Contemporaneous alkaline and subalkaline intraplate magmatism in the Dunedin Volcanic Group, NZ, caused by mantle heterogeneity

ABSTRACT

Monogenetic volcanism in the Maniototo Basin in the Dunedin Volcanic Group was unusual because the eroded lavas, plugs and dikes comprise spatially and temporally restricted silica-saturated and silica-undersaturated magmas. ⁴⁰Ar/³⁹Ar dating coupled with existing data and field relationships indicate that volcanism was focused at ∼11 Ma. Major and trace element data show the volcanic rocks to have mafic compositions and little evidence for fractionation and/or crustal contamination, but to have been variably altered. The volcanic rocks comprise an alkaline group and an alkaline to subalkaline transitional group, with one intermediate unit. With the exception of one flow, the transitional group has more radiogenic ⁸⁷Sr/⁸⁶Sr and less radiogenic ¹⁴³Nd/¹⁴⁴Nd, ²⁰⁶Pb/²⁰⁴Pb and ²⁰⁸Pb/²⁰⁴Pb than the alkaline group, as well as smaller Ti and K primitive mantle-normalised anomalies. Overall, the different magma compositions cannot be related to varying degrees of melting of equivalent mantle sources. Trace element ratios indicate that residual Ti and K-bearing mantle phases (amphibole, phlogopite) probably played a major role in the source of alkaline group but a smaller role in the transitional group. Therefore, the varying compositions of the Maniototo volcanic rocks relates to melting of a heterogeneous lithospheric mantle, or/and varying degrees of interaction with mantle hydrous metasomes as the magmas ascended.

KEYWORDS:

• Volcanism
• Zealandia; intraplate
• Dunedin Volcanic Group; alkaline; subalkaline

Introduction

The origin of dispersed intraplate volcanism is a contentious topic. Asthenospheric and/or lithospheric mantle involvement is commonly invoked to explain the range of magma compositions found in intraplate volcanic provinces (e.g. Finn et al. 2005; Hoernle et al. 2006; Pilet et al. 2008; Dasgupta et al. 2010; Rooney et al. 2017; Mather et al. 2020; Smith et al. 2021). Classically, such provinces are subdivided into sub-alkaline (silica-saturated) and alkaline (silica-undersaturated) evolution trends (Coombs and Wilkinson 1969; Irvine and Baragar 1971). However, with the collection of more data, it is becoming increasingly recognised that silica-saturated and silica-undersaturated magmas have occurred contemporaneously within single New Zealand magmatic provinces (Cook et al. 2005; Brenna et al. 2012; Pontesilli et al. 2021). These cases require the magmas to be derived from melting of multiple lithologically discrete mantle sources (e.g. McGee et al. 2013) and/or to have been modified during ascent through the lithosphere (e.g. Brenna et al. 2021).

Here, we investigate rocks in the Maniototo Basin in the NW area of the Dunedin Volcanic Group (Figure 1). The intraplate volcanic field represented by the Dunedin Volcanic Group was distributed over ∼7,800 km² with eruptions over a ∼15 Myr period that show no spatial systematic variation (Coombs et al. 1986, 2008; Scott et al. 2020a). Limited data collected as part of a regional review revealed that some volcanic rocks in the Maniototo Basin have sub-alkaline or alkaline compositions and are distinct from all other monogenetic components in this province (Scott et al. 2020a). Here, we provide a comprehensive analysis of the distribution, composition and mantle sources for these significant volcanic rocks through sampling, petrography, whole rock major and trace element data, Sr-Nd-Pb isotopic data and ⁴⁰Ar/³⁹Ar dating.

Geological setting

Intraplate volcanism has been occurring in Zealandia for ∼100 million years (Timm et al. 2010; Hoernle et al. 2006; van der Meer et al. 2017; Mortimer and Scott 2020). This volcanism in NZ is commonly characterised by small monogenetic eruptions and a lack of systemic relationship between volcanic field location and age or composition (Coombs et al. 1986; Cooper 1986; Reay et al. 1991; Baker et al. 1994; Hoernle et al. 2006; Panter et al. 2006; Sprung et al. 2007; Coombs et al. 2008; Timm et al. 2009, 2010; McCoy-West et al. 2010; Timm et al. 2010; van der Meer et al. 2013; Scott et al. 2013; McLeod and White 2018; van der Meer et al. 2016, 2017; Gamble et al. 2018; Scott and Turnbull 2018; Scott et al. 2020a, 2020b; Cooper 2020; Hopkins et al. 2021; Smith and Cronin 2021; Scott et al. 2023).

The Dunedin Volcanic Group in the South Island of New Zealand comprises more than 150 dispersed but eroded eruptive centres across an area of ∼ 7,800 km² (Coombs et al. 1986, 2008; Scott et al. 2020a) (Figure 1). ⁴⁰Ar/³⁹Ar and K-Ar dates indicate that eruptions occurred from 24 to 9 Ma (McDougall and Coombs 1973; Hoernle et al. 2006; Coombs et al. 2008; Timm et al. 2010; Fox et al. 2015; Kaulfuss et al. 2018; Scott et al. 2020a), although we have some reservations about the ages < 11 Ma due to their low radiogenic Ar yields (Scott et al. 2020b). The Dunedin Volcanic Group therefore represents one of the largest and longest-lived episodes of intraplate volcanism in Zealandia. Chemical and isotopic data for the province show the magmas to have been dominantly alkaline in composition (Coombs et al. 1986, 2008; Reay et al. 1991; Scott et al. 2020a), although some mafic magmas appear to have been sub-alkaline (Scott et al. 2020a; Pontesilli et al. 2021, 2022). The Dunedin Volcanic Group comprises rocks of:

1. The well-studied Dunedin Volcano (Figure 1), which was a zone of focused melting and magma fractionation centred on Dunedin from ∼16 to 11 Ma (Benson 1941, 1942; Coombs et al. 1960; Coombs and Wilkinson 1969; Price and Compston 1973; Price and Taylor 1973; Price and Chappell 1975; Coombs et al. 2008; Price et al. 2003; Hoernle et al. 2006; Sprung et al. 2007; Scott et al. 2020a; Pontesilli et al. 2021, 2022; Baxter et al. 2022); and

2. the ∼24 to 9 Ma less well-studied remnants of outlying, widespread and typically monogenetic maars and lava flows dispersed in the surrounding Otago region (McDougall and Coombs 1973; Coombs et al. 1986; Reay et al. 1991; Nemeth and White 2003; Coombs et al. 2008; Scott et al. 2020a) (Figure 1). These outlying volcanic rocks were formerly attributed to the Waipiata Volcanic Field, but this term is unnecessary because these rocks are spatially and temporally part of the Dunedin Volcanic Group (see Scott et al. (2020a) for a discussion).

Rocks of the Dunedin Volcanic Group are primarily silica-undersaturated and represent a range of magma types. Basanite is most common but fractionation has resulted in phonolites and trachytes (Price et al. 2003; Coombs et al. 2008; Scott et al. 2020a; Pontesilli et al. 2021; Baxter et al. 2022; Pontesilli et al. 2022). Most of the Dunedin Volcano’s primitive components have potassic or sodic compositions whereas the outlying lavas are dominantly sodic (Reay et al. 1991; Coombs et al. 2008; McLeod and White 2018; Scott et al. 2020a; Pontesilli et al. 2021). Strontium, Nd, Pb and Hf isotopic data indicate that the mantle source to the bulk of the volcanic rocks trends towards the HIMU or FOZO mantle reservoirs (Price and Compston 1973; Hoernle et al. 2006; Sprung et al. 2007; Scanlan et al. 2020; Scott et al. 2020a; Pontesilli et al. 2021). Important recent discoveries are that the isotopic compositions of the alkaline volcanic rocks largely overlap with the isotopic properties of metasomatised mantle lithosphere beneath Otago (Scott et al. 2014a,b; McCoy-West et al. 2016; Scott et al. 2016; Dalton et al. 2017; Scott et al. 2023).

Central and eastern Otago – the area of our study – is characterised by a series of basins and ranges, of which one of the largest is the Maniototo Basin (Youngson et al. 1998; Craw et al. 2016) (Figure 1). The basin basement comprises greenschist facies Otago Schist (Mortimer 2000) overlain by Cenozoic sediments, which are in turn overlain and intruded by Dunedin Volcanic Group rocks (e.g. Bishop 1974; Coombs et al. 1986; Youngson et al. 1998). While volcanic rocks in the Maniototo Basin have long been examined (e.g. Benson 1941, 1942), key observations are only provided by Bishop (1974), Coombs et al. (1986) and Nemeth and White (2003). Of the > 90 dates for the entire volcanic field, only three are from the abundant volcanic rocks this basin (McDougall and Coombs 1973; Youngson et al. 1998) and the only published isotopic (Sr-Nd-Pb-Mg) compositions have been reported by Scott et al. (2020a) as part of a regional synthesis.

Methods

Major and trace element analysis

Whole rock major and trace element analysis was performed on rocks that were first trimmed using a diamond tipped rock saw, with all cut surfaces then ground using 100 grade grit to remove saw marks. Samples were then double bagged and smashed with a hammer, and then crushed to a powder in a tungsten carbide rock mill. Whole rock analysis was undertaken by ALS Geochemistry in Brisbane, Australia. Loss on ignition (LOI) values were measured using ALS standard method ME-GRAO05, with 1 g of sample heated to 1000°C before being cooled, reweighed and loss on ignition calculated. Trace elements were measured by inductively coupled plasma mass spectrometry (ICP-MS), following standard method ME-MS81 on solutions made by dissolution of glass beads from lithium-borate fusion.

Sr-Nd-Pb isotope analysis

Whole rock isotopic analysis was undertaken at the Department of Geological Sciences, University of Cape Town MC-ICP-MS facility. All analyses were conducted using a NuPlasma HR mass-spectrometer with a DSN-100 desolvating nebuliser, using laboratory and analytical procedures detailed by Harris et al. (2015). For each sample, 50 mg of sample powder was dissolved in a 4:1 HF/HNO₃ acid mixture before being converted to nitrates by applying a concentrated HNO₃ solution. Strontium, Nd, and Pb fractions were isolated through sequential column chemistry (Pin et al. 2014). The neodymium elemental fractions were analysed as 50 ppb 2% HNO₃ solutions and data normalised to 0.512115 for bracketing analyses, using Jndi-1 reference material (Tanaka et al. 2000). All Nd isotope data were corrected for Ce and Sm interference using the measured signals for ¹⁴⁰Ce and ¹⁴⁷Sm and the natural isotopic abundances of Sm and Ce. Instrumental mass fractionation was accounted for using the exponential law and a ¹⁴⁶Nd/¹⁴⁴Nd value of 0.7219. Strontium elemental fractions were analysed as 200 ppb 0.2% HNO₃ solutions with data normalised against 0.710255 for the bracketing analyses, with NIST SRM987 used as a reference material. ⁸⁷Sr/⁸⁶Sr data were corrected for Rb interference using the measured signal for ⁸⁵Rb and the natural ⁸⁵Rb/⁸⁷Rbratio, while instrumental mass fractionation was accounted for using the exponential law and an ⁸⁶Sr/⁸⁸Sr value of 0.1194. Lead elemental fractions were analysed as 50 ppb 2% HNO₃ solutions. Aliquots of NIST SRM997 Tl standard were added to all standards and samples analysed to give ± 10:1 Pb:Tl ratios. NIST SRM981 was the reference standard, with the following reference values used: ²⁰⁸Pb/²⁰⁴Pb: 36.7219, ²⁰⁷Pb/²⁰⁴Pb: 15.4963, and ²⁰⁶Pb/²⁰⁴Pb 16.9405 (Galer and Abouchami 1998). Instrumental mass fractionation of the measured Pb isotope ratios corrected using the reference ²⁰⁵Tl/²⁰³Tl value of 2.3889 and the exponential law. BHVO-2 standards processed and analysed concurrently yielded: ¹⁴³Nd/¹⁴⁴Nd = 0.512988 ± 0.000013 (n = 3) (published value of 0.512984 ± 0.000011), ⁸⁷Sr/⁸⁶Sr = 0.703488 ± 0.000011 (n = 3) (published value of 0.703479 ± 0.000020), and ²⁰⁶Pb/²⁰⁴Pb = 18.6238 ± 0.0011 (n = 2) (published value of 18.6290 ± 0.0092), ²⁰⁷Pb/²⁰⁴Pb = 15.5281 ± 0.0011 (n = 2) (published value of 15.5334 ± 0.0094), and ²⁰⁸Pb/²⁰⁴Pb = 38.2193 ± 0.0036 (n = 2) (published value of 38.2367 ± 0.0182). Published results are from Weis et al. (2006).

To assess the effect of post-magmatic alteration on the bulk-rock Sr analyses, in situ ⁸⁷Sr/⁸⁶Sr data for plagioclase and carbonate were obtained using a RESOlution 193 m ArF laser ablation system with an M-50 laser ablation cell coupled to a Nu Instruments Nu Plasma-HR MC-ICP-MS at the University of Otago, following the method and analytical configuration was reported in Scanlan et al. (2018). Each analysis was performed in raster mode in a pure He atmosphere and ablated material transported into the MC-ICP-MS using an Ar carrier gas with a flow rate of ∼1 L/min. The beam was 90 μm beam with a 10 Hz repetition rate and an on-sample fluence of 2.5 J/cm². Each analysis had a pre-ablation stage to remove possible surface contamination, followed by a 35 s cell gas purge before 20 to 40 s of sample ablation. Ion beam intensities at masses 88, 87, 86, 85.5, 85, 84, 83.5, 83, and 82 were monitored across an array of Faraday collectors operating with 10¹¹ohm resistors. Instrumental mass fractionation was corrected for using the measured ⁸⁶Sr/⁸⁸Sr ratio normalized to 0.1194 and the exponential mass fractionation law. The ⁸⁷Sr/⁸⁶Sr results were checked after correction for Rb, molecular dimer, and REE interferences, as well as instrumental mass bias for accuracy and precision and normalised to Tridacna clam (0.709176; Neymark et al. 2014) with an in-house anorthoclase standard KAN providing a reference check. KAN analyses yielded values of ⁸⁷Sr/⁸⁶Sr = 0.70290 ± 0.00006, which overlap the published value (0.70290; Burgin et al., 2023).

⁴⁰Ar/³⁹Ar Isotope analysis

To determine their age, two samples were irradiated along with Fish Canyon sanidine standard for 17.5 h in the off axis position at the United States Geological Survey TRIGA Reactor. The irradiated samples were analysed using the automated LDEO-AGES GV 5400 noble gas mass spectrometer with MassSpec software from the Berkeley Geochronology Center. Measurements were made in peak hopping mode on an analogue multiplier, and corrections were made for backgrounds based on blank runs approximately every 4^th sample run, mass discrimination based on air measurements approximately every 12^th sample run, and nuclear interferences based on production ratios reported in Dalrymple et al. (1981). The J-value applied, 3.648e⁻³ ± 2.5e⁻⁶ (0.07%) is based on the average values for Fish Canyon sanidine monitors from 4 pits. The individual dates are reported at ± 1σ analytical uncertainties, relative to the Fish Canyon age of 28.201 (Kuiper et al. 2008) and using the decay constants of Min et al. (2000).

Results

The volcanic rocks mapped and sampled across the Maniototo Basin (Figures 1 and 2, Table 1) are described below. We first give a field and petrographic description of each unit, and then subsequently use this framework to present the whole rock major, trace and isotope analyses.

Figure 1. Geological map of the Maniototo Basin. Volcanic rocks are shown in red. Buried lavas were interpreted by Scott et al. (2020a) using regional geophysical data.

Figure 2. A, Geological maps of the eastern Maniototo Basin, showing the distribution of the different geological units discussed in the paper, with a close up B, of the area comprising a variety of different magmatic rocks on the northern edge of the Waipiata flow. Geology from QMAP and our own mapping.

Table 1. Summary of the volcanic rocks units. Mineral abbreviations: Ol, olivine; Pl, plagioclase; Cpx, clinopyroxene; Il, ilmenite; Ti-mag, titanomagnetite, Carb, carbonate; Zeol, zeolite; Serp, serpentinite; Ne, nepheline; Gl, glass; Pal, palagonite; Idd, iddingsite.

	Age	Classification	Main mineralogy	Grouping
Haughton Hill	10.9 ± 0.3 Ma (^40Ar/^39Ar)	Basalt-basanite, alkaline-subalkaline	Ol-Pl phenocrysts, Pl-Ol-Cpx-Il-Ti-mag groundmass, Carb-Zeol alteration.	Transitional
Kokonga	11.8 ± 0.7 Ma (K-Ar)	Basalt, alkaline-subalkaline	Ol-Pl phenocrysts, Pl-Cpx-Il-Ti-mag groundmass, Carb alteration	Transitional
Swinburn North	13.4 ± 0.3 Ma (K-Ar)	Basanite, alkaline	Ol-phenocrysts/xenocrysts, Pl-Cpx-oxides groundmass	Alkaline
Swinburn South	16.5 ± 1.3 Ma (K-Ar) 15.6 ± 0.6 Ma (K-Ar)	Basalt-basanite, alkaline-subalkaline	Ol-Cpx-Pl phenocrysts, Pl-Cpx-(Ol)-oxide groundmass, some Carb-Zeol alteration	Transitional
Taieri River	undated	Basalt, alkaline	Ol-Pl-Ti-mag phenocrysts, Pl-Cpx-Ol-Ti-mag groundmass, Serp alteration	Intermediate
Tuneheketake	11.1 ± 0.7 Ma	Basanite-basalt, alkaline	Ol phenocrysts, Ol-Ti-mag-Il-Pl-Ne – Cpx-Gl groundmass, Zeol alteration	Alkaline
Waipiata	10.95 ± 0.11 Ma 11.12 ± 0.10 Ma (^40Ar/^39Ar)	Basalt, alkaline-subalkaline	Ol-Pl phenocrysts, Ol-Cpx-Pl-Il-Ti-mag-(Sp) groundmass, Carb-Pal-Idd alteration	Transitional
Dikes	undated	Basanite-basalt, alkaline	See text.	Alkaline

Haughton Hill

Haughton Hill is an isolated flow in rolling farmland on the western side of the Maniototo Basin (Figure 1). It is best exposed as a slightly greater than 10 m thick lava in an old quarry on the SW side of the hill (Figure 3A). It overlies a thin tuff layer, which in turn sits on Miocene terrestrial sediments. The lava is medium to fine grained, holocrystalline and has a diktytaxitic porphyritic texture. It contains phenocrysts of olivine and plagioclase and rare glomerocrysts of olivine. The olivine crystals form subhedral, corroded megacrysts ∼1200 μm long that contain inclusions of < 20 μm spinel and have calcite-pseudomorphed rims and fractures. The plagioclase phenocrysts are subhedral to anhedral, up 1000 μm long, and have been partially grown over by groundmass, which consists of plagioclase, olivine, clinopyroxene, ilmenite and titanomagnetite. Sub-horizontal coarsely crystalline segregation veins running thorough the basalt comprise plagioclase that often contains needle-like apatite inclusions, pink and green zoned clinopyroxene, opaque minerals, zeolites and interstitial carbonate (Figures 3B and 4A). Calcite-filled amygdales are common and carbonate is present in the ground mass. Haughton Hill has a ⁴⁰Ar/³⁹Ar date of 10.9 ± 0.3 Ma (unpublished data of D. Lee reported in Scott et al. (2020a)).

Figure 3. A, Exposure in the Haughton Hill quarry, with inset B, showing an example of the sub-horizontal segregation veins. Fine-grained Waipiata flow overlying coal-bearing silty-mudstone. C, Swinburn North volcanic rocks dip gently northwest. D, The Swinburn South volcanic rocks form a large plateau. E, Swinburn South contains coarse segregation patches.

Kokonga

Kokonga is located on the east of the major Waipiata lava (Figure 2A). It is a uniformly fine-grained, medium grey rock best exposed as an at least 5 m thick layer dipping gently northwards in the main roadside outcrop. Contacts with the adjacent Waipiata lava are unclear but it clearly overlies sedimentary rocks. Kokonga is holocrystalline, and has a porphyritic, intergranular texture with phenocrysts of olivine and plagioclase (Figure 4B). Olivine forms clusters of subhedral phenocrysts ∼1500-2000 μm long or as subhedral and rounded grains that often contain inclusions of spinel and have sieve textures close to the rim. The plagioclase phenocrysts are subhedral and are mainly ∼500 μm long and ∼200 μm wide. The crystalline groundmass consists of clinopyroxene, plagioclase, ilmenite and titanomagnetite, with secondary calcite present within small vesicles. It has a K-Ar age of 11.8 ± 0.7 Ma (Coombs et al. 2008).

Figure 4. Range of lithologies in the Maniototo basaltic rocks. A, Haughton Hill basalt with coarse segregations containing late stage minerals and green clinopyroxene overgrowths. B, Kokonga lava with diktytaxitic texture and vesicles infilled by carbonate. C, Swinburn North olivine-phyric lava with microcrystalline groundmass. D, Swinburn South coarsely-crystalline basalt (dolerite) with ophitic intergrowth of clinopyroxene and plagioclase. E, Tuneheketake finely crystalline lava containing kink-banded olivine xenocrysts. F, Waipiata coarsely crystalline lava with iddingsitised olivine phenocrysts. G, olivine phenocrysts in coarsely crystalline lava with unaltered rims over iddingsitised mantles. H, Waipiata finely crystalline dyke with zoned olivine phenocrysts. Mineral abbreviations: carb = carbonate, cpx = clinopyroxene, idd = iddingsite, plag = plagioclase, ol = olivine, zeol = zeolites.

Swinburn North

Swinburn North is located north and east of State Highway 85 (Figure 2A), and consists of thin lava (or lavas) that rest unconformably on Miocene terrestrial sediments that in turn overlie Cenozoic marine sediments (Youngson et al. 1998) (Figure 3C). These rocks have a microlitic groundmass of clinopyroxene, plagioclase and Fe-Ti oxides with euhedral to subhedral microphenocrysts of olivine (Figure 4C). Samples commonly contain many small cm or smaller sized peridotite (lherzolite) xenoliths. A Swinburn North sample has an unpublished K-Ar date of 13.4 ± 0.3 Ma (Youngson et al. 1998).

Swinburn South

Swinburn South is a major unit located along the eastern side of the Maniototo Basin (Figures 2A and 3D). It is dominated by a flow that is relatively thick compared with others in the Maniototo Basin, but also comprises multiple scoria cones, tuff rings, small diatremes and feeder dykes. It sits above thin pyroclastic layers, which in turn rest on Miocene lacustrine sediments of varied thickness. The lava is locally >30 m thick in the southern area where it fills a well-defined paleo-valley, and is consistently coarsely crystalline basalt with impressive very coarsely crystalline segregation structures. There is a broad variety of lava textures and pyroclastic rock types, often in restricted, m-scale areas, with different units having complex intrusion and mingling relationships. Lithologies vary from those with distinctive ophitic to poikilitic texture, defined by clinopyroxene and olivine oikocrysts (0.5-2 mm wide) surrounding plagioclase chadacrysts (<2 mm long) (Figure 4D), to finely crystalline olivine-free units. Segregation bodies have larger crystal sizes (5–20 mm long) and host late-stage evolved minerals such as rims of green clinopyroxene on pink cores, amphibole and zeolites. A slightly altered Swinburn South rock from a large disused quarry has two K-Ar ages of 16.5 ± 1.3 Ma and 15.6 ± 0.6 Ma (McDougall and Coombs 1973).

Taieri River

Taieri River is a dark grey, fine-grained body forming a small hill on the northeast side of the coarse-grained Waipiata flow (Figures 2B and 5A). It is holocrystalline and has an intergranular texture and contains phenocrysts of olivine, plagioclase and titanomagnetite. The olivine phenocrysts are subhedral with spinel inclusions, but are have fractures containing green-yellow (serpentinite?) alteration. The groundmass consists of plagioclase, clinopyroxene, olivine and titanomagnetite. The clinopyroxene is euhedral and pale grey with hints of pink, existing as both individual crystals and clumps, with rare inclusions. It is undated and the stratigraphic position is unclear other than being in close proximity to Eocene and younger sediments.

Tuneheketake

Tuneheketake, named after a historical lake in the region, is a dark grey body exposed along the northern side of the Waipiata flow close to the Taieri River flow (Figure 2A). The contacts between flows are not exposed and the flow has probably been extensively eroded. It has columnar jointing and has a vertical exposure of nearly 50 m in the southeastern portion (Figure 5B), which may be a plug. The lithology contains abundant ∼200 μm olivine phenocrysts as well as coarser, generally <500 μm larger anhedral kink banded olivine set in a fine groundmass consisting of glass, olivine, titanomagnetite and rare ilmenite, plagioclase, nepheline, and clinopyroxene (Figure 4E). The clinopyroxene is euhedral to subhedral and pale brown to pink and often forms rosette-like clumps. Green to brown alteration and rare zeolites are also present. This unit petrographically matches the sample reported as ‘Waipiata-Kokonga’ and K-Ar dated as 11.1 ± 0.7 Ma by Coombs et al. (2008).

Figure 5. A, Major outcrop of Taieri River basalt. B, The Tuneheketake unit is exposed over 50 m vertical meters in the southern area of the unit and consist of vaguely columnar-jointed rocks. Person in the centre for scale. C, The Waipiata flow is one of the largest in Otago. It dips gently northwards towards the camera, and most of the background is composed of the coarse-grained facies. The photo is taken from the roadside on the fine-grained facies, looking south. D, The fine-grained Waipiata flow overlies coaliferous silt-mudstone sediments that may be laucustrine. This image is from close to the bridge over the Taieri River, south of Waipiata. E, Dike 3 intrudes Cenozoic sediments and is readily visible at low flow in the Taieri River.

Waipiata

The Waipiata flow occurs over ∼ 30 km² and is one of the largest in the entire Dunedin Volcanic Group (Figures 1 and 2A and 5C). It has medium to coarse-grained facies forming the largest portion, which has undergone extensive slumping, as well as a fine-grained facies that occurs along the western to north-western side (Figure 2A and B). The medium to coarse-grained facies is up to ∼20 m thick but the fine-grained facies is only 2.5 m thick. Both sit stratigraphically above sedimentary rocks, although an exposed basal contact with underlying coalified (laucustrine?) sediments has only been observed beneath the fine-grained portion (Figure 5D).

The medium to coarse-grained facies (herein: ‘coarse-grained’) is medium to pale grey, crystal rich and contains abundant micro-vesicles (Figure 3B). It is holocrystalline with a porphyritic diktytaxitic texture, containing phenocrysts up to 2 mm of olivine and 0.5 mm zoned plagioclase (Figure 4F). The groundmass consists of clinopyroxene, fine-grained plagioclase, ilmenite and titanomagnetite, and olivine rimed by clinopyroxene. Olivine phenocrysts are variably iddingsitised and also show unaltered overgrowths on iddingsite mantles (Figure 4G).

The fine-grained facies has a porphyritic diktytaxitic texture. The phenocrysts consist of olivine, mostly between 0.5 and 1 mm, and rare euhedral to subhedral plagioclase (∼0.5 mm) with oscillatory zoning and tiny inclusions of clinopyroxene (< 20 μm). The groundmass consists of twinned plagioclase, clinopyroxene, olivine, ilmenite, titanomagnetite and spinel. Vesicles vary in size and shape, and are mainly filled with calcite and other secondary minerals. Some areas of the unit contain a small quantity of glass that has mostly altered to palagonite and other associated minerals.

⁴⁰Ar/³⁹Ar dates were obtained for the coarse- and fine-grained facies, and full raw data are provided in Supplementary Table 1. For each sample, the data were examined on a step heating diagram with ages calculated assuming atmospheric initial (Lee et al. 2006), each was plotted on an isochron, and each was plotted on a second step-heating diagram using the initial from the isochron. In the ideal case all three of these approaches would yield the same result and all data would fall on the plateau. As this is rarely the case, an unemotional strategy for eliminating steps must be used. We only omitted steps at the beginning and end of the experiment, leaving the maximum number of steps in the middle, while dropping the MSWD to acceptable levels (ideally ∼1). In the case of LW 23GW, all but the final 3 steps (92.7% of the ³⁹Ar) lie on an isochron with age of 10.95 ± 0.15 Ma, initial of 295 ± 4, and MSWD of 1.1. In addition to significantly increasing the age scatter, the Ca/K of the last three steps is notably higher than the others and the %radiogenic is negative (values lower than atmosphere). The plateau age using the isochron initial is 10.95 ± 0.11 Ma, with MSWD. of 0.98 and p = 0.44 (Figure 6A). The data have an isochron age of 10.95 ± 0.15 Ma with an ⁴⁰Ar/³⁶Ar intercept of 295 ± 4, which overlaps the atmospheric value and indicates that this sample contains very little (or no) excess Ar (Figure 6B). The plateau age using the isochron initial is therefore the preferred age.

Figure 6. ⁴⁰Ar/³⁹Ar dates from two Waipiata samples. A, B, coarse-grained Waipiata; C, D, fine-grained Waipiata. For the plateaus, the dark shaded areas represent the preferred plateau steps used to calculate an age. For the isochrons, data excluded are shown in red and the letters correspond to step-heating stages. Errors are shown by the size of the boxes and ellipses.

Sample LW 2GW contains several steps at the beginning and end of analysis with low apparent ages calculated using the assumption of atmospheric initial (Figure 6C). The five steps that define the plateau account for 72.6% of the ³⁹Ar and these yield a plateau age of 11.12 ± 0.10 Ma, with MSWD of 0.7 and p of 0.59. The isochron plot is messy, with MSWD of 13 and p of 0 (Figure 6D). The 5 steps that define the plateau yield an anomalously low MSWD of 0.082 with p of 0.92. Accordingly the plateau age using the Lee et al. (2006) initial is considered the best estimate of the age of this sample. The two sample ages LW 23GW = 10.95 ± 0.11 Ma and LW 2GW = 11.12 ± 0.10 Ma are indistinguishable, and we note with interest that both end with steps that have much higher Ca/K and anomalously low ⁴⁰Ar/³⁶Ar, potentially related to the eruption or early alteration processes.

Dikes

Several dikes and/or plugs occur in sediments (Figures 2B and 5E) but are not readily attributed to any other component. Dike 1 is dark grey with olivine phenocrysts in a very fine groundmass and olivine, plagioclase, opaque minerals and rare clinopyroxene (Figure 4H). Dike 2, consisting of multiple offshoots emplaced into fine sediments and very well exposed in the riverbed, is black, weathering to dark brown, and fine-grained. Small (≤ 0.5 mm) crystals of olivine, clinopyroxene and plagioclase are visible, as well as small xenoliths of incorporated sediment (< 3%; < 5 mm across). Dike 3 forms a singular isolated outcrop between the Waipiata and Kokonga flows and emplaced into the same sediments. The body is light to medium grey in colour, with varying concentrations of vesicles, small phenocrysts of olivine and plagioclase and rare (< 1%), small (< 1 cm) schist and tiny peridotite xenoliths.

Whole rock geochemistry

As the volcanic rocks of the Maniototo Basin show signs of alteration, we first assess the affect of this on major and trace element compositions before classifying the suites. Using our new data coupled with that from Donnelly (1996), Coombs et al. (2008) and Scott et al. (2020a) (Table 2), there is major element variation that correlates with LOI. CaO, for example, shows a positive trend against LOI in Taieri River, Waipiata, Haughton Hill and, possibly, Swinburn South samples (Figure 7A). There is slight K₂O loss in the Swinburn South samples (Figure 7B). Fe₂O₃ and MgO have negative trends in Taieri River, Swinburn South and Haughton Hill rocks (Figure 7C, D). The MgO content of the Waipiata samples either shows a slight positive trend with LOI, or scatter towards low MgO contents at low LOI (Figure 7D). On the basis of trends against LOI, we therefore exclude from the characterisation: Haughton Hill 1aHH and OU66560; Swinburn North SP; Swinburn South G15 and STOP3; Waipiata OU66561, 5aW, 20aUL and 25aFL; and Taieri River 11aUL.

Figure 7. Selected major elements plotted against loss on ignition.

Table 2. Whole rock major and trace element analyses of samples analysed in this study, together with those from Donelly (1996), Coombs et al. (2008) and Scott et al. (2020a).

	HAUGHTON HILL				KOKONGA				WAIPIATA
Source	Donnelly (1996)	this study	Donnelly (1996)	this study	Coombs et al. (2008)	Donnelly (1996)	this study	this study	this study	Donnelly (1996)	this study	this study	this study	this study	this study	this study	this study	this study	this study
Field No.	OU66560	1aHH	OU66588	1lHH	OU54890	OU62555	6aK	5aW	2aW	OU66561	25aFL	20aUL	3aML	21aUL	30aUL	23aUL	28aUL	12aUL	10aUL
								fg	fg	cg	cg	cg	cg	cg	cg	cg	cg	cg	cg
SiO2	43.70	44.60	45.10	45.20	42.89	44.83	46.90	43.70	45.70	42.57	44.10	45.90	46.00	46.60	47.30	47.60	47.80	48.10	48.70
Al2O3	13.00	13.50	13.40	13.75	12.81	13.22	13.95	12.80	13.45	12.42	13.70	14.15	14.05	13.75	14.90	14.70	14.70	15.40	15.30
Fe2O3	13.40	12.95	13.50	12.85	13.91	15.02	12.40	12.00	12.80	11.94	12.35	13.05	13.35	13.75	13.60	13.75	13.50	13.30	13.45
CaO	10.90	11.65	10.20	9.67	11.20	9.80	10.15	11.75	9.31	13.09	12.60	12.10	11.00	9.65	11.05	9.83	10.25	10.45	10.40
MgO	7.15	6.94	8.42	8.58	10.95	9.29	9.67	8.56	9.02	9.49	8.44	8.02	8.12	8.17	7.82	7.96	7.53	6.44	5.06
Na2O	2.73	2.74	2.63	2.48	3.34	3.01	2.59	2.31	2.49	2.43	2.86	2.69	2.73	2.52	2.74	2.77	2.96	3.00	3.09
K2O	0.76	0.68	0.74	0.74	1.32	0.84	0.82	0.72	0.73	0.85	0.79	0.84	0.81	0.86	0.88	0.85	0.85	0.92	0.96
Cr2O3		0.05		0.05			0.06	0.06	0.05		0.05	0.05	0.05	0.07	0.06	0.06	0.05	0.04	0.04
TiO2	1.86	1.89	1.82	1.79	2.43	2.09	2.21	1.82	1.94	1.96	1.86	1.96	1.97	1.96	2.07	1.98	2.01	2.21	2.27
MnO	0.17	0.16	0.17	0.16	0.21	0.26	0.15	0.18	0.15	0.32	0.16	0.19	0.17	0.18	0.18	0.19	0.17	0.18	0.19
P2O5	0.26	0.23	0.26	0.26	0.77	0.37	0.33	0.32	0.24	0.40	0.27	0.31	0.28	0.35	0.34	0.29	0.29	0.36	0.43
LOI	5.52	6.17	3.80	3.07	−0.38	1.63	−0.11	3.25	2.24	4.03	4.08	2.22	1.50	1.06	0.82	0.50	0.67	0.59	0.90
Total	99.45	101.56	100.04	98.60	99.45	100.36	99.12	97.47	98.12	99.50	101.26	101.48	100.03	98.92	101.76	100.48	100.78	100.99	100.79
Ba	218	236	245	372	596	240	238	248	183	314	198	346	254	462	275	405	244	291	274
Ce		35		36	102	51	46	45	36		36	40	40	45	43	38	39	47	46
Cr	298	390	294	340	339	372	470	460	390	388	360	410	420	490	390	430	380	300	280
Cs		0.64		1			0.3	0.24	0.3		0.37	0.24	0.31	0.21	0.16	0.27	0.35	0.35	0.46
Dy		4.3		4.3			4.8	4.4	4.4		4.3	4.4	4.5	5.0	4.8	4.5	4.6	5.0	5.2
Er		2.3		2.3			2.5	2.3	2.3		2.3	2.3	2.4	2.7	2.5	2.5	2.4	2.7	2.7
Eu		1.5		1.5			1.8	1.6	1.6		1.6	1.6	1.7	1.8	1.8	1.6	1.7	1.9	2.0
Ga		20		19	19	20	20	18	19		19	19	20	19	20	20	20	21	22
Gd		4.6		4.8			5.5	5.0	4.8		4.8	5.0	5.0	5.7	5.6	5.0	5.1	5.8	5.9
Hf		3.1		3.2			3.8	3.4	3.5		3.3	3.6	3.6	3.8	3.8	3.7	3.6	4.2	4.2
Ho		0.81		0.88			0.93	0.84	0.85		0.83	0.83	0.86	0.95	0.92	0.87	0.90	0.96	0.99
La	15	17	22	17	68	24	22	22	17	29	17	19	19	23	24	18	19	22	22
Lu		0.27		0.28			0.29	0.27	0.29		0.28	0.3	0.28	0.33	0.3	0.3	0.29	0.31	0.32
Nb	21	22	21	23	199	29	31	33	24	33	23	25	25	31	27	25	25	30	29
Nd	24	18	18	19	41	28	24	22	20	25	19	21	21	24	23	21	22	25	25
Pr		4.4		4.5		13.0	5.8	5.5	4.5		4.7	4.9	5.0	5.8	5.6	4.9	5.1	6.0	6.0
Rb	22	18		18	36	19	19	17	18	25	18	17	18	19	18	18	20	21	22
Sm		4.5		4.4			5.4	4.7	4.6		4.4	4.7	4.7	5.5	5.0	4.8	4.9	5.5	5.8
Sr	442	535		512	730	414	455	450	369	450	419	417	446	446	443	441	423	474	458
Ta		1.3		1.4			1.8	2.0	1.4		1.4	1.5	1.4	1.8	1.6	1.4	1.5	1.7	1.8
Tb		0.72		0.70			0.80	0.74	0.76		0.73	0.73	0.77	0.87	0.84	0.76	0.80	0.89	0.90
Th		2.4		2.4	10.0	2.0	3.2	3.0	2.5		2.5	2.7	2.8	3.0	2.9	2.7	2.7	3.2	3.2
Tm		0.30		0.31			0.33	0.31	0.32		0.30	0.32	0.33	0.35	0.33	0.32	0.32	0.35	0.36
U		0.67		0.62	7.00	1.00	0.81	0.78	0.65		0.65	0.65	0.74	0.72	0.46	0.67	0.62	0.80	0.69
V	197	230	199	208	195	186	236	213	214	182	217	218	224	244	236	235	234	242	260
Y	17	21		21	33	24	23	21	21	19	21	22	22	26	24	23	23	25	26
Yb		1.9		1.8			1.9	1.7	1.9		1.8	1.9	1.9	2.1	1.9	1.9	2.0	2.1	2.1
Zr	113	127		128	196	132	150	133	135	121	136	140	142	151	152	146	146	167	167

Table 2. Continued

	SWINBURN SOUTH												TUNAHEKETEKE
Source	Scott et al. (2020a)	Scott et al. (2020a)	Scott et al. (2020a)	Scott et al. (2020a)	Scott et al. (2020a)	Scott et al. (2020a)	Scott et al. (2020a)	Scott et al. (2020a)	Scott et al. (2020a)	Scott et al. (2020a)	Scott et al. (2020a)	Donnelly (1996)	this study	this study
Field No.	STOP3	G15	E9	E16	J12	SWIN-1	E14	D10	J19	E5	A2	OU66587	8aUL	16aUL
SiO2	43.20	44.30	45.20	45.20	45.20	45.40	45.50	46.20	46.30	46.70	46.80	48.00	43.50	43.70
Al2O3	13.80	14.15	14.20	14.30	14.35	14.40	14.30	14.10	13.95	14.20	13.90	14.30	13.20	13.40
Fe2O3	13.15	12.80	13.95	13.75	13.90	13.35	13.75	13.20	13.45	13.55	12.65	11.20	13.60	13.90
CaO	11.55	9.87	9.78	10.40	9.97	9.85	9.38	9.01	9.13	9.30	9.29	10.90	11.10	10.80
MgO	8.32	9.77	9.62	9.73	9.84	9.60	10.90	9.05	10.50	10.15	9.92	5.47	10.90	10.70
Na2O	3.04	1.45	2.83	2.37	2.75	2.69	2.54	2.50	2.69	2.61	2.59	2.88	3.55	2.53
K2O	0.49	0.47	0.85	0.89	0.68	0.78	0.58	0.61	0.71	1.00	0.80	1.13	1.40	1.08
Cr2O3													0.06	0.05
TiO2	1.52	1.43	1.76	1.71	1.63	1.57	1.45	1.51	1.63	1.76	1.55	2.17	2.41	2.47
MnO	0.18	0.16	0.18	0.18	0.18	0.18	0.18	0.17	0.17	0.17	0.17	0.16	0.20	0.20
P2O5	0.23	0.20	0.26	0.27	0.27	0.26	0.21	0.24	0.29	0.31	0.27	0.37	0.75	0.71
LOI	4.69	4.92	0.34	1.38	0.74	2.43	1.31	2.75	1.29	0.66	1.64	3.00	−0.32	1.44
Total	100.17	99.52	98.97	100.18	99.51	100.51	100.10	99.34	100.11	100.41	99.58	99.58	100.35	100.98
Ba	188	169	208	222	205	173	162	236	250	221	479	352	538	422
Ce	40	41	54	59	56	43	42	42	48	52	46		108	90
Cr	310	330	310	300	300	320	330	440	420	390	470	75	420	400
Cs	0	0	1	0	1	1	0	1	1	1	1		0.45	0.38
Dy	3.7	3.5	4.3	4.2	3.9	3.8	3.5	3.7	3.5	4.0	3.6		6.5	6.1
Er	1.9	1.7	2.1	2.1	2.1	2.0	1.8	1.8	1.8	2.0	1.8		3.3	3.0
Eu	1.5	1.4	1.7	1.7	1.6	1.5	1.4	1.5	1.7	1.7	1.6		2.8	2.7
Ga	17	17	20	20	20	18	18	19	20	20	19		19	20
Gd	3.7	3.7	4.6	4.6	4.4	4.0	3.8	4.2	4.0	4.5	4.2		8.1	7.5
Hf	2.4	2.5	3.0	2.9	2.9	2.5	2.4	2.6	3.1	3.2	2.8		4.8	4.6
Ho	0.70	0.66	0.76	0.80	0.76	0.74	0.67	0.69	0.67	0.73	0.71		1.22	1.13
La	21	22	29	31	29	23	22	22	25	27	24	38	58	47
Lu	0.2	0.2	0.3	0.3	0.3	0.3	0.2	0.2	0.2	0.2	0.2		0.41	0.36
Nb	27	27	41	43	38	30	27	26	30	32	27	32	85	72
Nd	19	18	24	25	23	20	19	20	22	24	22	35	47	41
Pr	4.6	4.6	6.1	6.5	5.9	5.0	4.7	4.9	5.4	5.9	5.3		12.2	10.4
Rb	14	12	22	23	21	35	16	17	25	26	30		40	29
Sm	4.0	3.9	5.1	5.3	4.8	4.1	4.1	4.5	4.6	5.0	4.7		8.6	8.2
Sr	391	402	357	395	349	419	319	553	1115	404	921		779	673
Ta	1.7	1.8	2.9	2.8	2.6	1.8	1.9	1.8	2.0	2.2	1.9		4.7	4.0
Tb	0.59	0.61	0.75	0.74	0.67	0.64	0.59	0.64	0.66	0.70	0.65		1.13	1.09
Th	3.5	3.4	4.3	4.4	4.4	3.6	3.3	3.2	3.9	4.2	3.6		7.3	5.7
Tm	0.26	0.25	0.32	0.29	0.29	0.28	0.26	0.24	0.24	0.27	0.25		0.45	0.40
U	0.94	0.74	0.96	0.97	1.01	0.87	0.71	0.73	0.88	0.94	0.83		2.03	1.49
V	239	210	224	227	229	231	206	201	205	218	195	222	230	247
Y	21	18	22	22	21	22	18	19	18	20	19		31	29
Yb	1.7	1.5	1.8	1.7	1.8	1.7	1.6	1.5	1.5	1.6	1.5		2.7	2.4
Zr	95	93	123	122	111	106	94	106	121	132	117		214	195

Table 2. Continued

		SWINBURN NORTH								DIKES			TAIERI RIVER
Source	Coombs et al. (2008)	Scott et al. (2020a)	Scott et al. (2020a)	Scott et al. (2020a)	Scott et al. (2020a)	Scott et al. (2020a)	Scott et al. (2020a)	Scott et al. (2020a)	Scott et al. (2020a)	this study	this study	this study	this study	this study
Field No.	54921	SP (Dyke)	I5	F1B	L4	L14	H8	L15	I14	18aD	7aKK	13aUL	11bUL	11aUL
										2	3	1
SiO2	47.59	39.60	41.40	41.40	41.60	42.30	43.00	43.70	44.10	42.70	43.60	47.60	44.80	45.30
Al2O3	14.21	12.65	13.15	13.40	13.00	13.05	13.65	13.15	13.95	12.95	13.45	13.10	13.10	15.05
Fe2O3	13.39	13.45	13.70	14.30	14.20	14.75	13.75	14.35	14.50	13.40	12.60	12.95	13.35	12.05
CaO	9.76	9.33	9.35	10.20	11.40	10.35	10.35	9.93	9.72	11.30	11.50	10.15	10.40	11.85
MgO	8.26	7.75	8.63	9.55	7.78	9.32	9.36	8.66	8.55	10.15	11.20	9.26	11.00	6.62
Na2O	3.55	4.70	4.05	4.04	4.83	3.86	3.82	3.94	4.07	2.51	2.43	3.24	2.06	2.18
K2O	0.85	1.77	1.43	1.32	1.21	1.44	1.34	1.66	1.47	1.17	1.14	1.12	1.22	1.66
Cr2O3										0.06	0.06	0.06	0.06	0.04
TiO2	2.02	2.26	2.40	2.34	2.51	2.52	2.25	2.37	2.48	2.33	2.47	2.22	2.10	2.36
MnO	0.19	0.23	0.21	0.23	0.21	0.23	0.23	0.20	0.26	0.23	0.17	0.26	0.18	0.16
P2O5	0.30	0.93	0.67	0.67	1.01	0.98	0.66	0.89	0.72	0.71	0.76	0.58	0.42	0.54
LOI	0.09	5.60	2.51	3.69	1.50	0.63	2.22	0.14	1.06	1.41	0.89	0.36	1.26	2.19
Total	100.21	98.27	97.50	101.14	99.25	99.43	100.63	98.99	100.88	98.92	100.27	100.90	99.95	100.00
Ba	259	662	495	429	684	593	428	560	486	541	1290	377	286	353
Ce	36	166	101	101	160	147	93	129	110	100	110	74	58	73
Cr	403	290	340	340	310	350	360	330	350	450	440	390	430	300
Cs		1	0	0	1	1	0	1	0	0.53	0.26	0.63	0.44	0.54
Dy		7.1	5.7	5.3	6.6	6.8	5.2	6.0	5.8	6.2	6.5	5.3	5.0	5.8
Er		3.4	2.7	2.6	3.2	3.1	2.5	2.8	2.7	3.0	3.2	2.8	2.5	3.0
Eu		3.7	2.8	2.7	3.7	3.7	2.6	3.2	3.0	2.7	2.9	2.2	2.0	2.4
Ga	18	19	19	20	20	20	19	19	20	19	19	18	18	21
Gd		8.8	6.8	6.6	8.9	8.6	6.5	7.7	7.0	7.6	8.1	6.5	5.8	7.0
Hf		6.7	5.4	5.1	6.2	5.7	4.8	5.5	5.6	4.6	4.9	4.3	3.9	4.3
Ho		1.36	1.05	0.96	1.19	1.23	0.98	1.10	1.04	1.19	1.22	1.02	0.96	1.09
La	27	93	52	53	87	78	49	67	57	53	59	39	30	37
Lu		0.4	0.3	0.3	0.4	0.4	0.3	0.3	0.3	0.37	0.38	0.32	0.3	0.36
Nb	22	121	76	72	109	93	65	79	80	77	82	54	45	55
Nd	24	65	44	43	64	61	41	53	46	43	47	35	27	34
Pr	3.0	17.9	11.5	11.2	16.9	15.9	10.5	13.7	12.1	11.5	12.4	8.8	7.1	8.7
Rb	19	58	32	27	23	27	36	48	36	34	31	30	30	35
Sm		10.9	8.3	8.3	11.6	11.1	8.1	10.0	8.8	8.3	9.0	6.7	5.9	7.0
Sr	390	1140	850	778	1085	1065	744	879	828	755	1140	580	464	642
Ta		8.0	4.8	4.8	7.0	5.8	4.5	5.2	5.4	4.3	4.7	3.1	2.6	3.2
Tb		1.27	1.00	1.01	1.30	1.28	0.98	1.10	1.07	1.11	1.13	0.95	0.85	1.03
Th		13.8	6.5	5.8	9.5	8.4	5.4	7.5	6.4	6.6	7.3	5.1	4.0	4.8
Tm		0.47	0.36	0.35	0.43	0.43	0.35	0.38	0.39	0.41	0.44	0.38	0.34	0.40
U		3.28	1.65	1.67	2.41	1.98	1.39	1.85	1.57	1.72	1.71	1.28	1.04	1.29
V	208	195	209	196	195	207	196	198	200	234	236	214	226	252
Y	22	37	29	27	34	33	26	30	28	30	31	26	24	28
Yb		3.0	2.2	2.1	2.5	2.6	2.1	2.4	2.2	2.5	2.6	2.2	2.1	2.4
Zr	131	319	240	227	283	256	210	239	248	204	214	180	156	186

The filtered Maniototo Basin volcanic rocks comprise two major groups. On a total alkali versus silica diagram, the Haughton Hill, Kokonga, Swinburn South and Waipiata samples are mostly classified as basalt and plot close to or across the alkaline-subalkaline divide (Figure 8A). They vary between silica-saturated and undersaturated (Figure 8B) and these rocks are herein referred to as the transitional group. On the other hand, the Swinburn North, Tuneheketake and the three dike samples are mostly basanites and fall within the alkaline field (Figure 8A). They are moderately to strongly silica-undersaturated (3.8 to 21.8 wt% normative nepheline) and are herein referred to as the alkaline group. Taieri River plots between the alkaline and transitional groups; it is basalt (Figure 8A) and only weakly alkaline (Figure 8B). Mg numbers (Mg#, where Mg# = 100*Fe²⁺/(Fe²⁺+Mg)) of the groups range between 44 and 65 with no correlation with alkali content, other than the alkaline group having a very weak negative correlation (Figure 8B).

Figure 8. A, Total alkalis versus silica diagram (Le Bas et al. 1986) showing the classification of the different magmatic rocks in the Maniototo Basin as basalt or basanite. The grey field is other Dunedin Volcanic Group rocks, taken from Scott et al. (2020a). B, Mg# versus normative nepheline illustrating the degree of alkalinity of the Maniototo Basin components.

In terms of major elements, the filtered alkaline group tends to have lower SiO₂ and Al₂O₃ (Figure 9A) but higher Na₂O (Swinburn North only) and K₂O (Figure 9B), P₂O₅ (not shown), TiO₂ (Figure 9C) than the filtered transitional group (Table 2). The two groups have very similar MgO except in samples above ∼47 wt% SiO₂ (Figure 9D). Taieri River is anomalous in that it has the high SiO₂ content mostly seen in the transitional group (Figure 9A), but a K₂O content equivalent to the alkaline group, as well as lower Na₂O than most other samples (Figure 9B). One dike and one Tuneheketake sample also have major element compositions more similar to the transitional group than the alkaline group; however, their trace element abundance and isotopic ratios (discussed below) indicate that they belong to the alkaline group. Within the transitional group, the Swinburn South rocks are slightly anomalous because they have lower TiO₂ (Figure 9C) and slightly higher MgO (Figure 9D) than the other rocks. The Light Rare Earth Element (REE) to Heavy REE concentrations and ratios (La/Yb) show the alkaline group samples to have steeper REE concentrations than the transitional group (Figure 9E). The Sr concentrations of the alkaline group are also higher than those of the transitional group (Figure 9F), except that several of the Swinburn South rocks have distinctly high and anomalous Sr abundances (Table 2).

Figure 9. Diagrams showing relationships between the different magmatic groups and trends in the elements and element ratios.

A trace element normalized diagram of all analyses shows that the transitional group is relatively depleted compared to the alkaline group, with Taieri River plotting at intermediate values (Figure 10). Transitional group rocks exhibit small Ti* ( = Ti_N/(Eu_N x Gd_N)^0.5, where _N = primitive mantle-normalised) of 0.9 to 1.1 compared to the alkaline group (0.6 to 0.9). The trend is similar in K*( = K_N/(Nb_N x La_N)^0.5): the transitional group (K* = 0.7 to 1.1) has a smaller negative anomaly compared to the alkaline group (0.3 to 0.6), with the exception of Swinburn South (K* of 0.4 to 0.8, median 0.6). This widespread negative K* anomaly cannot be a function of alteration, even though K may be slightly affected by this process (Figure 7B), because the anomaly is pervasive. Similarly, the positive P anomaly (Figure 10) is also pervasive. Zirconium_N/Hf_N is also slightly lower in the transitional rocks (1.03 to 1.14) compared to the alkaline group (1.15 to 1.31), once again excluding Swinburn South (1.03 to 1.17). Taieri River is has intermediate ratios (Zr_N/Hf_N  = 1.10 to 1.19).

Figure 10. Primitive mantle-normalised data illustrating the trace element differences and similarities between the different Maniototo Basin volcanic rocks. Data are from this study and also Scott et al. (2020a).

Sr-Nd-Pb isotopes

New Sr-Nd-Pb isotopic analyses are presented for 21 Maniototo Basin volcanic rocks to aid distinction of the magmas and characterisation of their sources. In addition, we incorporate published data from two Haughton Hill samples, four Swinburn South, three Swinburn North, one Waipiata and one Kokonga sample that were reported by Scott et al. (2020a) (Table 3). The ⁸⁷Sr/⁸⁶Sr and ¹⁴³Nd/¹⁴⁴Nd data are all corrected to 11 Ma based on the interpretation of existing K-Ar and new ⁴⁰Ar/³⁹Ar data, although in detail this makes little difference to the results because the age correction from measured values is commonly less than 2 standard errors on the measurements. All Pb data are uncorrected due to a lack of sufficient quality Pb, U and Th concentration data.

Table 3. Sr-Nd-Pb isotopic data for the studied samples. Several data are from Scott et al. (2020a).

	Source (if not this study)	Rb	Sr	^87Rb/ ^86Sr	^87Sr/^86Sr	+/- 2 se	^87Sr/^86Sr 11 Ma		Nd	Sm	^147Sm/ ^144Nd	^143Nd/ ^144Nd	+/- 2 se		^143Nd/ ^144Nd 11 Ma	ϵNd		^206Pb/^204Pb	+/- 2 se	^207Pb/^204Pb	+/- 2 se	^208Pb/^204Pb	+/- 2 se
		ppm	ppm						ppm	ppm
Haughton Hill
1aHH		18	535	0.10	0.70462	12	0.70461		18	4.5	0.15	0.51289	14		0.51287	4.6		19.222	9	15.646	9	38.986	3
1lHH		18	512	0.10	0.70478	12	0.70477		19	4.4	0.14	0.51289	13		0.51288	4.7		19.217	9	15.648	8	38.997	3
66560	Scott et al. (2020a)			0.14	0.70448	12	0.70446				0.13	0.51288	12		0.51287	4.5		19.226	10	15.652	10	39.007	3
66588	Scott et al. (2020a)			0.18	0.70416	12	0.70413				0.13	0.51290	12		0.51289	4.9		19.218	10	15.649	10	38.994	3
FG Waipiata
2aW		18	369	0.14	0.70374	12	0.70372		20	4.6	0.14	0.51287	14		0.51286	4.3		19.233	10	15.648	10	38.995	3
5aW		17	450	0.11	0.70330	11	0.70328		22	4.7	0.13	0.51291	12		0.51290	5.2		19.508	10	15.65	9	39.180	3
Waipiata
3aML		18	446	0.12	0.70375	15	0.70373		21	4.7	0.14	0.51287	11		0.51286	4.3		19.227	9	15.647	8	38.992	3
10aUL		22	458	0.14	0.70341	10	0.70338		25	5.8	0.14	0.51289	12		0.51288	4.7		19.239	9	15.65	10	39.006	4
12aUL		21	474	0.13	0.70336	13	0.70334		25	5.5	0.13	0.51289	12		0.51288	4.7		19.255	9	15.649	10	39.009	3
20aUL		17	417	0.12	0.70349	10	0.70347		21	4.7	0.13	0.51288	14		0.51287	4.5		19.212	11	15.644	12	38.978	4
21aUL		19	446	0.12	0.70338	10	0.70336		24	5.5	0.14	0.51288	9		0.51287	4.5		19.318	11	15.643	10	39.044	3
23aUL		18	441	0.12	0.70363	12	0.70361		21	4.8	0.14	0.51288	12		0.51287	4.6		19.224	9	15.648	9	38.992	3
25aFL		18	419	0.13	0.70364	10	0.70362		19	4.4	0.14	0.51289	12		0.51288	4.7		19.217	11	15.65	12	38.985	4
28aUL		20	423	0.14	0.70345	14	0.70343		22	4.9	0.14	0.51288	10		0.51287	4.5		19.22	13	15.65	14	38.989	4
30aUL		18	443	0.12	0.70340	14	0.70338		23	5.0	0.13	0.51290	12		0.51289	4.9		19.225	10	15.647	10	39.000	3
66561	Scott et al. (2020a)			0.16	0.70326		0.70323				0.13	0.51293	12		0.51292	5.5		19.609		15.659	10	39.274	3
Kokonga
6aK		19	455	0.12	0.70329	11	0.70327		24	5.38	0.14	0.51290	11		0.51289	4.9		19.301	10	15.642	11	39.030	3
62555	Scott et al. (2020a)				0.70329	12	0.70329				0.13	0.51289			0.51288	4.7		19.25	10	15.644		38.993
Swinburn South
SWIN1	Scott et al. (2020a)			0.24	0.70424	12	0.70420				0.12	0.51285	12		0.512839	3.9		19.982	10	15.699	10	39.542	3
A2	Scott et al. (2020a)			0.94	0.70568	12	0.70554				0.13	0.51286	12		0.512846	4.1		19.519	10	15.666	10	39.220	3
E16	Scott et al. (2020a)			0.17	0.70312	12	0.70309				0.13	0.51287	12		0.512863	4.4		20.288	10	15.717	10	39.734	3
J12	Scott et al. (2020a)			0.17	0.70315	12	0.70312				0.13	0.51287	12		0.512861	4.4		20.219	10	15.711	10	39.682	3
Taieri River
11aUL		35	642	0.16	0.70306	10	0.70303		34	7.0	0.12	0.51291	14		0.51290	5.1		19.64	9	15.65	10	39.27	3
11bUL		30	464	0.19	0.70318	10	0.70315		27	5.9	0.13	0.51291	10		0.51290	5.1		19.62	12	15.65	11	39.27	4
Tunaheketeke
8aUL		40	779	0.15	0.70294	13	0.70292		47	8.6	0.11	0.51294	9		0.51293	5.8		19.961	10	15.666	10	39.524	3
16aUL		29	673	0.12	0.70293	11	0.70291		41	8.2	0.12	0.51294	12		0.51293	5.7		19.918	11	15.661	10	39.479	3
Dikes
13aUL (1)		30	580	0.15	0.70314	15	0.70312		35	6.7	0.12	0.51292	15		0.51291	5.4		19.643	11	15.659	11	39.295	4
18aD (2)		34	755	0.13	0.70317	10	0.70315		43	8.3	0.12	0.51293	13		0.51292	5.5		19.944	10	15.662	7	39.501	3
7aKK (3)		31	1140	0.08	0.70425	15	0.70424		47	9.0	0.12	0.51293	11		0.51292	5.4		19.953	9	15.663	9	39.506	3
Swinburn North
H8	Scott et al. (2020a)			0.13	0.70308	12	0.70306				0.12	0.51292	12		0.512910	5.3		19.872	10	15.653	12	39.390	3
L15	Scott et al. (2020a)			0.16	0.70300	12	0.70298				0.09	0.51291	12		0.512901	5.1		19.765	10	15.655	12	39.347	3
I5	Scott et al. (2020a)			0.11	0.70310	12	0.70308				0.11	0.51291	12		0.512905	5.2		19.959	10	15.655	12	39.444	3

The transitional group (Haughton Hill, Waipiata, Swinburn South and Kokonga) range from ⁸⁷Sr/⁸⁶Sr_{11 Ma} of 0.70309 to 0.70554 (Table 3). The Waipiata fine-grained and coarse-grained flows are isotopically indistinguishable and are very similar to Kokonga, whereas Haughton Hill and Swinburn South extend to significantly more radiogenic Sr values (Figure 11A). However, as a test for whether the bulk rock Sr in those radiogenic samples could be due to post-magmatic alteration, the compositions of magmatic feldspars and post-magmatic carbonate in the two Haughton Hill rocks were analysed via in-situ laser ablation. These data show the feldspars to have ⁸⁷Sr/⁸⁶Sr values of 0.70326 ± 0.00016 and the post-magmatic carbonates to have ⁸⁷Sr/⁸⁶Sr of 0.70785 ± 0.00016 (Table 4). Similarly, the most radiogenic Swinburn South sample also has the highest Sr content (921 ppm compared to the average of 388 ppm for the other three isotopically analysed samples) (Figure 9F) and contains post-magmatic carbonate. The measured ⁸⁷Sr/⁸⁶Sr data for the alkaline rocks (Taieri River, Tuneheketake, Swinburn North and the three dikes) range from 0.70293 to 0.70425, and are age-corrected to 0.70291 to 0.70424 (Table 3). Dike 3 has anomalously high ⁸⁷Sr/⁸⁶Sr and Sr (1140 ppm. Figure 9F) compared to the other alkaline samples (⁸⁷Sr/⁸⁶Sr ≤ 0.70317, Sr ≤ 755 ppm).

Figure 11. A, Age-corrected ¹⁴³Nd/¹⁴⁴Nd data versus ⁸⁷Sr/⁸⁶Sr data, plotted against the larger Dunedin Volcanic Group, which has two clusters in Pb isotopes. In-situ ⁸⁷Sr/⁸⁶Sr data for Haughton Hill feldspar (n = 14) are shown with a single symbol that is larger than the errors, whereas the Haughton Hill carbonate data (0.7078-0.7079; Table 4) plot off the x-axis to the right. B, Age corrected ²⁰⁷Pb/²⁰⁴Pb and ²⁰⁶Pb/²⁰⁴Pb illustrate the broad differences between the different volcanic rocks. Data in the figures are from this study and the compiled data in Scott et al. (2020a) and from Pontesilli et al. (2021).

Table 4. In-situ ⁸⁷Sr-⁸⁶Sr analyses of feldspar and carbonate in the Haughton Hill flow.

	^87Rb/ ^86Sr	87Sr/86Sr	+/- 2 se
feldspar	0.01	0.70308	29
feldspar	0.01	0.70311	27
feldspar	0.01	0.70314	16
feldspar	0.01	0.70320	18
feldspar	0.00	0.70325	25
feldspar	0.02	0.70327	12
feldspar	0.01	0.70333	11
feldspar	0.01	0.70356	26
feldspar	0.22	0.70312	15
feldspar	0.01	0.70326	12
feldspar	0.02	0.70326	6
feldspar	0.15	0.70332	8
feldspar	0.13	0.70333	13
feldspar	0.05	0.70345	12
carbonate	0.00	0.70780	20
carbonate	0.00	0.70790	13

The ¹⁴³Nd/¹⁴⁴Nd data generally distinguish the two major groups (Figure 11A, Table 3). The transitional group has a ¹⁴³Nd/¹⁴⁴Nd spread of 0.51286 to 0.51292 (ϵNd_{11 Ma} =  +3.9 to +5.5, median of +4.7), which tends to be slightly less radiogenic than most of the alkaline rocks (0.51290 to 0.51293; ϵNd_{11 Ma} =  +5.1 to +5.8, median of +5.4). The Taieri River rocks plot at the intersection of both groups.

Like Nd, the Pb isotopic data largely subdivide the two major groups (Figure 11B, Table 3). The alkaline group has ²⁰⁶Pb/²⁰⁴Pb of 19.64 to 19.96 and ²⁰⁸Pb/²⁰⁴Pb of 39.30 to 39.52, whereas the transitional group has ²⁰⁶Pb/²⁰⁴Pb of 19.21 to 19.61 and ²⁰⁸Pb/²⁰⁴Pb of 38.99 to 39.27. Taieri River once again has values at the intersection of the two groups. All ²⁰⁷Pb/²⁰⁴Pb data plot between 15.64 to 15.67 and do not distinguish the two suites (Figure 11B), except for Swinburn South data, which plot on an array of elevated ²⁰⁷Pb/²⁰⁴Pb for a given ²⁰⁶Pb/²⁰⁴Pb or ²⁰⁸Pb/²⁰⁴Pb value. This array intercepts the other transitional-subalkaline magmas in the Maniototo Basin.

Discussion

Crystal fractionation and crustal contamination

Intraplate monogenetic basalts are often near-primary magmas and therefore can provide insight into mantle source regions (e.g. McGee and Smith 2016). However, prior to investigating any source associations, it is necessary to evaluate potential modification by crystal fractionation or crustal assimilation. Although the studied transitional rocks have MgO contents that range from 5 to 11 wt%, most of the rocks cluster between ∼8 and 11% (Figure 9D). The only systematic variation in MgO with increasing SiO₂ occurs within the Waipiata (Figure 9D) and may be caused by a combination of alteration, crystal fractionation and/or samples with different crystal abundances. The alkaline group also ranges from 8-11 wt% MgO and there is no obvious trend.

Thorium and Nb ratios are often used as guide for crustal assimilation because the crust is high in Th (e.g. Hofmann et al. 1986). All Nb-Th-U data from the volcanic rocks in the Maniototo Basin fall between depleted mantle (Nb/Th = 14 vs Nb/U = 50) and primitive mantle (Nb/Th = 8 vs Nb/U = 30) (Sun and McDonough 1989; McDonough and Sun 1995) (Figure 12A). The lowest values of Nb/Th and Nb/U occur in the transitional lavas. Assuming a depleted mantle source, the variation on display could result from up to ∼1% crustal contamination (Nb/Th has no relationship to LOI); however, given the heterogeneity in other chemical parameters, especially the isotopic data, it is more likely that the range observed is controlled by heterogeneity within the mantle source. The combination of high Nb/Th and Nb/U, as well as a lack of systematic relationship between Ba/Nb and MgO (Figure 12B), suggests there has been little if any crustal contamination. Thus, we interpret that most of the rocks retain the trace element and isotopic character of their parental source-derived magmas. This interpretation has the important implication that the isotopic range displayed should also therefore represent the mantle sources, with the exception of the elevated Sr due to post-magmatic alteration. The isotopic distinction between the alkaline group and transitional group (Figure 11B) also means that these magmas cannot be caused by simply different degrees of melting of the same source.

Figure 12. Element ratio diagrams illustrating the relationships and trends in the Maniototo Basin volcanic rocks discussed in the text. MgO-normalised is assuming anhydrous bulk rock compositions.

Mantle sources for maniototo basin volcanic rocks

The Zealandia sub-continental lithospheric mantle has experienced varying levels of re-enrichment due to metasomatism, most obviously manifested through wide-scale isotopic and chemical changes in mantle peridotite xenolith clinopyroxene compositions via reaction with carbonatitic melts (e.g. Scott et al. 2014a, 2014b; McCoy-West et al. 2015, 2016; Scott et al. 2016a, b; Dalton et al. 2017; van der Meer et al. 2019) or the formation of pyroxenites via reaction with silicate melts (Bonnington et al. 2023). Ti/Eu may be used to investigate the contribution of a carbonatite-metasomatized component to a magma because carbonatites carry abundant REE but little Ti (e.g. Klemme et al. 1995) and the negative correlation between Ti/Eu and La/Yb (Figure 12C) indicates that the strongest ‘carbonatitic’ signature occurs in the alkaline group. However, this interpretation is complicated if phlogopite and amphibole, both Ti-rich phases, were residual in the mantle sources – as will be outlined below.

Metasomatic process may form veins of hydrous minerals within the lithospheric mantle. These veins usually contain amphibole and/or phlogopite (e.g. Foley 1992). Amphibole and phlogopite have been found in peridotite – albeit extremely sparsely – across Otago (e.g. Klemme 2004; Scott et al. 2014b). The occurrence of these minerals is significant because, assuming typical F-free compositions, they are stable at temperature and pressure conditions only within the lithosphere. Phlogopite exists up to ∼1300°C at < 5 GPa (Sato et al. 1997), while F-free pargasitic amphibole, which is more common than K-richterite in mantle xenoliths in alkaline basalts, is stable to ∼1100°C and 2.5 Gpa (e.g. Foley 1992; Green et al. 2014; Pilet et al. 2008). Amphibole and phlogopite can act as residual phases during partial melting, with these phases manifest as a primitive mantle-normalised negative K* and Ti* anomalies (e.g. LaTourrette et al. 1995; Späth et al. 2001; Panter et al. 2006). Alkaline basalt trace element compositions have been experimentally created through the melting of amphibole-rich metasomatic veins, which, when reacted with ambient lherzolite, create the range from high alkali basanite to mid alkaline basalt (e.g. Merrill and Wylie 1975; Pilet et al. 2008). Many of Maniototo volcanic rocks show a negative K*, of which the strongest ne-normative alkaline samples have the lowest K* and lowest Ti* (Figure 12D). The transitional samples subtly show this trend, except for Swinburn South, which shows no variation in Ti* with decreasing K*. These relationship suggest that residual amphibole and/or phlogopite may be controlling these anomalies, with a greater residual volume in the alkaline rocks. Swinburn South evidently had a source with residual mineralogy that retained K but not Ti.

The High Field Strength Elements (HFSE; Nb, Ta, Zr, Hf) are often used to investigate processes of partial melting involving variably metasomatized peridotite. Nb and Ta are preferentially incorporated in Ti-bearing oxide minerals, while major silicate phases usually contain only small abundances and have partition coefficients < 1 for both elements (e.g. Green 1995; Kalfoun et al. 2002). Despite their overall incompatibility, mantle lithosphere minerals such as amphibole and phlogopite can also incorporate considerable Nb and Ta (Ionov and Hofmann 1995; Pfänder et al. 2007). Zr and Hf tend to have greater compatibility in silicate phases compared to Nb and Ta (Green et al. 2000, Pfänder et al. 2007, Tiepolo et al. 2001). When considering the major metasomatic mineral phases (amphibole, phlogopite, clinopyroxene), there are some distinctions in partitioning that can be employed to discriminate the influence of different minerals. Despite some uncertainty, ^cpx/meltD_{Hf or Zr} tend to be one order of magnitude larger than ^cpx/meltD_{Nb or Ta}, ^phlog/meltD_{Hf or Ta} tend to be one order of magnitude larger than ^phlog/meltD_{Nb or Zr}, whereas ^amph/meltD HFSE are all within similar magnitude (Green et al 2000, Pfänder et al. 2007, Tiepolo et al. 2001). Within the more primitive Maniototo rocks, the transitional group has mostly lower Zr/Hf (37-42.5) compared to the alkaline group (42-48) (Figure 12E). Primitive mantle has relatively low Zr/Hf (26-37; Sun and McDonough 1989; McDonough and Sun 1995) and metasomatically modified mantle commonly has higher Zr/Hf as a result of the fluid chemistries (e.g. Bizimis et al. 2003). Considering the respective D values, phlogopite would be an effective residual metasomatic phase to fractionate Zr with respect to Hf, hence affecting their ratio. In this case the high-Zr/Hf alkaline rocks appear to have a greater residual phlogopite influence. In contrast, there is only small variation in Nb/Ta (13-19) in the Maniototo Basin basalts and overall the range is similar for both alkaline and transitional groups. The data are similar to the average of OIB (Pfänder et al. 2007), although it is clear that Swinburn South and Swinburn North extend to slightly lower values than Tuneheketake and the dikes. A greater distinction exists in Zr/Ta (or likewise Hf/Nb) (Figure 12E) and either residual clinopyroxene or amphibole would be the most effective phase in fractionating Zr (and Hf) with respect to Ta (and Nb). Again, the alkaline group having lower Zr/Ta suggests a role for residual metasomatic mantle minerals during magma generation or modification processes. Swinburn South is distinct as it has both low Zr/Hf and Zr/Ta (Figure 12E) in addition to the anomalous K* but not Ti* outlined above. Although the latter characteristics might indicate a phlogopite influence, the relatively low Zr/Ta and Zr/Hf points to a lack thereof. Alternatively, phlogopite may occur in conjunction with metasomatic amphibole and clinopyroxene that would complicate/reverse chemical signatures. Overall, however, on the basis of their trace element chemistry, the Maniototo Basin alkaline rocks were therefore likely derived from, or interacted with, a mantle that contained a greater metasomatic component than the Maniototo Basin transitional lavas.

Zones of amphibole and/or phlogopite-rich mantle are known as metasomes (Rooney et al. 2017) and these have been proposed as possible melt sources or components in the formation of alkaline lavas both within the Dunedin Volcanic Group (e.g. Price et al. 2003; Sprung et al. 2007; Scott et al. 2020a; Pontesilli et al. 2021). However, the combination of the difference in trace element anomalies between alkaline and transitional lavas (e.g. negative Zr-Hf anomaly) and the systematic isotopic placement of the Maniototo Basin transitional group largely outside the known Otago lithospheric mantle range (Figure 11A and B) suggests that there is a source component that is only tapped or clearly represented in the latter melts. Pontesilli et al (2021) proposed that the range of high to low alkaline basalts of Dunedin Volcano could have been generated by asthenospheric melts (low-alkaline) interacting with lithospheric metasomes, with the result being transformation towards high-alkaline magma (e.g. Foley 1992; Rooney et al. 2017). In this case, the higher ⁸⁷Sr/⁸⁶Sr and lower ¹⁴³Nd/¹⁴⁴Nd in the low-alkaline rocks of the Maniototo Basin would likely be from components within the convecting mantle, with this composition diluted and overprinted depending on the volume of reaction with the lithospheric mantle metasomes to give rise to increasingly alkaline rocks (e.g. Pontesilli et al. (2021)). The origin of the intermediate composition Taieri River flow remains unclear; however, an alkaline melt reacting with peridotite would raise the SiO₂ via orthopyroxene dissolution (Pilet et al. 2008) yet likely retain the alkaline isotopic ratios.

Conclusions

A range of alkaline and subalkaline Miocene-erupted volcanic rocks occur in the Maniototo Basin in the northwest of the Dunedin Volcanic Group. We present the first comprehensive description of their occurrences, mineralogy, geochemistry and Sr-Nd-Pb isotopes as well as two ⁴⁰Ar^/39Ar dates. None of the rocks show significant crustal contamination, although some show post-magmatic modification of Sr isotope compositions and there is evidence of some post-depositional alteration. The alkaline magmas are trace element enriched compared to the transitional (alkaline to subalkaline) group and also have larger Ti and K negative anomalies indicative of residual amphibole and/or phlogopite in one of the mantle sources. The transitional group of magmas show less effect of amphibole and phlogopite. Since these two hydrous phases are present only in the lithosphere (unless as high-T K-richterite), an explanation for the variation in the Maniototo Basin volcanic rocks is that the melts represent different degrees of mantle lithospheric modification as magmas ascended, or the melts were derived from different mantle sources. In either case, the variability appears to be due to mantle-related processes and heterogeneity.

Acknowledgments

Field and analytical costs of the MSc of Wilson and the PhD of Giacalone were covered by the University of Otago Geology department. Giacalone was supported by an Otago Doctoral Scholarship. We thank all the landowners for permission to cross their land. Two reviewers (anonymous) and the editor (H. Dalton) are thanked for their comments.

Disclosure statement

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

Data availability statement

All geochemical and geochronological data used in this paper are openly available in the text or in the supplemental files deposited in figshare at https://doi.org/10.6084/m9.figshare.24427378.v2.