016/0377-0273(95)00006-G
Article Google Scholar

Download references

Acknowledgements

LRP was supported by a PhD Scholarship from Victoria University of Wellington, the ECLIPSE Programme (contract RTVU1704) from the New Zealand Ministry of Business, Innovation and Employment (MBIE), and GNS Science. The Ngati Tūwharetoa Iwi (represented through the Department of Conservation and GNS Science) are thanked for permission to undertake scientific sampling. Bubs Smith is thanked for his support of this work. The Earthquake Commission of New Zealand grant 16/U735 supported analytical costs. GSL and DBT were partly funded through MBIE Core Funding via the Minerals, Urban and Regional Geology and Volcanic Hazards programmes at GNS Science. Andrew Galloway is thanked for assistance and morale in the field. We also thank Anita Grunder, Jim Cole and Kevin Norton for constructive feedback, Tom Sisson and Ben Ellis for formal reviews and Daniela Rubatto for editorial handling.

Funding

Open Access funding enabled and organized by CAUL and its Member Institutions. The Earthquake Commission of New Zealand grant 16/U735 supported analytical costs. Salary for LRP was provided by the sources noted in the acknowledgements above.

Author information

Authors and Affiliations

School of Geography, Environment and Earth Sciences, Victoria University of Wellington, PO Box 600, Wellington, 6140, New Zealand
Leo R. Pure, Colin J. N. Wilson, Bruce L. A. Charlier & John A. Gamble
Environment Approvals Division, Department of Climate Change, Energy, the Environment and Water, GPO Box 3090, Canberra, ACT, 2601, Australia
Leo R. Pure
School of Biological, Earth and Environmental Science, University College Cork, Cork, Ireland
John A. Gamble
Avalon Research Centre, GNS Science, PO Box 30-368, Lower Hutt, 5040, New Zealand
Dougal B. Townsend & Graham S. Leonard

Authors

Leo R. Pure
View author publications
You can also search for this author in PubMed Google Scholar
Colin J. N. Wilson
View author publications
You can also search for this author in PubMed Google Scholar
Bruce L. A. Charlier
View author publications
You can also search for this author in PubMed Google Scholar
John A. Gamble
View author publications
You can also search for this author in PubMed Google Scholar
Dougal B. Townsend
View author publications
You can also search for this author in PubMed Google Scholar
Graham S. Leonard
View author publications
You can also search for this author in PubMed Google Scholar

Contributions

LR Pure: Conceptualisation, Methodology, Formal analysis, Investigation, Resources, Data curation, Funding acquisition, Writing—original draft, Writing—review and editing. CJN Wilson: Conceptualisation, Project administration, Funding acquisition, Supervision, Writing—review and editing. BLA Charlier: Methodology, Resources, Supervision, Writing—review and editing. JA Gamble: Conceptualisation, Investigation, Supervision, Writing—review and editing. DB Townsend: Conceptualisation, Methodology, Resources, Supervision, Writing—review and editing. GS Leonard: Conceptualisation, Supervision, Writing—review and editing.

Corresponding author

Correspondence to Leo R. Pure.

Ethics declarations

Conflict of interest

The authors declare that they have no conflict of interest.

Additional information

Communicated by Daniela Rubatto.

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Below is the link to the electronic supplementary material.

Supplementary file1 (XLSX 63 KB)

Supplementary file2 (DOCX 639 KB)

Supplementary file3 (XLSX 19 KB)

Supplementary file4 (XLSX 149 KB)

Supplementary file5 (XLSX 53 KB)

Supplementary file6 (XLSX 114 KB)

Supplementary file7 (XLSX 129 KB)

Supplementary file8 (XLSX 72 KB)

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/

Reprints and permissions

About this article

Cite this article

Pure, L.R., Wilson, C.J.N., Charlier, B.L.A. et al. The compositional diversity and temporal evolution of an active andesitic arc stratovolcano: Tongariro, Taupō Volcanic Zone, New Zealand. Contrib Mineral Petrol 178, 30 (2023). https://doi.org/10.1007/s00410-023-02004-0

Download citation

Received: 22 October 2022
Accepted: 04 March 2023
Published: 25 April 2023
DOI: https://doi.org/10.1007/s00410-023-02004-0

Keywords

Use our pre-submission checklist

Avoid common mistakes on your manuscript.

The compositional diversity and temporal evolution of an active andesitic arc stratovolcano: Tongariro, Taupō Volcanic Zone, New Zealand

Abstract

Similar content being viewed by others

Introduction

Geological background

Analytical methods

Results

Petrographic characteristics

Geochemistry

Major oxides

Trace elements

Sr–Nd–Pb isotopes

Discussion

Magma assembly processes at Tongariro

Crystallisation differentiation

Assimilation and recharge

Assimilation with fractional crystallisation

Geochemical modelling of magma evolution

Time–composition relationships and the organisation of the magmatic system

Long-term drivers and aspects of magmatic systems at Tongariro and neighbouring volcanoes

Time–composition patterns at Tongariro vs. Ruapehu

Contemporaneous magmatic events at Tongariro, Ruapehu and Taupō volcanoes

Conclusions

Data availability

References

Acknowledgements

Funding

Author information

Authors and Affiliations

Contributions

Corresponding author

Ethics declarations

Conflict of interest

Additional information

Publisher's Note

Supplementary Information

Rights and permissions

About this article

Cite this article

Share this article

Keywords

Search

Navigation