Journal of Volcanology and Geothermal Research 421 (2022) 107430 Contents lists available at ScienceDirect Journal of Volcanology and Geothermal Research j ourna l homepage: www.e lsev ie r .com/ locate / jvo lgeores Inter- and intra-crystal quartz δ18O homogeneity at Okataina volcano, Aotearoa New Zealand: Implications for rhyolite genesis May Sas a,⁎, Phil Shane b, Noriyuki Kawasaki c, Naoya Sakamoto d, Georg F. Zellmer e, Hisayoshi Yurimoto c,d a Geology Department, Western Washington University, Bellingham, Washington 98225, USA b School of Environment, University of Auckland, Auckland 1142, New Zealand c Natural History Sciences, Hokkaido University, Sapporo 060-0810, Japan d Isotope Imaging Laboratory, Creative Research Institution, Hokkaido University, Sapporo 001-0021, Japan e Volcanic Risk Solutions, Massey University, Private Bag 11222, Palmerston North 4442, New Zealand ⁎ Corresponding author. E-mail address: sasm@wwu.edu (M. Sas). https://doi.org/10.1016/j.jvolgeores.2021.107430 0377-0273/© 2022 The Authors. Published by Elsevier B.V a b s t r a c t a r t i c l e i n f o Article history: Received 24 June 2021 Received in revised form 11 November 2021 Accepted 16 November 2021 Available online 20 November 2021 The sources and processes involved in the genesis of the voluminous rhyolitic magmas of cataclysmic caldera- forming eruptions, and the intervening lower-volume intra-caldera extrusions, have been subject to much de- bate. To better understand generation of high-volume and low-volume silicic eruptions within a single volcanic centre, and how they may differ, we examined ten volumetrically varied high-SiO2 rhyolite eruptions from the Okataina Volcanic Centre (OVC) in Aotearoa New Zealand. The OVC is one of the world's most recurrently active silicic volcanoes. In the last ~600 ky, theOVCwas the focus of three known caldera-forming events and numerous intermittent dome-building and fissure eruption episodes, with rhyolitic eruption activity as recent as 1314 CE. To elucidate how mass contributions from the mantle and crust may have fluctuated over the lifespan of the OVC magmatic system, oxygen isotopic ratios (δ18O) of quartz in rhyolites were investigated for the first time at inter-crystal and intra-crystal scales. Quartz crystals from four eruption episodes (two caldera-forming events, Utu, ~557 ka, Rotoiti, ~45 ka, and two intra-caldera dome-building events, Rotoma, ~9.5 ka, and Kaharoa, ~0.7 ka) yielded intra-crystal δ18O isotopic homogeneity (±0.23‰, 2sd) based on secondary ion mass spectrometry (SIMS). These samples also display inter-crystal and inter-unit homogeneity within slightly lower precision (7.6 ± 0.5‰, 2sd). Whole-crystal quartz from the same four units, as well as six other units (two intra-caldera dome-building episodes, Okareka, ~21.8 ka, Whakatane, ~5.5 ka, three pre-Rotoiti extra-caldera domes, Round Hill, Haparangi, Kakapiko, and one immediately post-Rotoiti eruption, Earthquake Flat), were then examined using high-precision laser fluorination. Single crystals also yielded mostly homogenous ratios with average δ18O = 7.6 ± 0.5‰ (2sd), which is consistent with intra-crystal SIMS analyses, albeit for a larger set of samples. Stable and radiogenic isotope mixing models using the newly obtained δ18O ratios demonstrate that OVC rhyolites can be produced by ≥25% assimilation of a regional (Torlesse-like) metasedimentary endmember by a depletedmantle sourcewith slightly variable amounts of subduction flux, and that any incorporation of hydro- thermally alteredmaterial to the system is limited to<5% in caldera and intra-caldera eruptions. The δ18O records of the OVC are among the most homogenous currently known and indicate stable and consistent mantle and crustal contributions across the lifespan of the magmatic system, with assimilation largely occurring prior to segregation of rhyolitic melts within the silicic reservoir. This isotopic homogeneity may be due to a relatively high-volume and constant magma flux at the OVC, which contrasts to other rhyolitic caldera volcanoes with greater isotopic variability. © 2022 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http:// creativecommons.org/licenses/by-nc-nd/4.0/). Keywords: Stable isotopes Caldera Petrogenesis Rhyolite Quartz 1. Introduction In addition to fractional crystallization, incorporation of crustal mass is thought to be significant in rhyolite genesis,with some studies suggest- ing asmuch as 25–30% assimilation of crustalmaterials (e.g., Long Valley, . This is an open access article under Bindeman and Valley, 2002; Maroa, McCulloch et al., 1994). These estimates of crustal versus mantle contributions are based on com- positional and isotopic models, the latter of which commonly act as evi- dence for assimilation processes. Major shifts (e.g., >1‰) in oxygen isotopic ratios (δ18O) of rocks and phenocrysts associated with caldera collapse at rhyolite volcanoes have been noted globally (δ18O ‰ = [Rsample/Rstandard – 1] × 1000, R = 18O/16O; standard is Vienna Standard Mean OceanWater, VSMOW; Baertschi, 1976). Since oxygen the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). http://crossmark.crossref.org/dialog/?doi=10.1016/j.jvolgeores.2021.107430&domain=pdf http://creativecommons.org/licenses/by-nc-nd/4.0/ http://creativecommons.org/licenses/by-nc-nd/4.0/ https://doi.org/10.1016/j.jvolgeores.2021.107430 mailto:sasm@wwu.edu https://doi.org/10.1016/j.jvolgeores.2021.107430 http://creativecommons.org/licenses/by-nc-nd/4.0/ http://www.sciencedirect.com/science/journal/ www.elsevier.com/locate/jvolgeores M. Sas, P. Shane, N. Kawasaki et al. Journal of Volcanology and Geothermal Research 421 (2022) 107430 is the most abundant element in magmatic systems, even minor fluctuations (sub-per mil) in δ18O can amount to large mass transformations resulting from crustal assimilation or mafic influx (Bindeman, 2008). The significantdrops (i.e., ~6‰) inmagmatic δ18O ra- tios following caldera formationhavebeen interpreted tobe the result of large-scale incorporation of low-δ18O hydrothermally-altered volcanic rock into the magmatic system (e.g., Yellowstone; Hildreth et al., 1984; Bindeman and Valley, 2000, 2001; Bindeman et al., 2007). Similarly, some calderas in extensional or rift settings exhibit changes in δ18O be- tween different caldera-collapse events, which are thought to result from mafic influx and re-melting of isotopically variable roof material (e.g., Southwestern Nevada Volcanic Field; Bindeman and Valley, 2003). Furthermore, in continental arc settings, intra-crystal zoning of δ18O ratios in quartz have been attributed to assimilation of altered crust shortly prior to a catastrophic caldera eruption (e.g., Toba volcano, Sumatra; Budd et al., 2017). The present study explores δ18O ratios in quartz phenocrysts from rhyolites from the Okataina Volcanic Centre (OVC), a caldera volcano in Aotearoa New Zealand. This magmatic sys- tem has been frequently active on a millennial-scale and is well- studied (e.g., Nairn, 2002; Smith et al., 2005, 2010; Shane et al., 2008a, 2008b). Unlike the volcanoes mentioned above, it is situated in an ac- tively rifting arc where the continental crust is thin (~25 km; Bannister et al., 2004). Quartz was selected as it is ubiquitous in high-SiO2 rhyolites from the OVC, and because it is particularly useful for mineral-specific isoto- pic studies as it commonly abundant, coarse-grained, and resilient to secondary isotope exchange (Bindeman, 2008). At the OVC, quartz and other phenocrysts have revealed considerable insight to the petro- genesis of magmas. The previous examination of zoning, trace element compositions, and melt inclusions in OVC quartz phenocrysts suggest mingling of discrete parental melts as well as rapid crystallization (Shane et al., 2008b; Smith et al., 2010; Matthews et al., 2012). Several zircon U–Th disequilibrium and trace elements studies (e.g., Storm et al., 2011, 2012, 2014; Rubin et al., 2017) identify crystal populations with differing compositional and thermal histories that indicate the presence of long-lived, structurally disconnected magmatic reservoirs. Similarly, the accompanying amphibole population requires repeated remobilization of a thermally zoned system with minimal lateral con- nectivity (Shane and Smith, 2013). Investigation of oxygen fugacity (fO2) from Fe–Ti oxides and orthopyroxene compositions called on the presence of cold and wet, as well as hot and dry, rhyolite magmas (Deering et al., 2010). Plagioclase phenocryst cores were found to be antecrystic and nucleate in crustally-contaminated melts of variable compositions (Shane, 2015; Sas et al., 2021). All of these studies indicate the presence of separate but contemporaneousmelt bodies in a volumi- nous crystal mush with restricted lateral connectivity. Furthermore, re- cent studies examining intra- and whole-crystal Sr–Pb isotopic ratios in plagioclase from OVC rhyolites reveal fairly consistent isotopic signa- tures over the last ~0.6 My (Sas et al., 2019, 2021). To examine how magmatic records in a later-crystalizing phase compare to those in pla- gioclase, and to elucidate the effects of open system processes through the lifetime of this dynamic magmatic system, textures and stable oxy- gen isotopic ratios of quartz were investigated. Albeit a well-studied region, there are few δ18O-based studies at the OVC and adjacent silicic caldera volcanoes (e.g., plagioclase separates from Maroa Volcanic Centre rhyolites, McCulloch et al., 1994). At the OVC, although Blattner and Reid (1982) and Blattner et al. (1996) provide δ18O ratios of quartz from some rhyolites, the analyses are of bulk samples (many crystals) and are limited to younger units. This study presents the first single crystal- and intra-crystal- scale δ18O data in OVC quartz, using secondary ion mass spectrometry and laser fluorination techniques. The sample suite was selected to characterize potential temporal and spatial δ18O ratio variations at the volcanic centre. Unlike at other caldera centers, this work identifies a well-balanced system at the OVC, with consistent inputs from the mantle and crust, and constant mechanisms of crustal assimilation across the duration of the magmatic system. 2 2. Geologic Background The OVC has been active for ~600 ka, erupting dominantly rhyolitic magmas from a locus that has had at least three caldera-collapse events (Nairn, 2002; Cole et al., 2014). The volcanic center is situated within the northeastern segment of the Taupō Volcanic Zone (TVZ) – an actively-rifting, 2million-year-old volcanic arc that is located in the cen- tral North Island, Aotearoa New Zealand (Fig. 1) (Wilson et al., 1995). The TVZ is the 200 km subaerial terminus of the ~3000 km long Tonga-Kermadec volcanic arc and is a result of the westward sub- duction of the Pacific Plate beneath the Australian Plate (Wilson et al., 1995). The rates of subduction are reduced from northeast off- shore North Island, 58 mm/yr, to southwest offshore North Island, 19 mm/yr, as subduction becomes more oblique (Wallace et al., 2009). In addition to subduction-related magma generation, magma produc- tion in the TVZ is also impacted by extension-induced crustal thinning (~25 km crystal thickness, Bannister et al., 2004) and concomitant decompression melting of the underlying mantle. The coupled subduction-extension setting results in high rates of magma emplace- ment within the crust (Stern and Benson, 2011; Cole et al., 2014). The extension rate at the OVC is ~12 mm/yr (Wallace et al., 2004). The uppermost crust in the TVZ, down to ~3 km depth at the OVC, consists of Pleistocene volcanics and volcaniclastics (Cole et al., 2010, 2014; Seebeck et al., 2010). Underlaying these volcanic rocks, in the middle to upper crust, are interbedded and structurally complex sand- stones and mudstones (collectively referred to as greywacke in litera- ture) that have been weakly metamorphosed, albeit higher-grade xenoliths in southern TVZ lavas suggest increased metamorphism in the middle crust (Price et al., 2012, 2015; Milicich et al., 2021). The metasediments comprise two Mesozoic, arc-sourced terranes, the Torlesse Composite Terrane (Jurassic-Cretaceous) to the east and the Waipapa Composite Terrane (Permian-Jurassic) to the west (Fig. 1; Spörli, 1978; Price et al., 2015). The Torlesse Terrane constitutes some- what equal portions of lithics, quartz and feldspar, with high average SiO2 contents of ~70 wt% (Reid, 1983; Price et al., 2015). The Waipapa Terrane contains a larger portion of volcanic lithics relative to other components, thereby resulting in lower average SiO2 contents of ~63 wt% (Reid, 1983; Price et al., 2015). The boundary between the two terranes in debated; Price et al. (2015) suggest an oblique boundary based on the presences of Torlesse xenoliths in dacite lava north of the Taupō Volcanic Centre. Conversely, Milicich et al. (2021) suggest a steeply dipping boundary that crosses through the southern region of the OVC, although uncertainty of the boundary location ex- tends approximately 20 km northwest of the OVC. Notably, Milicich et al. (2021) propose that Quaternary extension has reactivated the boundary between the twometaseimentary terranes, and likely heavily impacted the location of the TVZ. In contrast to middle to upper crust compositions, lower crust compositions below the OVC, and the TVZ in general, are poorly constrained. However, metaigneous xenoliths found in andesite lavas from the southernmost TVZ suggest lower crust rocks are of oceanic crust origin (Graham et al., 1990; Price et al., 2012). The coupled extension-subduction setting of the TVZ results in strong NE-SW lineaments of structural (e.g., faults and grabens) and volcanic (e.g., fissure eruptions, caldera boundaries) features (Nairn, 2002; Seebeck et al., 2010; Cole et al., 2014). At the OVC, this is apparent by the formation of two parallel vent alignments that are oriented NE- SW, Tarawera to the south and Haroharo to the north (Fig. 1), which formed following the most recent caldera-collapse event at ~45 ka (Cole et al., 2010; Danišík et al., 2012). Tarawera and Haroharo are both composed of predominantly rhyolitic deposits (domes and pyro- clastics) that have formed over several eruption periods (Nairn, 2002). Although there are minor mineralogical differences across the OVC (e.g., cummingtonite is the dominant ferromagnesian phase in Haroharo deposits, but is scarce in Tarawera deposits), the majority of deposits are high-SiO2 rhyolites with cognate plagioclase + quartz Fig. 1. Map of the Okataina Volcanic Centre (OVC) showing sample locations (circles), major vent locations (triangles), caldera boundaries (thin solid lines), and intra-caldera volcanic complexes and extra-caldera domes (shaded regions). Sample locations are also included in Table 1. The location of the OVC within the Taupō Volcanic Zone (TVZ) and the two regional metasedimentary formations, Torlesse and Waipapa, (Price et al., 2015; Milicich et al., 2021) are shown in the inset map. Okataina inferred caldera boundaries, ring structure (dashed line), dome complexes, and grabens (green lines) are from Nairn (2002). Major vent locations are from Nairn (2002), Nairn et al. (2004), Smith et al. (2006), and Shane et al. (2008a). Locations of basalt samples used for modelling (stars, inset map) are from Graham et al. (1992) and Macpherson et al. (1998). M. Sas, P. Shane, N. Kawasaki et al. Journal of Volcanology and Geothermal Research 421 (2022) 107430 >> Fe-Ti oxides + amphibole + orthopyroxene ± biotite ± clinopyroxene (Nairn, 2002; Smith et al., 2005). This study focuses on ten representative rhyolitic units from seven spatially, temporally and volumetrically varied eruptive episodes that cover the span of eruptive activity at the OVC and include caldera- forming eruptions, as well as intra-caldera and extra-caldera eruption episodes (Table 1, Fig. 1). The caldera-forming eruptions include Utu, the oldest known collapse event (~557 ka; Leonard et al., 2010), and Rotoiti, the most recent significant collapse event (~45 ka; Nairn, 2002; Shane et al., 2005). Intra-caldera small volume eruptions consist of Rotoma (9.5 ka) andWhakatane (5.5 ka) fromHaroharo, andOkareka (21.8 ka) and Kaharoa (0.7 ka) from Tarawera (Nairn, 2002; Smith et al., 2005, 2006). Extra-caldera small volume eruptions consist of the Round Hill, Haparangi, and Kakapiko domes, which are part of the Haparangi- Kapenga dome complex (~200–100 ka) that preceded Rotoiti (Nairn, 2002). One additional unit, Earthquake Flat, erupted weeks to months post-Rotoiti, with vents along the southwestern edge of the OVC ring structure (Nairn and Kohn, 1973). Earthquake Flat and the extra- caldera domes contain coarser crystals (≤4 mm) and are also more crystal-rich (20%–40%) than intra-caldera deposits (generally ≤2 mm crystals and ≤15% crystals), and may represent remobilized portions of a mostly-crystalline reservoir (Molloy et al., 2008; Sas et al., 2021). With the exception of a few dacitic Earthquake Flat clasts, all units fall in the rhyolite field on a total alkali silica diagram (Fig. 2). In units where detailed stratigraphy is available, stratigraphic units selected from each deposit represent the most volumetrically significant unit and/or represent hybrid magmatic products that consist of mingled or 3 mixed rhyolitic melts (Table 1 and references therein). The studied units are summarized in Table 1 and sample locations are shown in Fig. 1. 3. Methods 3.1. Sample preparation The samples studied here are all previously collected (Table 1 and references therein) non-altered juvenile tephras or non-welded, juve- nile ignimbrites whose loose nature requiredminimal pressure to liber- ate the crystals (Sas et al., 2021). Following liberation, several hundred crystals, which exhibited a bipyramidal habit and were >100 μm in size, were individually picked from each unit using a binocular micro- scope. Any crystals exhibiting surface alteration were avoided. For in-situ secondary ion mass spectrometry (SIMS) analyses, untreated crystals were mounted onto a single 1-in. diameter epoxy plug, with standard crystals (NBS28) imbedded in the center of the plug. Crystals were grinded to expose crystal cores, and a sub-micron polish was ap- plied to the plug. For whole-crystal laser fluorination (LF) analyses, non-weathered, inclusion-free quartz crystals were selected using a high-magnification binocular microscope where possible. For units where inclusions were unavoidable (Rotoiti and Kaharoa; Smith et al., 2010), a low total inclusion volume was achieved based on high- magnification optical examination and measurement of melt inclusions using a microscale. Specifically, for Kaharoa, <1 vol% was obtained due to the paucity and size of the inclusions in individual crystals, and for Ta bl e 1 Su m m ar y of rh yo lit ic O V C un it s. Er up ti on K ah ar oa W ha ka ta ne Ro to m a O ka re ka Ea rt hq ua ke Fl at Ro to it i K ak ap ik o H ap ar an gi Ro un d H ill U tu Er up ti on st yl e D om e D om e D om e D om e D om e Ca ld er a D om e D om e D om e Ca ld er a A ge (k a) 0. 7 5. 5 9. 5 21 .8 ~4 5 45 20 0– 45 20 0 20 0– 45 55 7 V ol um e (k m 3 )a 9. 1 11 .3 8 5 10 12 0 – – – 90 St ra t un it b T2 (1 ) W T1 (7 ) RT 2( 7, 9) T1 (1 0) – T1 (1 5) – – – – Si O 2 c 76 .4 4 76 .6 0 76 .3 8 75 .9 8 72 .2 4 73 .6 6 73 .9 4 73 .6 8 74 .1 5 73 .4 4 T (° C) d 72 4 ± 16 74 5 ± 11 76 0 ± 9 79 4 ± 12 75 4 ± 52 76 6 ± 13 72 2 ± 13 71 8 ± 15 72 4 ± 13 76 2 ± 8 P (k ba r) e 0. 6– 2. 5 0. 7– 2. 8 0. 7– 1. 9 0. 8– 1. 8 0. 7– 1. 4 0. 6– 2. 5 – – – – fO 2 (N N O )f + 0. 00 ± 0. 54 + 0. 34 ± 0. 08 + 0. 62 ± 0. 06 + 0. 82 ± 0. 08 − 0. 26 to + 0. 32 + 0. 92 ± 0. 09 − 0. 19 ± 0. 10 − 0. 47 ± 0. 12 − 0. 28 ± 0. 11 − 0. 52 ± 0. 10 H 2 O g 5. 18 Q m i 4. 48 Q m i 5. 26 Q m i 5. 12 Q m i 5. 04 P m i 4. 99 Q m i 3. 97 g 5. 07 g 3. 87 g – M in er al as se m bl ag eh pl g > qt z> > bi o > Fe -T i ± hb l, op x, cg t qt z > pl ag > > hb l > cg t> Fe -T i > op x pl g > qt z> > Fe -T i > op x > cg t pl g > qt z> > op x + hb l + cg t + Fe -T i> > bi o pl g > qt z> > bi o > hb l> Fe -T i ± op x pl g > qt z> > cg t > Fe -T i > hb l > op x qt z > pl g> > bi o > hb l > op x + Fe -T i qt z > pl g> > bi o> > hb l> op x + Fe -T i qt z > pl g> > bi o > hb l > op x + Fe -T i qt z/ pl g> bi o> > hb l> op x + Fe -T i δ1 8 O an al ys es i SI M S (n = 7) , LF (n = 2) LF (n = 3) SI M S (n = 7) ,L F (n = 2) LF (n = 3) LF (n = 2) SI M S (n = 8) ,L F (n = 2) LF (n = 3) LF (n = 8) LF (n = 3) SI M S (n = 7) , LF (n = 2) Lo ca ti on j − 38 .2 73 ,1 76 .5 15 − 38 .1 40 , 17 6. 51 5 − 38 .0 17 ,1 76 .5 90 − 38 .1 85 ,1 76 .5 46 − 38 .2 99 , 17 6. 27 7 − 37 .9 22 ,1 76 .3 42 − 38 .2 33 ,1 76 .2 05 − 38 .2 07 , 17 6. 22 3 − 38 .2 60 ,1 76 .2 39 − 38 .2 04 , 17 6. 66 0 Re fe re nc es k 1– 4 5– 8 7– 9 6, 8, 10 11 –1 4 9, 15 –1 7 6, 18 18 18 6, 19 –2 2 a V ol um es ex tr ac te d fr om lit er at ur e ar e de ns e ro ck eq ui va le nt (D RE ). b St ra ti gr ap hi c un it fr om w hi ch cr ys ta ls w er e ex tr ac te d (w he re av ai la bl e) ar e ba se d on su bd iv is io ns es ta bl is he d by th e re sp ec ti ve st ud y, w hi ch is in cl ud ed as a nu m be r in pa re nt he se s fo llo w in g th e un it an d is lis te d be lo w in Re fe re nc es k . St ra t- ig ra ph ic un it s se le ct ed ar e th e m os t vo lu m et ri ca lly si gn ifi ca nt an d/ or hy br id m ag m at ic pr od uc ts . c A ve ra ge w ho le ro ck Si O 2 co nt en ts ex tr ac te d fr om lit er at ur e. d -f Re pr es en ta ti ve un it Fe –T io xi de s te m pe ra tu re s (± 2s d) an d fO 2 (± 2s d) ex tr ac te d fr om lit er at ur e, an d am ph ib ol e pr es su re ra ng es fr om Sh an e an d Sm it h (2 01 3) . g H 2 O co nt en ts ex tr ac te d fr om lit er at ur e. Q m i: H 2 O fr om qu ar tz -h os te d m el t in cl us io ns ,P m i: H 2 O fr om pl ag io cl as e- ho st ed m el ti nc lu si on s, g: es ti m at ed H 2O fr om gl as s. h M in er al ab br ev ia ti on s ar e: pl g: pl ag io cl as e, qt z: qu ar tz ,b io :b io ti te ,h bl :h or nb le nd e, cg t: cu m m in gt on ite ,o px :o rt ho py ro xe ne ,F e- Ti :F e– Ti ox id es . i A na ly se s in cl ud e in tr a- cr ys ta ls ec on da ry io n m as s sp ec tr om et ry (S IM S) an d w ho le -c ry st al la se r flu or in at io n (L F) ;n is th e to ta ln um be r of cr ys ta ls an al yz ed pe r m et ho d. j Lo ca ti on s of sa m pl es ar e pr ov id ed in de ci m al de gr ee s. k 1: N ai rn et al .( 20 04 ); 2: Le on ar d et al .( 20 02 ); 3: Sa he ta py -E ng el et al .( 20 14 ); 4: H og g et al .( 20 03 ); 5: K ob ay as hi et al .( 20 05 ); 6: N ai rn (2 00 2) ;7 :S m it h et al .( 20 06 ); 8: Lo w e et al .( 20 08 ); 9: Sm it h et al .( 20 05 ); 10 :S ha ne et al .( 20 08 a) ;1 1: M ol lo y et al .( 20 08 ); 12 :N ai rn an d K oh n (1 97 3) ;1 3: Fr og ga tt an d Lo w e (1 99 0) ;1 4: D av is (1 98 5) ;1 5: Sh an e et al .( 20 05 ); 16 :S ch m it z an d Sm it h (2 00 4) ;1 7: D an iší k et al .( 20 12 ); 18 :S as et al .( 20 21 ); 19 :C ol e et al .( 20 10 ); 20 :C ol e et al .( 20 14 ); 21 : Le on ar d et al .( 20 10 ); 22 :D ee ri ng et al .( 20 10 ). M. Sas, P. Shane, N. Kawasaki et al. Journal of Volcanology and Geothermal Research 421 (2022) 107430 4 Fig. 2. Total alkali silica (TAS) diagram of OVC rhyolite units included in this study (whole rock; Leonard et al., 2002; Nairn, 2002; Schmitz and Smith, 2004; Smith et al., 2006;Molloy et al., 2008; Shane et al., 2008a; Deering et al., 2008; Shane and Smith, 2013; Sas et al., 2021), basalt samples used formodelling (Havre Trough and Rumble IV, whole rock; Gamble et al., 1994), a representative OVC basalt (Tarawera, whole rock; Gamble et al., 1993; Nairn, 2002; Nairn et al., 2004; Hiess et al., 2007; Zellmer et al., 2020), as well as a hydrothermally altered granitoid found in OVC rhyolite (TW1, whole rock; Brown et al., 1998). Rhyolite samples from which quartz crystals were extracted have enlarged symbols and a thick, black outline. M. Sas, P. Shane, N. Kawasaki et al. Journal of Volcanology and Geothermal Research 421 (2022) 107430 Rotoiti, <2 vol% was achieved. Since Rotoiti crystals contain fresh, unal- tered (sealed) melt inclusions (Smith et al., 2010), their contribution to δ18O is considered to be insignificant (Bindeman and Valley, 2002). In order to remove any attached glass or unsealed inclusions, quartz crys- tals analyzed using LF were first purified by etching in 30% hydrofluoric acid for 5 min, following the technique of Bindeman and Valley (2002). A summary of units analyzed, methods employed, and total number of crystals per unit are included in Table 1. 3.2. Cathodoluminescence imaging and in situ oxygen isotope analysis Cathodoluminescence (CL) images of quartz crystals were obtained using a Gatan Mini-CL installed on a field emission scanning electron microscope (FE-SEM; JEOL JSM-7000F at Hokkaido University). Be- tween 25 and 30 crystals per unit were imaged using SEM-CL, and brightness and contrast settings were maintained across all images for comparison of textures. A 15 kV accelerating voltage and a current of 10 nA was employed in our study. The polished section was coated with a carbon thin film (~20 nm). Analyses of δ18O zones in individual quartz crystals from four pilot units were obtained by in situO-isotopemeasurements using a SIMS in- strument (CAMECA IMS-1280HR at Hokkaido University). Units se- lected for SIMS analyses consist of two large-volume eruptions, Utu and Rotoiti, and two small-volume intra-caldera eruptions, Rotoma and Kaharoa (Table 1). Quartz textures were first examined using SEM-CL imaging. Then, ideal (i.e., complete sections from core to rim where possible) and 7–8 representative crystals were selected for SIMS analyses, where zones with differing CL intensities (brightness) were analyzed. Before measurements, the polished section was coated with a gold thin film (~70 nm). The SIMS measurements were performed during two different ses- sions with different analytical precisions. Procedures essentially followed those described in Kawasaki et al. (2018). For one set of 5 analyses (low-precision mode), a 133Cs+ primary beam (20 keV, 1 nA) with a diameter of ~6 μm was used. Negative secondary ions, 16O− and 18O−, were measured simultaneously in multi-collection mode using two Faraday cups. A normal incident electron flood gun was used for electrostatic charge compensation of the analyzed area during the analyses. The mass resolution of M/ΔM was set at ~2200. The sec- ondary ion intensity of 16O− was typically ~1.0 × 109 cps. Before the analyses, the analyzed area was sputtered with a rastered beam of ~11 μm for 30 s. Each analysis consisted of 5 cycles of collecting the sec- ondary ions for 3 s. NBS28 (δ18O = 9.29‰, Kusakabe and Matsuhisa, 2008), imbedded in the center of the sample plug, was used as a stan- dard for correcting instrumental mass fractionation (IMF). Because of some heterogeneity of δ18O among grains of the NBS28 standard (e.g., Appleby et al., 2008), we measured multiple grains of the NBS28 standard for every bracketing (to account for this heterogeneity). Every analytical cycle consisted of 10 analyses of NBS28, 10 analyses of samples, and 10 analyses of NBS28. The data were then corrected using a standard bracketing method using NBS28 and normalizing to δ18O = 9.29‰ (Kusakabe and Matsuhisa, 2008). Uncertainty of the IMF correction was 0.15–0.27‰. Analytical errors were typically ~0.43‰ 2σ in δ18O, including a statistical error of individual analysis cal- culated from counting statistics of total ion counts and the uncertainty of the IMF correction. The reproducibility of repetitive analyses of a sin- gle grain of the NBS28 standard was 0.54‰ (2 standard deviation, n = 10). Analyzed spots were carefully observed by FE-SEM. For another set of the analyses (high-precision mode), a 133Cs+ pri- mary beam (20 keV, 2 nA) with a diameter of ~9 μmwas used. The sec- ondary ion intensity of 16O− was typically ~2.0 × 109 cps. Before the analyses, the analyzed area was sputtered with a rastered beam of ~14 μmfor 30 s. Each analysis consisted of 10 cycles of collecting the sec- ondary ions for 10 s. Other settings were identical to those of the low- precision mode. The IMF correction with bracketing analyses of the NBS28 standard could not be readily applied for the high-precision M. Sas, P. Shane, N. Kawasaki et al. Journal of Volcanology and Geothermal Research 421 (2022) 107430 mode, because both the heterogeneity of NBS28 and background drift of Faraday cup detectors were significant in this mode. Averages of δ18O ratio for each sample analyzed in high-precisionmodewere normalized to averages of δ18O ratios for the same sample analyzed in low-precision mode, corrected by multiple bracketing. We note that the high- precision mode was applied to reveal the degree of homogeneity/het- erogeneity of samples with an improved reproducibility. Analytical er- rors were typically ~0.16‰ (2σ) in δ18O, including a statistical error of individual analysis calculated from counting statistics of total ion counts and the uncertainty of the normalization. The reproducibility of repeti- tive analyses of a single grain of the NBS28 standard was 0.23‰ (2sd, n = 10). 3.3. Laser fluorination (LF) analytical procedure Quartz phenocrysts from all of the samples (Table 1) were analyzed using LF (Sharp, 1990). Aliquots of quartzweighing between 1 and 2mg (dominantly representing a single crystal) were analyzed at the Stable Isotope Laboratory, University of Oregon, following the CO2-LFmethods of Bindeman (2008). Samples were fluorinated using BrF5 and a CO2 laser. Following reaction, O2 gas was cryogenically purified, cleaned of excess F2 using liquid Hg vapors, and converted to CO2, which was analyzed for oxygen isotopic compositions using a stable isotope ratio mass spectrometer (Thermo FinniganMAT253). A University of Oregon garnet (UOG) with δ18O= 6.52‰ VSMOW (Troll et al., 2013) was used Fig. 3. Cathodoluminescence (CL) images showing high-precision secondary ionmass spectrom Rotoma, (e–f) Rotoiti, and (g–h) Utu. The error bars represent the respective 2σ for each analys for each unit are shown in Supplementary Figs. A1–A4. Quartz crystals imaged using CL and an 6 as a standard to correct unknowns to VSMOW scale (corrections were 0.10–0.20‰) with 0.11–0.13‰ (2sd) external reproducibility across the analytical sessions. 4. Results 4.1. Quartz textures Approximately 100 crystals from the four samples (Kaharoa, Rotoma, Rotoiti and Utu; Table 1) examined via SEM-CL imaging exhibit similar textures that are typical of magmatic quartz (e.g., Wilcock et al., 2013). All of the quartz crystals displayweak oscillatory zoning of bright (high CL intensity) and dark (low CL intensity) bands that range from 5 to 100 μm inwidth (Fig. 3; Supplementary Fig. A1–A4). One exception is a single Utu crystal that exhibits diffuse boundaries between zones (Fig. 3h). Quartz crystals can be divided into three broad categories based on variations in CL brightness between cores and rims: (1) subtle differences in brightness between the core and rim (Q1; e.g., Fig. 3c); (2) notably brighter cores with darker rims (Q2; e.g., Fig. 3b); and (3) notably darker cores with brighter rims (Q3; e.g., Fig. 3e). The boundaries between quartz core/mantle and rim zones are consistently resorbed (i.e., truncated growth patterns). These textural groups are found in the Kaharoa, Rotoma, Rotoiti and Utu samples (Table 1; Sup- plementary Fig. A1–A4), with Q1 being the most common (~40%), and Q2 and Q3 having roughly equal distributions (~30% each). The same etry (SIMS) δ18O analyses of representative OVC quartz crystals from (a–b) Kaharoa, (c–d) is. SIMS δ18O analyses and their respective 2σ are listed in Table 2, and all crystals analyzed alyzed using SIMS were not modified or treated before being embedded in epoxy. Table 2 Intra-crystal MC-SIMS δ18O (‰) analyses of OVC quartz. Unit/Crystala distanceb δ18O 2σc Unit/Crystala distanceb δ18O 2σc OVC rhyolites Kaharoa RTQ-2 56 7.37 0.16 KAQ1 64 7.53 0.44 142 7.64 0.16 222 7.62 0.44 245 7.63 0.16 395 7.72 0.44 468 7.64 0.16 KAQ2 128 7.18 0.40 650 7.70 0.16 422 7.42 0.40 RTQ-2 99 7.59 0.40 672 7.19 0.40 repeat 311 7.65 0.40 KAQ3 30 7.47 0.42 577 7.79 0.40 507 7.78 0.42 RTQ-3 21 7.77 0.16 653 7.64 0.42 93 7.78 0.16 KAQ4 53 7.03 0.46 255 7.76 0.16 415 7.40 0.46 379 7.86 0.16 670 7.77 0.46 544 7.91 0.16 KAQ-5 10 7.30 0.15 RTQ-3 66 7.42 0.41 229 7.19 0.15 repeat 333 7.41 0.41 394 7.33 0.15 525 7.16 0.41 556 7.55 0.15 RTQ4 94 7.76 0.41 768 7.57 0.15 312 7.99 0.40 KAQ-5 61 7.46 0.42 702 8.12 0.40 repeat 389 7.07 0.42 RTQ-5 41 7.71 0.16 717 7.03 0.42 131 7.69 0.16 KAQ-6 11 7.20 0.15 286 7.72 0.16 78 7.44 0.15 414 7.70 0.16 179 7.37 0.15 606 7.66 0.16 271 7.48 0.15 RTQ-5 143 7.49 0.45 414 7.36 0.15 repeat 326 7.58 0.45 KAQ-6 14 7.97 0.44 563 7.53 0.45 repeat 270 7.11 0.44 RTQ-6 54 7.80 0.16 397 7.53 0.44 342 7.73 0.16 KAQ7 79 7.30 0.43 631 7.65 0.16 386 6.94 0.43 819 7.59 0.16 590 7.33 0.43 972 7.62 0.16 Rotoma RTQ-6 90 7.86 0.43 RMQ1 21 7.95 0.44 repeat 536 7.57 0.43 105 8.07 0.44 880 7.25 0.44 186 7.60 0.44 RTQ7 30 7.86 0.42 RMQ2 80 7.92 0.40 190 8.01 0.42 216 7.67 0.39 334 7.58 0.42 515 7.55 0.39 RTQ8 74 7.83 0.46 RMQ3 50 7.69 0.41 268 7.84 0.46 217 8.21 0.41 420 7.71 0.46 435 7.69 0.41 Utu RMQ-4 18 7.68 0.17 UTQ-1 79 7.58 0.16 90 7.72 0.17 112 7.71 0.16 235 7.70 0.17 152 7.64 0.16 350 7.78 0.17 312 7.91 0.16 430 7.80 0.17 423 7.62 0.16 RMQ-4 105 7.48 0.39 UTQ-1 28 7.46 0.44 repeat 253 7.82 0.39 repeat 265 7.69 0.44 400 7.42 0.39 323 7.77 0.44 RMQ-5 21 7.70 0.17 UTQ2 14 7.70 0.41 115 7.88 0.17 225 7.64 0.41 212 7.68 0.17 436 7.38 0.41 296 7.92 0.17 UTQ3 51 7.71 0.42 392 7.74 0.17 191 8.31 0.42 RMQ-5 57 8.43 0.44 280 7.53 0.43 repeat 189 7.88 0.44 UTQ-4 56 7.65 0.16 369 7.51 0.44 123 7.50 0.16 RMQ6 116 7.11 0.44 227 7.53 0.16 229 7.54 0.44 394 7.72 0.16 535 7.97 0.44 464 7.69 0.16 RMQ7 76 7.81 0.43 UTQ-4 130 7.55 0.40 203 7.83 0.43 repeat 330 8.03 0.40 297 8.46 0.44 486 7.96 0.40 Rotoiti UTQ5 75 7.59 0.42 RTQ-1 42 7.54 0.16 246 7.45 0.42 91 7.63 0.16 417 8.14 0.43 185 7.64 0.16 UTQ6 62 7.45 0.44 315 7.51 0.16 442 7.40 0.44 524 7.61 0.16 701 7.50 0.45 897 7.49 0.16 UTQ7 56 7.28 0.45 256 7.77 0.45 356 7.42 0.45 M. Sas, P. Shane, N. Kawasaki et al. Journal of Volcanology and Geothermal Research 421 (2022) 107430 textures are observed in previous studies of Okareka and Rotoiti quartz (Shane et al., 2008a; Smith et al., 2010), but differ slightly from Earth- quake Flat quartz CL images (Smith et al., 2010; Matthews et al., 2012), some of which show a bright CL band on the crystal rims in addi- tion to the aforementioned textures. Although quartz crystals exhibit similar textures in the samples, the core zones differ. Specifically, Utu quartz cores are commonly bright (i.e., Q2), and some cores exhibit resorption surfaces that have amixture of bright and dark intensities (e.g., Fig. 3g). In contrast, Rotoiti quartz crystals commonly have dark cores (i.e., Q3; Fig. 3e). Other Rotoiti cores, as well as many Rotoma cores, exhibit little to no difference in brightness between cores and rims (i.e., Q1; Fig. 3c–d,f). Kaharoa cores, similarly to Utu, are mostly bright (i.e., Q2; Fig. 3a–b), and several cores display a complex mixture of bright and dark intensities. 4.2. Oxygen isotopic compositions Intra-crystal δ18O data obtained via SIMS for the four pilot units (Utu, Rotoiti, Rotoma, and Kaharoa) are listed in Table 2, with examples of crystals and their respective SIMS analyses shown in Fig. 3, and all crys- tals and SIMS analyses shown in Supplementary Figs. A1–A4. We ana- lyzed 5–6 spots in each crystal from zones of variable CL intensity. Two crystals from each of Utu, Rotoma, and Kaharoa, and five crystals from Rotoiti, yielded intra-crystal homogeneity better than the analyti- cal reproducibility of the NBS28 standard in high precision mode (0.23‰, 2σ) (Table 2). To assess potential variations between different quartz grains in each sample, a total of seven to eight crystalswere examined per sample in a faster but lower-precision mode, analyzing 3 spots in each crystal. Quartz from all four samples exhibited inter-crystal δ18O homogeneity at a level similar to the analytical reproducibility of the NBS28 standard (0.54‰, 2σ) (Table 2). To further investigate sample-scale isotopic var- iability, the data were modelled as Gaussian distribution curves using kernel density estimation (KDE), with maximum analytical uncertainty (~0.5‰, 2σ) set as the bandwidth (Fig. 4). The standard deviation pro- duced by theGaussianmodels (0.6–0.7‰) is similar to the analytical un- certainty, indicating isotopic homogeneity within each sample ismostly within the resolution of this low-precision instrumental mode. The units do exhibit variation among modelled means, although they over- lap within error (Fig. 4). Assuming relative inter-crystal homogeneity, two additional whole crystals from each sample were analyzed using themore precise LF technique (Table 3, Fig. 5). Within-sample δ18O ho- mogeneity is indicated by these sub-samples of quartz (1–2 mg, com- monly representing a single crystal). Analyses are listed in Table 3 and shown in Fig. 5. Finally, the means of the analytical methods, SIMS and LF, are in good agreement and within analytical uncertainty of one another (Fig. 5). Moreover, both methods point to δ18O homogeneity (within SIMS error, ~0.5‰, 2σ) across the four units studied (Fig. 5). Quartz ho- mogeneity within and between units is also observed across the addi- tional six units analyzed using high precision LF (Table 3). However, there is one outlier beyond the LF analytical error: a Haparangi crystal with notably lower ratios (by 1.14‰) than the unit mean (Table 3). 5. Discussion 5.1. Implications of quartz δ18O homogeneity Changes in δ18O ratios up to ~6‰ have been reported in magmas from rhyolite volcanoes following caldera formation (e.g., Bindeman and Valley, 2001). This is not evident for the OVC. At intra-crystal- scale, δ18O homogeneity across differing textural growth zones are within analytical error (~0.2‰, 2σ) (Table 2, Fig. 3). Thus, during the pre-eruptive amalgamation of separate magma batches into the erupt- ible magma as envisaged by previous workers (e.g., Shane et al., 2005, 2008b; Smith et al., 2006, 2010; Shane, 2015), the δ18O ratios of quartz 7 Table 2 (continued) Unit/Crystala distanceb δ18O 2σc Unit/Crystala distanceb δ18O 2σc Reproducibility of NBS-28 Standard 2SD (n) Low-precision mode 0.54 (10) High-precision mode 0.23 (10) a Repeat analyses represent duplicate analyses of the crystal in low-precision mode. b Analyses are listed in the order from core to rim,with distance from the core provided in μm. c 2σ<0.20 indicate analyses completed in high-precisionmode, and 2σ>0.35 indicate analyses completed in low-precision mode. Fig. 4. Low-precision SIMS δ18O analyses for (a) Kaharoa, (b) Rotoma, (c) Rotoiti, and (d) Utu, modelled by Gaussian distribution curves using kernel density estimation (KDE), with maximum analytical uncertainty (± 0.54 2σ) set as the bandwidth. Similarity between modelled KDE and SIMS analytical error indicates intra-unit homogeneity with the resolution of this study. M. Sas, P. Shane, N. Kawasaki et al. Journal of Volcanology and Geothermal Research 421 (2022) 107430 indicate neither significant assimilation of hydrothermally altered roof rocks nor that influx of isotopically distinct melts from deeper in the system occurred (i.e., mantle-derived melts; e.g., Bindeman and Valley, 2003; Budd et al., 2017). Inter-crystal isotopic homogeneity is also observed and demonstrates that the parental reservoirs of the dis- cretemelts fromwhich the quartz crystals are derived (e.g., Shane et al., 2008b; Smith et al., 2010) display considerable overlap in δ18O ratios (based on the differing textural patterns of crystals, i.e., cores versus rims). This suggests a spatial δ18O uniformity (to within 0.5‰, 2σ) across the OVC plumbing system. A similar uniformity in radiogenic iso- topic ratios is observed in plagioclase, an earlier crystallizing phase, ex- tracted from the same OVC rhyolite units. Specifically, based on plagioclase textures and trace element contents, Sas et al. (2021) deter- mined that plagioclase from the rhyolites nucleated from more mafic melts (albeit not basalts), while plagioclase rims grew in equilibrium with silicicmelts. Despite these findings, core and rim Sr–Pb isotopic ra- tios of plagioclase were largely homogeneous, indicating that the OVC reservoir reached this level of isotopic homogeneity prior to rhyolite formation. The observed homogeneity spans ~550 thousand years and demonstrates temporal δ18O and Sr–Pb isotopes uniformity in the mag- matic system. These observations suggest that the OVC magmatic sys- tem has maintained a relatively consistent balance between mantle and crustal mass contributions. The same δ18O homogeneity is further supported by the additional six samples analyzed via LF, as variations between the means of all ten units analyzed in this study are within analytical error (≤0.5‰, 2sd) (Tables 2 and 3; Fig. 5). The new δ18O data (Tables 2 and 3) are similar to reported quartz δ18O ratios from the OVC based on aliquots of bulk quartz (5–30 mg; Blattner et al., 1996) (Fig. 5), supporting the overall observation of isotopic uniformity.Wenote a lava dome outside the cal- dera rim (Haparangi; Fig. 1) contains an outlier crystal that is beyond the uncertainty of the sample mean (Fig. 5). The Haparangi outlier has a lower δ18O ratio (6.7‰) than other crystals in this study (Table 3; Fig. 5). Hydrothermal alteration may have led to the lower ratio in the Haparangi outlier. Evidence of subsurface hydrothermal alteration of rocks (e.g., comagmatic granitoid lithics found in ignimbrites) in the TVZ is sparse (n = 3; Brown et al., 1998), although a granitoid lithic from Kaharoa with low δ18O ratios (2.8 ± 1.1‰, adjusted to NBS28 ra- tios in this study; Supplementary Table A1) indicates that interaction between shallow magmatic bodies and meteoric water is possible (Shane et al., 2012). Bindeman (2008) demonstrated that hydrothermal alteration generally leads to more erratic δ18O ratios throughout crystal populations, while the Haparangi sample population is relatively uni- form (mean of 7.8 ± 0.3‰, 2sd, n = 7). Therefore, it is possible that the Haparangi outlier ratio represents crystallization in a melt that ex- perienced less crustal assimilation or interaction with mafic influx. Al- ternatively, it is also possible that incorporation of hydrothermally altered materials is volumetrically sparse, and/or that the ratios of hy- drothermally altered materials (1.5–3.1‰, adjusted; Brown et al., 1998) are too moderate to significantly impact the isotopic ratios of OVC magmas, as is explored below. 8 5.2. Crustal contamination and rhyolite magma genesis at OVC Recentmodels for rhyolite petrogenesis, including those for the OVC, include: (1) fractionating basaltic parental melts that experience some Table 3 Laser fluorination analyses of OVC quartz. δ18O (‰) 2sd δ18O (‰) 2sd δ18O (‰) 2sd OVC rhyolites Kaharoa Earthquake Flat Haparangi 7.72 0.11 7.80 0.13 6.68 0.11 7.55 0.11 7.71 0.13 7.63 0.11 Whakatane Rotoiti 7.83 0.13 7.48 0.13 7.78 0.11 7.72 0.13 7.65 0.13 7.61 0.11 8.02 0.13 7.53 0.13 Utu 7.91 0.13 Rotoma 7.77 0.11 7.92 0.13 7.73 0.11 7.96 0.11 7.72 0.13 7.63 0.11 Kakapiko Round Hill Okareka 7.77 0.13 7.40 0.13 7.42 0.13 7.28 0.13 7.55 0.13 7.73 0.13 7.27 0.13 7.41 0.13 7.64 0.13 UOG standard Reporteda Analyzed ± 2sd (n) Session I 6.52 6.53 0.11 (3) Session II 6.52 6.55 0.13 (3) a (Troll et al., 2013). M. Sas, P. Shane, N. Kawasaki et al. Journal of Volcanology and Geothermal Research 421 (2022) 107430 degree of crustal contamination (via assimilation of crust and/ormixing with crustalmelts), (2)magmatic storage largely consisting of a crystal- line mush zone with melt(s) accumulation near the roof of the system, sometimes in semi-isolated bodies, and (3) magmatic rejuvenation through heat, volatile, and/or mass transfer that are provided from mantle-derived melts that underlie the silicic reservoir (e.g., Hildreth, 2004; Bachmann and Bergantz, 2004, 2008a, 2008b; Shane et al., 2008a; Smith et al., 2010; Bachmann and Huber, 2016). Constraining magma sources via radiogenic isotopes is challenging at the OVC due to the overlap of isotopic ratios of TVZ rhyolites and basalts, and the likely basement lithologies involved (Graham et al., 1992; McCulloch et al., 1994; Gamble et al., 1996; Price et al., 2015). However, δ18O ratios of the likely dominant basement lithology (Mesozoic Torlesse Terrane metasediments, ≥ 9‰) and TVZ basalts (5.3–6.8‰) (Blattner and Reid, 1982; McCulloch et al., 1994) do not overlap (Fig. 5) (δ18O ratios from literature have been adjusted to the NBS28 ratio used in this study; Sup- plementary Table A1). Hence, OVC quartz δ18O ratios may elucidate mantle versus crustal contributions to the magmatic system. Work by Graham et al. (1992) has demonstrated that, composition- ally and isotopically, Waipapa metasediments are not ideal dominant crustal assimilants for TVZ magmas. Early models suggested that the δ18O ratios of TVZ rhyolites are too low to reflect substantial assimilation of Torlesse Terrane metasediments (i.e., Blattner and Reid, 1982; Graham et al., 1992). In contrast, otherwork using radiogenic and stable isotopes demonstrate that Torlesse-like compositions serve as ideal assimilants for TVZ magmas (Gamble et al., 1990, 1993; McCulloch et al., 1994; Macpherson et al., 1998). Specifically, Gamble et al. (1993) suggest that TVZ basalts result from ≤10% Torlesse crustal con- tamination of a primitive TVZ basalt. For rhyolites, McCulloch et al. (1994) suggest as much as 25% crustal contamination of a basaltic par- ent based on Sr-Nd-Pb-O isotopes of mineral separates (Maroa Volcanic Centre rhyolites). Recent work reexamined assimilation-fractional crystallization (AFC) and mixing processes at the OVC based on whole rock composi- tions, Sr–Pb isotopes, and plagioclase Sr–Pb isotopes (Sas et al., 2021). AFC andmixingmodels of Sas et al. (2021) call on ≥20–30% assimilation of a Torlesse-like crustal source by a mid-ocean ridge (MOR)-like man- tle source (DMM) with varying amounts of subduction flux. To test whether models using quartz δ18O ratios are consistent with models based on radiogenic isotopes, we modelled simple mixing between ba- saltic parentalmelts and Torlesse-likemetasediments. The δ18O ratios of potential crystallization melts were calculated from average quartz ra- tios for each unit (δ18Omelt ≅ δ18Oquartz – 0.4‰ for high-SiO2 rhyolites; Bindeman and Valley, 2003; Bindeman, 2008). The calculated δ18Omelt 9 ratios were combined with whole rock isotopic data (Sr–Nd) of OVC rhyolites from Burt et al. (1998), Nairn et al. (2004), and Schmitz and Smith (2004), and whole rock and plagioclase isotopic data (Sr–Pb) of Sas et al. (2021). All isotopic standards were adjusted to the same stan- dard ratios where possible (Supplementary Table A1). Two DMM pa- rental melts and two Torlesse-like metasedimentary assimilants were used for modelling. The DMM parental melts are: (1) a relatively prim- itive composition represented by a Havre Trough basalt sample (VUW158/4) from the Western Ngatoro Basin, which has been negligi- bly modified via subduction processes (Gamble et al., 1993, 1996; Macpherson et al., 1998), and (2) a Rumble IV basalt sample (VUW 162/1) from the EasternNgatoro Basin,which has been slightly contam- inated by a fluid-dominant subduction flux (Gamble et al., 1993, 1996; Macpherson et al., 1998). These off-shore basalts were selected as avail- able O-Sr-Nd-Pb analyses of TVZ basalts indicate crust and/or rhyolite contamination, including at the OVC (e.g., Rotokawau basalt, δ18O = 6.0‰ adjusted; Supplementary Table A1) (Brown et al., 1998). Sample locations of the basaltic endmembers relative to the OVC are shown in the inset map in Fig. 1. The crustal assimilant compositions are: (1) a Torlesse Kaweka member representative composition of Price et al. (2015), with δ18O ratios from Macpherson et al. (1998), and (2) a Torlesse-like southern TVZ metasedimentary xenolith with enriched Pb isotopes, with compositions from Price et al. (2012, 2015) and aver- age δ18O ratios from Blattner and Reid (1982). Additional TVZ crustal contaminants (i.e., southern TVZ metaigneous xenoliths, Waipapa metasediments) and parental melts (i.e., OVC basalts) were considered during modelling and are shown along with modelling results. Compo- sitions used for modelling are listed in Supplementary Table A2 and modelling results of δ18O versus radiogenic isotopes (Sr-Nd-Pb) are shown in Fig. 6. Modelling results using δ18O ratios indicate ≥25% assimilation of Torlesse-like metasedimentary crust by a DMM source with minor and variable amounts of subduction flux over the lifespan of the mag- matic system might explain the source of the erupted magmas (Fig. 6). Modelling results also demonstrate that (1) OVC basalts do not represent ideal parental melts as their Sr–Nd radiogenic isotope compositions are comparable to those of rhyolites, and (2) that other potential crustal contaminants would not reproduce the isotopic trends observed at the OVC isotopic compositions (i.e., mid-upper crust Waipapa metasediments, Pb isotopic ratios too low; lower crust metaigneous xenoliths found in southern TVZ andesites, Nd isotopic ratios too high; Price et al., 2012) (Fig. 6). Late-stage crustal contaminants were also considered. Specifically, although δ18O variability of the rhyolite units is largely within error (0.5‰, 2σ; Fig. 5), the potential impacts of rhyolite assimilation of hy- drothermally altered roof materials were assessed based on the pres- ence of the Haparangi outlier quartz (6.7‰) and the lower average whole rock ratios of the other two extra-caldera units, Kakapiko (6.9‰) and Round Hill (7.1‰) (Supplementary Table A1). Assimilation was modelled as both AFC (Supplementary Fig. A5) and simple mixing (Fig. 6; Supplementary Fig. A5) between a representative OVC rhyolite endmember (Rotoiti, δ18O = 7.3‰; Supplementary Table A1) and a co-magmatic, hydrothermally altered, granitoid lithic found in Kaharoa (2.8 ± 1.1‰; Brown et al., 1998) (Supplementary Table A2). Composi- tionally, the granitoid lithic is nearly identical to the rhyolites (Fig. 2; Supplementary Table A2), therefore AFC (at F < 0.5 and r = 0.3; McCulloch et al., 1994; Price et al., 2012; Sas et al., 2021) and mixing have little impacts on rhyolite compositions (Supplementary Fig. A5). Isotopically, modelling demonstrates two points. First, due to the moderate δ18O ratios and nearly identical Sr-Nd-Pb ratios (relative to rhyolites), <5% assimilation of hydrothermally alteredmaterial is unde- tectable in caldera and intra-caldera rhyolites. Second, incorporation of ~7–10% of hydrothermally altered material could produce the lower ra- tios observed in extra-caldera units (Fig. 6). Thus, it is possible that extra-caldera units (i.e., Kakapiko, Haparangi, Round Hill) assimilated altered roof material, albeit physical evidence for such substantial Fig. 5. δ18O ratios of OVC rhyolites plotted with δ18O ratio ranges for TVZ on-shore basalts, metasediments, hydrothermally altered granitoids, and meteoric water, as well as typical MOR basalt (MORB) ratios and global high silica rhyolite ratios. OVC rhyolite δ18O ratios include means of intra-crystal SIMS analyses (large red circles), whole-crystal LF analyses (large yellow circles), and bulk (multi-crystal) quartz analyses (small black circles; Blattner et al., 1996). Ratios of typical MORB, MORB-derived rhyolitic melts, high silica rhyolites, and Yellowstone rhyolites are from Bindeman (2008). Range of TVZ rhyolite-hosted, hydrothermally altered granitoid lithics are from Brown et al. (1998). Ranges of TVZ basalts, Torlesse and Waipapa metasediments, and southern TVZ metasedimentary xenoliths are from Blattner and Reid (1982) and McCulloch et al. (1994). All whole rock δ18O ratios from literature have been corrected toNBS28=9.29‰ (Kusakabe andMatsuhisa, 2008),with original ratios and corrected ratios listed in Supplementary Table A1.Meteoricwater data are from Taylor et al. (1977). M. Sas, P. Shane, N. Kawasaki et al. Journal of Volcanology and Geothermal Research 421 (2022) 107430 incorporation of hydrothermally altered material is absent in the TVZ (Brown et al., 1998). Alternatively, the trio may have experienced some degree of hydrothermal alteration similar to the granitoid lithic. Specifically, it is suggested that the extra-caldera units could potentially represent remobilized plutons based on their high crystallinity (20–40%), petrological and geochemical similarities, and proximity to Earthquake Flat (Sas et al., 2021), which Molloy et al. (2008) suggest was a rejuvenated, partially-crystalline, rhyolitic pluton. 10 Overall, the models are consistent with AFC and radiogenic model- ling results of OVC rhyolites based on whole rock and plagioclase (Sas et al., 2021), as well as recent studies that call on high percentages (20–30% based on radiogenic isotopes; Fig. 6) of DMM partial melting for TVZ intra-caldera basalt petrogenesis (Waight et al., 2017; Zellmer et al., 2020; Barker et al., 2020). Large volumes of melt in the upper mantle (≤12%; Stern and Benson, 2011), lower crust (≤2%; Heise et al., 2010), and middle crust (≤ 4%; Harrison and White, 2006) have also Fig. 6. Simple isotopicmixingmodels using off-shore TVZ DMMbasalts (stars) as parents and Torlesse-likemetasediments (dark grey diamonds) as assimilants. All compositions used for modelling are listed in Supplementary Table A2. Assimilant compositions used for modelling consist of the Torlesse Kaweka representative composition in (a–e) and a metasedimentary xenolith composition in (f–i) (Price et al., 2012, 2015). Representative compositions of other potential isotopic contaminants (Waipapa metasediments, light grey diamond; metaigneous xenoliths, brown square) and parents (OVC basalt, purple circle) are included for reference and are also listed in Supplementary Table A2. Square regions represent isotopic ranges for the Torlessemetasediments (dark grey, shaded),Waipapametasediments (light grey, shaded), andmetaigneous xenoliths from the southern TVZ that are thought to be representative of the lower crust (brown, dashed outline). Isotopic data for the metasedimentary xenoliths are limited and therefore no range is available (Supplementary Table A2; Price et al., 2012, 2015). Also included are simple isotopicmixing lines between a hydrothermally altered OVC granitoid (dark blue diamond; Brown et al., 1998) andOVC rhyolites, with compositionalmixing and AFC models shown in Supplementary Fig. A5. All isotopic ratios (O-Sr-Nd-Pb) have been adjusted to a common standard ratio where possible (e.g., all δ18O ratios have been corrected to NBS29; Kusakabe andMatsuhisa, 2008). The original reported isotopic ratios, the corrected isotopic ratios, and the respective references for each unit are listed in Supplementary Table A1. ForOVC rhyolites, only Pb isotopic data of Sas et al. (2021)were used (whole rock and plagioclase) as Pb ratios fromprevious studies could not be corrected (Supplementary TableA1). The LF error of δ18O ratios (±0.1‰, 2sd; Table 3) is included on the bottom right of panel j. Errors of whole rock Sr-Nd-Pb isotopic ratios (Supplementary Table A1) and plagioclase Sr and 206Pb/204Pb isotopic ratios (Sas et al., 2021) are smaller than their respective data symbols, thus they are not included. Where plagioclase errors (2sd) are larger than the respective data symbols, errors have been included on the bottom left of the panel. There are no plagioclase Nd isotopes, and whole Nd isotopic ratios are only available for Kaharoa, Earthquake Flat, and Rotoiti (Supplementary Table A1), therefore the isotopic range of OVC rhyolites in panels b and g is narrower. M. Sas, P. Shane, N. Kawasaki et al. Journal of Volcanology and Geothermal Research 421 (2022) 107430 been previously suggested by geophysical studies at the OVC. Fig. 7 pro- vides a simplified schematic model for rhyolite petrogenesis at the OVC based on these geophysical data, our models, and previously published 11 petrogenic models (e.g., Smith et al., 2005; Molloy et al., 2008; Shane et al., 2008a, 2008b; and others). This conceptual model shows that the zone of metasediment assimilation is largely restricted to the roots Fig. 7. Schematic model for magma genesis at the OVC based on geophysical and petrological models, as well as AFC andmixing isotopic models (Sas et al., 2021; this study). The mantle beneath the OVCmelts as a result of decompression and variable subduction flux, with the upper mantle containing as much as ~12%melt (Stern and Benson, 2011). The middle to lower crust beneath the OVC is then heavily intruded by mantle-derived basalt and contains as much as ~1–2% melt (Harrison and White, 2006). The basaltic melts stall at approximately 10–15 km depth, where they assimilate Torlesse-like metasedimentary crustal materials and fractionate (Sas et al., 2021; this study). Melt content at ~10 km depth is thought to be as high as 4% (Heise et al., 2010). Periodic basaltic melts penetrate the silicic reservoir and provide heat and volatiles to isolated silicic melt pods, consequently triggering movement of rhyolitic melts and eruption largely via contribution of heat and volatiles. Assimilation of hydrothermally altered materials, if present, is more dominant in extra-caldera settings, where the reservoir is mostly crystalline. M. Sas, P. Shane, N. Kawasaki et al. Journal of Volcanology and Geothermal Research 421 (2022) 107430 of the reservoir where parental magmas reside, and occurs prior to the segregation of fractionated melt based on the largely homogenous sta- ble (this study) and radiogenic (Sas et al., 2021) isotopic ratios, whereas any potential impact of hydrothermal alteration is restricted to the up- permost reservoir and is more prominent in extra-caldera settings. 5.3. Quartz δ18O records in caldera systems The observed isotopic homogeneity at the OVC will herein be discussed in context to other major global silicic centres in several tectono-magmatic settings. Changes in oxygen isotopic ratios of magmas, which have been ascribed to open system processes, have been recognized in pre-, syn-, and post- caldera-forming events at sev- eral silicic volcanic centers of various tectonic settings. Commonly, a decrease in δ18O ratios at these centers is attributed to mafic influx and/or melting and incorporation of low-δ18O, hydrothermally altered material. For instance, quartz and other minerals erupted from the intra-plate Yellowstone and Snake River Plain caldera centers exhibit significant δ18O ratio diversity. Specifically, crystals in deposits from three caldera eruptions at Yellowstone (~2 Ma, 1.3 Ma, and 0.64 Ma with ~2500 km3, ~300 km3, and ~1000 km3, respectively) exhibit notable within-unit heterogeneity (quartz δ18O ratios variability of ~0.5–1.8‰) (Hildreth et al., 1984; Bindeman and Valley, 2000, 2001). The differences between quartz crystals from the three Yellowstone cal- dera deposits are even larger (up to ~2.6‰). Furthermore, quartz crys- tals from post-2 Ma and post-0.64 Ma intra-caldera lavas reflect significant drops in δ18O ratios relative to their precursory caldera events (as much as ~6‰, Fig. 8a). Quartz crystals from extra-caldera lavas, in contrast, have maintained δ18O ratios that overlap with the 2 Ma caldera event (Hildreth et al., 1984; Bindeman and Valley, 2000, 2001). The authors note that the observed drops in δ18O ratios are too large to be explained by MOR basalt-like parental melts, and instead 12 require voluminous crustal assimilation of hydrothermally altered rock following evacuation of the silicic reservoir. Older rhyolite calderas from the Snake River Plain, namely Bruneau-Jarbidge, 12.7–8.1 Ma (Boroughs et al., 2005; Bindeman and Simakin, 2014), Picabo, 10.5–6.6 Ma (Drew et al., 2013), and Heise, 6.6–3.9 Ma (Bindeman et al., 2007;Watts et al., 2011), all have similar shifts in δ18O ratios asso- ciated with caldera collapse with similar proposed origins. Specifically, within-unit variations of quartz δ18O ratios at these centers reach 1.5‰, and between-unit variations are as large as 4‰. Drops in δ18O have also been noted in a subduction setting in Suma- tra (Indonesia), which is situated on thick continental crust (>35 km; Sakaguchi et al., 2006). An intra-crystal investigation of the youngest caldera-forming event at Toba volcano (Young Toba Tuff, ~2800 km3, 75 ka) also calls on incorporation of hydrothermally altered crustal ma- terial (Budd et al., 2017). The authors noted overall lower (relative to the core) δ18O ratios in quartz rims, with core-to-rim variations as large as 1.8‰, and several per mil inter-crystal variation (Fig. 8b). Similar fluctuations in δ18O ratios are also observed at caldera cen- ters in the continental extensional setting of the Basin and Range prov- ince (western United States). Quartz and other minerals in deposits from voluminous caldera eruptions (>900 km3, 12.8–11.45 Ma) at the Southwestern Nevada Volcanic Field show within-unit δ18O ratio vari- ability of 0.6–1.2‰ and between-unit variability of ~3.0‰ (Bindeman and Valley, 2003). The authors interpreted the isotopic variations to involve rapid magma generation through mafic influx, and subsequent assimilation of isotopically varied (due to hydrothermal alteration) crustal rocks. In contrast to these systems, the OVC is situated in an actively rifting arc (~12 mm/yr at the OVC; Wallace et al., 2004) with thin continental crust (~25 km; Bannister et al., 2004), where the modern subduction tectonic regime began ~16 Ma (Wilson and Rowland, 2016). Eruption frequency at the OVC is high (millennial timescales), the magmatic Fig. 8. Simplified illustration comparing rhyolite magma genesis models and silicic reservoir dynamics at caldera-forming centers of different tectonic settings, as well as their respective known number of caldera eruptions (n), estimated DRE volumes (km3), and variability noted in quartz δ18O isotopic ratios (‰). (a) The Yellowstone caldera, western United States, is situated in thick continental crust and is associated with hot spot magmatism (after Huang et al., 2015; ratios and volumes from Hildreth et al., 1984 and Bindeman and Valley, 2000, 2001). (b) The Toba caldera, Indonesia, is situated in a subduction setting (modified after Jaxybulatov et al., 2014; ratios and volumes from Budd et al., 2017). (c) The Long Valley caldera is situated in the Basin and Range extension province inwestern United States (after Hildreth andWilson, 2007; ratios and volumes from Bindeman and Valley, 2002 and Hildreth, 2004). (d) The OVC, North Island, Aotearoa New Zealand, is situated in a coupled subduction-extension setting where the continental crust is thin (simplified from Fig. 7; δ18O ratios from this study and volumes from Cole et al., 2014). M. Sas, P. Shane, N. Kawasaki et al. Journal of Volcanology and Geothermal Research 421 (2022) 107430 system is relatively young (~0.6 Ma; Leonard et al., 2010), and eruption volume magnitudes are comparatively lower (90–190 km3 per caldera collapse event, Table 1; Nairn, 2002; Smith et al., 2005; Shane et al., 2008b; Deering et al., 2011). This high frequency of volcanic activity, alongwith homogeneous isotopic compositions, indicate a steady influx of mantle material into the system and constant mechanisms of crustal assimilation. Although δ18O ratio differenceswere noted in lava external to the caldera and suggest that assimilation of hydrothermally altered material and/or hydrothermal alteration is plausible, they are un- common (i.e., Haparangi) or subtle (<1‰) and lie within the range of other rhyolites (i.e., Kakapiko, Round Hill), which, in addition to the strikingly similar compositions of hydrothermally altered material and rhyolites, make such assimilation challenging to confirm (Fig. 5, Fig. 6, Supplementary Fig. A5). Themassflux equilibriumobserved at theOVC differs from the other caldera systems discussed above (Fig. 8) but does resemble those of smaller volume (≤600 km3) caldera systems in extensional setting, such as the Long Valley caldera. The Long Valley caldera is of similar age (0.76–0.04 Ma; Bindeman and Valley, 2002) and is situated in an extensional setting, albeit intra-plate, in the western margin of the ~17 Ma Basin and Range extension province (western United States; Zoback et al., 1981). Although historical extension rates at the Long Val- ley caldera have varied significantly due to resurgence fluctuations (e.g., ≤70 mm/yr; Hill, 2006), extension rates 60 km SE of the caldera in Owens Valley are similar to extension rates at the OVC (~10 mm/yr; Dixon et al., 1995). Notably, the Long Valley caldera exhibits similar uni- formity in δ18O ratios to those observed at the OVC (Fig. 8c). Specifically, the majority of quartz erupted in the Bishop Tuff event (~600 km3, 760 ka) have homogeneous ratios that indicate significant (20–30%) crustal assimilation (Bindeman and Valley, 2002; Hildreth, 2004). Un- like at the OVC, however, average pre-caldera lavas (~0.1–1.1 Ma prior) have higher (by ~1‰) and more variable quartz δ18O ratios than the Bishop Tuff, indicating homogenization of the magmatic stor- age system prior to the climactic caldera-forming eruption (Bindeman and Valley, 2002). Hildreth and Wilson (2007) also demonstrated that the rhyolitic melts of the Bishop Tuff can be correlated via fractional crystallization, which is not the case for the rhyolitic melts at OVC (e.g., Smith et al., 2006). Similarly, δ18O fluctuations between the Otowi (1.61Ma, ~400 km3) and Tshirege (1.23Ma, ~250 km3)Members of the Bandelier Tuff at Valles caldera are <1‰ (Wolff et al., 2002). Like 13 the OVC and Long Valley systems, the intra-plate Valles system is situ- ated in an extensional setting in the Rio Grande Rift (southwestern United States). Albeit both the Valles magmatic system (~2 Ma; Wolff et al., 2002) and the tectonic regime (~26 Ma; Golombek et al., 1983) are older, extension rates in the region over the past 5 Ma are compara- ble to those of the OVC and Long Valley (~14 mm/yr; Golombek et al., 1983). The similarities of quartz δ18O ratios between these systems sug- gests that the extensional setting, as opposed to the subduction setting, has the largest impact on the mass flux dynamics of the OVC magmatic system (Fig. 8d). 6. Conclusions Despite the resorption textures observed in all quartz crystals from OVC rhyolites, all of the intra-crystal-scale growth zones reveal homo- geneous δ18O ratios (to within ~0.2‰, 2σ). While previous mineral- based studies require nucleation of crystals in discrete, segregated melts within the OVC reservoir, results from this study indicate that themelts fromwhich quartz nucleatedwere derived from isotopically homogeneous source(s) (to within ~0.5‰, 2σ). The isotopic homoge- neity further extends across quartz crystals from rhyolite magmas that differ in age, volume, and vent location. This homogeneity of quartz δ18O ratios supports the concept of an isotopically homoge- nized magmatic system below the OVC, with steady contributions from mantle and crust. Simple isotopic mixing models suggest rhyo- lite genesis involving a DMM basaltic parent with variable subduction flux and significant assimilation (≥25%) of Torlesse-like metasedi- ments. The large degrees of assimilation and the homogeneity of iso- topic character require the presence of a voluminous network of mafic melts in the middle to lower crust to provide a vigorous thermal regime, which is supported by seismic and resistivity studies of the re- gion. The isotopic homogeneity observed at the OVC differs frommost other rhyolite caldera centers, which commonly exhibit δ18O fluctua- tions of ≥1‰ on a crystal-scale and require voluminous incorporation of hydrothermally altered crust, especially following caldera-collapse events. In contrast, quartz from rhyolitic OVC eruptions demonstrates the existence of a relatively buffered isotopic system across the lifespan of the volcano, and no detectable involvement (<5% assimilation) of hydrothermally altered roof rocks in the genesis of intra-caldera and caldera eruption products. M. Sas, P. Shane, N. Kawasaki et al. Journal of Volcanology and Geothermal Research 421 (2022) 107430 Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influ- ence the work reported in this paper. Acknowledgements The authors would also like to thank Ilya Bindeman and Raimundo Brahm for their assistancewith some samples, aswell as Katy Chamber- lain and an anonymous reviewer for their comments, which greatly im- proved this manuscript. Support for this project comes from the New Zealand Ministry of Business, Innovation, and Employment grant MAUX1507 to GFZ, the University of Auckland to MS, and Monka-sho to HY. Appendix A. Supplementary data Supplementary data to this article can be found online at https://doi. org/10.1016/j.jvolgeores.2021.107430. References Appleby, S.K., Graham, C.M., Gillespie, M.R., Hinton, R.W., Oliver, G.J.H., EIMF, 2008. A cryptic record of magma mixing in diorites revealed by high-precision SIMS oxygen isotope analysis of zircons. Earth Planet. Sci. Lett. 269, 105–117. Bachmann, O., Bergantz, G.W., 2004. On the origin of crystal-poor rhyolites: extracted from batholithic crystal mushes. J. Petrol. 45, 1565–1582. Bachmann, O., Bergantz, G.W., 2008a. Rhyolites and their Source Mushes across Tectonic Settings. J. Petrol. 49, 2277–2285. Bachmann, O., Bergantz, G.W., 2008b. The magma reservoirs that feed supereruptions. El- ements 4, 17–21. Bachmann, O., Huber, C., 2016. Silicic magma reservoirs in the Earth’s crust. Am. Mineral. 101, 2377–2404. Baertschi, P., 1976. Absolute 18O Content of standardmean oceanwater. Earth Planet. Sci. Lett. 31, 341–344. Bannister, S., Bryan, C.J., Bibby, H.M., 2004. Shear wave velocity variation across the Taupo Volcanic Zone, New Zealand, from receiver function inversion. Geophys. J. Int. 159, 291–310. Barker, S.J., Rowe, M.C., Wilson, C.J.N., Gamble, J.A., Rooyakkers, S.M., Wysoczanski, R.J., Illsley-Kemp, F., Kenworthy, C.C., 2020. What lies beneath? Reconstructing the prim- itive magmas fueling voluminous silicic volcanism using olivine-hosted melt inclu- sions. Geology 48, 504–508. Bindeman, I., 2008. Oxygen isotopes in mantle and crustal magmas as revealed by single crystal analysis. Rev. Mineral. Geochem. 69, 445–478. Bindeman, I.N., Simakin, A.G., 2014. Rhyolites—Hard to produce, but easy to recycle and sequester: integrating microgeochemical observations and numerical models. Geosphere 10, 930–957. Bindeman, I.N., Valley, J.W., 2000. Formation of low-δ18O rhyolites after caldera collapse at Yellowstone, Wyoming, USA. Geology 28, 719–722. Bindeman, I.N., Valley, J.W., 2001. Low-δ18O rhyolites from yellowstone: magmatic evo- lution based on analyses of zircons and individual phenocrysts. J. Petrol. 42, 1491–1517. Bindeman, I.N., Valley, J.W., 2002. Oxygen isotope study of the Long Valley magma sys- tem, California: isotope thermometry and convection in large silicic magma bodies. Contrib. Mineral. Petrol. 144, 185–205. Bindeman, I.N., Valley, J.W., 2003. Rapid generation of both high- and low-δ18O, large- volume silicic magmas at the Timber Mountain/Oasis Valley caldera complex, Nevada. Geol. Soc. Am. Bull. 18. Bindeman, I.N., Watts, K.E., Schmitt, A.K., Morgan, L.A., Shanks, P.W.C., 2007. Voluminous low δ18Omagmas in the late Miocene Heise volcanic field, Idaho: implications for the fate of Yellowstone hotspot calderas. Geology 35, 1019–1022. Blattner, P.E., Reid, F., 1982. Origin of lavas and ignimbrites of the Taupo Volcanic Zone, New Zealand, in the light of oxygen isotope data. Geochim. Cosmochim. Acta 46, 1417–1429. Blattner, P.E., Rui-Zhong, H., Graham, I.J., Houston-Eleftheriadis, C., 1996. Temperatures and isotopic evolution of silicic magmas, Taupo Volcanic Zone and Coromandel, New Zealand. N. Z. J. Geol. Geophys. 39, 353–362. Boroughs, S., Wolff, J., Bonnichsen, B., Godchaux, M., 2005. Large-volume, low-δ18O rhy- olites of the Central Snake River Plain, Idaho, USA. Geology 33, 821–824. Brown, S.J.A., Burt, R.M., Cole, J.W., Krippner, S.J.P., Price, R.C., Cartwright, I., 1998. Plutonic lithics in ignimbrites of Taupo Volcanic Zone, New Zealand; sources and conditions of crystallisation. Chem. Geol. 148, 21–41. Budd, D.A., Troll, V.R., Deegan, F.M., Jolis, E.M., Smith, V.C., Whitehouse, M.J., Harris, C., Freda, C., Hilton, D.R., Halldórsson, S.A., Bindeman, I.N., 2017. Magma reservoir dy- namics at Toba caldera, Indonesia, recorded by oxygen isotope zoning in quartz. Sci. Rep. 7, 40624. Burt, R.M., Brown, S.J.A., Cole, J.W., Shelley, D., Waight, T.E., 1998. Glass-bearing plutonic fragments from ignimbrites of the Okataina caldera complex, Taupo Volcanic Zone, 14 New Zealand: remnants of a partially molten intrusion associated with preceding eruptions. J. Volcanol. Geotherm. Res. 84, 209–237. Cole, J.W., Spinks, K.D., Deering, C.D., Nairn, I.A., Leonard, G.S., 2010. Volcanic and struc- tural evolution of the Okataina Volcanic Centre; dominantly silicic volcanism associ- ated with the Taupo Rift, New Zealand. J. Volcanol. Geotherm. Res. 190, 123–135. Cole, J.W., Deering, C.D., Burt, R.M., Sewell, S., Shane, P.A.R., Matthews, N.E., 2014. Okataina Volcanic Centre, Taupo Volcanic Zone, New Zealand: a review of volcanism and syn- chronous pluton development in an active, dominantly silicic caldera system. Earth- Sci. Rev. 128, 1–17. Danišík, M., Shane, P., Schmitt, A.K., Hogg, A., Santos, G.M., Storm, S., Evans, N.J., Keith, Fifield L., Lindsay, J.M., 2012. Re-anchoring the late Pleistocene tephrochronology of New Zealand based on concordant radiocarbon ages and combined 238U/230Th dis- equilibrium and (U–Th)/He zircon ages. Earth Planet. Sci. Lett. 349–350, 240–250. Davis,W.J., 1985. Geochemistry and Petrology of the Rotoiti and Earthquake Flat Pyroclas- tic Deposits. M.S. thesisUniversity of Auckland. Deering, C.D., Cole, J.W., Vogel, T.A., 2008. A rhyolite compositional continuum governed by lower crustal source conditions in the Taupo Volcanic Zone, New Zealand. J. Petrol. 49, 2245–2276. Deering, C.D., Gravley, D.M., Vogel, T.A., Cole, J.W., Leonard, G.S., 2010. Origins of cold- wet-oxidizing to hot-dry-reducing rhyolite magma cycles and distribution in the Taupo Volcanic Zone, New Zealand. Contrib. Mineral. Petrol. 160, 609–629. Deering, C.D., Cole, J.W., Vogel, T.A., 2011. Extraction of crystal-poor rhyolite from a hornblende-bearing intermediate mush: a case study of the caldera-forming Matahina eruption, Okataina volcanic complex. Contrib. Mineral. Petrol. 161, 129–151. Dixon, T.H., Robaudo, S., Lee, J., Reheis, M.C., 1995. Constraints on present-day Basin and Range deformation from space geodesy. Tectonics 14, 755–772. Drew, D.L., Bindeman, I.N., Watts, K.E., Schmitt, A.K., Fu, B., McCurry, M., 2013. Crustal- scale recycling in caldera complexes and rift zones along the Yellowstone hotspot track: O and Hf isotopic evidence in diverse zircons from voluminous rhyolites of the Picabo volcanic field, Idaho. Earth Planet. Sci. Lett. 381, 63–77. Froggatt, P.C., Lowe, D.J., 1990. A review of late Quaternary silicic and some other tephra formations from New Zealand: their stratigraphy, nomenclature, distribution, vol- ume, and age. N. Z. J. Geol. Geophys. 33, 89–109. Gamble, J.A., Smith, I.E.M., Graham, I.J., Peter, Kokelaar B., Cole, J.W., Houghton, B.F., Wilson, C.J.N., 1990. The petrology, phase relations and tectonic setting of basalts from the Taupo Volcanic Zone, New Zealand and the Kermadec Island Arc - Havre Trough, SW Pacific. J. Volcanol. Geotherm. Res. 43, 253–270. Gamble, J.A., Smith, I.E.M., McCulloch, M.T., Graham, I.J., Kokelaar, B.P., 1993. The geo- chemistry and petrogenesis of basalts from the Taupo Volcanic Zone and Kermadec Island Arc, SW Pacific. J. Volcanol. Geotherm. Res. 54, 265–290. Gamble, J.A., Wright, I.C., Woodhead, J.D., McCulloch, M.T., 1994. Arc and back-arc geo- chemistry in the southern Kermadec arc-Ngatoro Basin and offshore Taupo Volcanic Zone, SW Pacific. In: Smellie, J.L. (Ed.), Volcanism Assoc. Ext. Consum. Plate Margins Geol. Soc. Spec. Publ. 81, pp. 193–212. Gamble, J., Woodhead, J., Wright, I., Smith, I., 1996. Basalt and sediment geochemistry and magma petrogenesis in a transect from Oceanic Island arc to rifted continental mar- gin arc: the Kermadec—Hikurangi margin, SW Pacific. J. Petrol. 37, 1523–1546. Golombek, M.P., McGill, G.E., Brown, L., 1983. Tectonic and geologic evolution of the Espanola Basin, Rio Grande Rift: Structure, rate of extension, and relation to the state of stress in the western United States. Tectonophysics 94, 483–507. Graham, I.J., Blattner, P., McCulloch, M.T., 1990. Meta-igneous granulite xenoliths from Mount Ruapehu, New Zealand: Fragments of altered oceanic crust? Contrib. Mineral. Petrol. 105, 650–661. Graham, I.J., Gulson, B.L., Hedenquist, J.W., Mizon, K., 1992. Petrogenesis of Late Cenozoic volcanic rocks from the Taupo Volcanic Zone, New Zealand, in the light of new lead isotope data. Geochim. Cosmochim. Acta 56, 2797–2819. Harrison, A., White, R.S., 2006. Lithospheric structure of an active backarc basin: the Taupo Volcanic Zone, New Zealand. Geophys. J. Int. 167, 968–990. Heise, W., Caldwell, T.G., Bibby, H.M., Bennie, S.L., 2010. Three-dimensional electrical re- sistivity image of magma beneath an active continental rift, Taupo Volcanic Zone, New Zealand. Geophys. Res. Lett. 37, L10301. Hiess, J., Cole, J.W., Spinks, K.D., 2007. Influence of the crust and crustal structure on the location and composition of high-alumina basalts of the Taupo Volcanic Zone, New Zealand. N. Z. J. Geol. Geophys. 50, 327–342. Hildreth, W., 2004. Volcanological perspectives on Long Valley, Mammoth Mountain, and Mono Craters: several contiguous but discrete systems. J. Volcanol. Geotherm. Res. 136, 169–198. Hildreth, W., Wilson, C.J.N., 2007. Compositional Zoning of the Bishop Tuff. J. Petrol. 48, 951–999. Hildreth, W., Christiansen, R.L., O’Neil, J.R., 1984. Catastrophic isotopic modification of rhyolitic magma at times of caldera subsidence, Yellowstone Plateau Volcanic Field. J. Geophys. Res. 89, 8339–8369. Hill, D.P., 2006. Unrest in Long Valley Caldera, California, 1978–2004. Geol. Soc. Lond. Spec. Publ. 269, 1–24. Hogg, A.G., Higham, T.F.G., Lowe, D.J., Palmer, J.G., Reimer, P.J., Newnham, R.M., 2003. A wiggle-match date for Polynesian settlement of New Zealand. Antiquity 77, 116–125. Huang, H.-H., Lin, F.-C., Schmandt, B., Farrell, J., Smith, R.B., Tsai, V.C., 2015. The Yellow- stone magmatic system from the mantle plume to the upper crust. Science 348, 773–776. Jaxybulatov, K., Shapiro, N.M., Koulakov, I., Mordret, A., Landes, M., Sens-Schonfelder, C., 2014. A large magmatic sill complex beneath the Toba caldera. Science 346, 617–619. Kawasaki, N., Simon, S.B., Grossman, L., Sakamoto, N., Yurimoto, H., 2018. Crystal growth and disequilibrium distribution of oxygen isotopes in an igneous Ca-Al-rich inclusion from the Allende carbonaceous chondrite. Geochim. Cosmochim. Acta 221, 318–341. https://doi.org/10.1016/j.jvolgeores.2021.107430 https://doi.org/10.1016/j.jvolgeores.2021.107430 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0005 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0005 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0005 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0010 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0010 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0015 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0015 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0020 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0020 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0025 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0025 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0030 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0030 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0035 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0035 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0035 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0040 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0040 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0040 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0045 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0045 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0050 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0050 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0050 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0055 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0055 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0060 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0060 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0060 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0065 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0065 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0065 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0070 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0070 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0070 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0075 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0075 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0075 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0080 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0080 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0080 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0085 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0085 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0085 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0090 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0090 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0095 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0095 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0095 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0100 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0100 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0100 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0105 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0105 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0105 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0105 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0110 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0110 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0110 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0115 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0115 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0115 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0115 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0120 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0120 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0120 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0125 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0125 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0130 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0130 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0130 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0135 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0135 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0135 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0140 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0140 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0140 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0140 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0145 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0145 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0150 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0150 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0150 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0150 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0155 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0155 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0155 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0160 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0160 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0160 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0165 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0165 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0165 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0170 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0170 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0170 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0170 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0175 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0175 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0175 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0180 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0180 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0180 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0185 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0185 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0185 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0190 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0190 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0190 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0195 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0195 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0200 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0200 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0200 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0205 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0205 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0205 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0210 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0210 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0210 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0215 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0215 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0220 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0220 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0220 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0225 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0225 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0230 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0230 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0235 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0235 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0235 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0240 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0245 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0245 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0245 M. Sas, P. Shane, N. Kawasaki et al. Journal of Volcanology and Geothermal Research 421 (2022) 107430 Kobayashi, T., Nairn, I., Smith, V., Shane, P., 2005. Proximal stratigraphy and event se- quence of the c. 5600 cal. yr BP Whakatane rhyolite eruption episode from Haroharo volcano, Okataina Volcanic Centre, New Zealand. N. Z. J. Geol. Geophys. 48, 471–490. Kusakabe, M., Matsuhisa, Y., 2008. Oxygen three-isotope ratios of silicate reference mate- rials determined by direct comparison with VSMOW-oxygen. Geochem. J. 42, 309–317. Leonard, G.S., Cole, J.W., Nairn, I.A., Self, S., 2002. Basalt triggering of the c. AD 1305 Kaharoa rhyolite eruption, Tarawera volcanic complex, New Zealand. J. Volcanol. Geotherm. Res. 115, 461–486. Leonard, G.S., Begg, J.G.,Wilson, C.J.N., 2010. Geology of the Rotorua Area. Institute of Geo- logical and Nuclear Sciences 1:250,000, Geological Map 5 (1 sheet + 102 p). Lowe, D.J., Shane, P.A.R., Alloway, B.V., Newnham, R.M., 2008. Fingerprints and age models for widespread New Zealand tephra marker beds erupted since 30,000 years ago: a framework for NZ-INTIMATE. Quat. Sci. Rev. 27, 95–126. Macpherson, C.G., Gamble, J.A., Mattey, D.P., 1998. Oxygen isotope geochemistry of lavas from an oceanic to continental arc transition, Kermadec–Hikurangi margin, SW Pa- cific. Earth Planet. Sci. Lett. 160, 609–621. Matthews, N.E., Huber, C., Pyle, D.M., Smith, V.C., 2012. Timescales ofmagma recharge and reactivation of large silicic systems from Ti diffusion in quartz. J. Petrol. 53, 1385–1416. McCulloch, M.T., Kyser, T.K., Woodhead, J.D., Kinsley, L., 1994. Pb-Sr-Nd-O isotopic con- straints on the origin of rhyolites from the Taupo Volcanic Zone of New Zealand: ev- idence for assimilation followed by fractionation from basalt. Contrib. Mineral. Petrol. 115, 303–312. Milicich, S.D., Mortimer, N., Villamor, P., Wilson, C.J.N., Sagar, M.W., Ireland, T.R., 2021. The Mesozoic terrane boundary beneath the Taupo Volcanic Zone, New Zealand, and po- tential controls on geothermal system characteristics. N. Z. J. Geol. Geophys. 64, 518–529. Molloy, C., Shane, P., Nairn, I., 2008. Pre-eruption thermal rejuvenation and stirring of a partly crystalline rhyolite pluton revealed by the earthquake flat pyroclastics deposits, New Zealand. J. Geol. Soc. 165, 435–447. Nairn, I.A., 2002. Geology of the Okataina Volcanic Centre. Nairn, I.A., Kohn, B.P., 1973. Relation of the Earthquake Flat Breccia to the Rotoiti Breccia, central North Island, New Zealand. N. Z. J. Geol. Geophys. 16, 269–279. Nairn, I.A., Shane, P.R., Cole, J.W., Leonard, G.J., Self, S., Pearson, N., 2004. Rhyolite magma processes of the ∼AD 1315 Kaharoa eruption episode, Tarawera volcano, New Zealand. J. Volcanol. Geotherm. Res. 131, 265–294. Price, R.C., Gamble, J.A., Smith, I.E.M., Maas, R., Waight, T., Stewart, R.B., 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. Price, R., Mortimer, N., Smith, I., Maas, R., 2015. Whole-rock geochemical reference data for Torlesse and Waipapa terranes, North Island, New Zealand. N. Z. J. Geol. Geophys. 58, 213–228. Reid, F., 1983. Origin of the rhyolitic rocks of the Taupo Volcanic Zone, New Zealand. J. Volcanol. Geotherm. Res. 15, 315–338. Rubin, A.E., Cooper, K.M., Till, C.B., Kent, A.J., Costa, F., Bose, M., Gravley, D., Deering, C., Cole, J., 2017. Rapid cooling and cold storage in a silicic magma reservoir recorded in individual crystals. Science 356, 1154–1156. Sahetapy-Engel, S., Self, S., Carey, R.J., Nairn, I.A., 2014. Deposition and generation of mul- tiple widespread fall units from the c. AD 1314 Kaharoa rhyolitic eruption, Tarawera, New Zealand. Bull. Volcanol. 76, 836. Sakaguchi, K., Gilbert, H., Zandt, G., 2006. Converted wave imaging of the Toba Caldera, Indonesia. Geophys. Res. Lett. 33, L20305. Sas, M., Kawasaki, N., Sakamoto, N., Shane, P., Zellmer, G.F., Kent, A.J.R., Yurimoto, H., 2019. The ion microprobe as a tool for obtaining strontium isotopes in magmatic plagio- clase: a case study at Okataina Volcanic Centre, New Zealand. Chem. Geol. 513, 153–166. Sas, M., Shane, P., Kuritani, T., Zellmer, G.F., Kent, A.J.R., Nakagawa, M., 2021. Mush, melts and metasediments: a history of rhyolites from Okataina Volcanic Centre, New Zealand, as captured in plagioclase. J. Petrol. 62. Schmitz, M.D., Smith, I.E.M., 2004. The petrology of the rotoiti eruption Sequence, Taupo Volcanic Zone: an example of fractionation and mixing in a rhyolitic system. J. Petrol. 45, 2045–2066. Seebeck, H., Nicol, A., Stern, T.A., Bibby, H.M., Stagpoole, V., 2010. Fault controls on the ge- ometry and location of the Okataina Caldera, Taupo Volcanic Zone, New Zealand. J. Volcanol. Geotherm. Res. 190, 136–151. Shane, P.A.R., 2015. Contrasting plagioclase textures and geochemistry in response to magma dynamics in an intra-caldera rhyolite system, Okataina volcano. J. Volcanol. Geotherm. Res. 297, 1–10. Shane, P.A.R., Smith, V.C., 2013. Using amphibole crystals to reconstruct magma storage temperatures and pressures for the post-caldera collapse volcanism at Okataina vol- cano. Lithos 156–159, 159–170. Shane, P.A.R., Nairn, I.A., Smith, V.C., 2005. Magmamingling in the ∼50 ka Rotoiti eruption from Okataina Volcanic Centre: implications for geochemical diversity and chronol- ogy of large volume rhyolites. J. Volcanol. Geotherm. Res. 139, 295–313. 15 Shane, P.A.R., Nairn, I.A., Smith, V.C., Darragh, M., Beggs, K., Cole, J.W., 2008a. Silicic re- charge of multiple rhyolite magmas by basaltic intrusion during the 22.6 ka Okareka Eruption Episode, New Zealand. Lithos 103, 527–549. Shane, P.A.R., Smith, V.C., Nairn, I.A., 2008b. Millennial timescale resolution of rhyolite magma recharge at Tarawera volcano: insights from quartz chemistry andmelt inclu- sions. Contrib. Mineral. Petrol. 156, 397–411. Shane, P., Storm, S., Schmitt, A.K., Lindsay, J.M., 2012. Timing and conditions of formation of granitoid clasts erupted in recent pyroclastic deposits from Tarawera Volcano (New Zealand). Lithos 140–141, 1–10. Sharp, Z.D., 1990. A laser-based microanalytical method for the in situ determination of oxygen isotope ratios of silicates and oxides. Geochim. Cosmochim. Acta 54, 1353–1357. Smith, V.C., Shane, P.A.R., Nairn, I.A., 2005. Trends in rhyolite geochemistry, mineralogy, and magma storage during the last 50 kyr at Okataina and Taupo volcanic centres, Taupo Volcanic Zone, New Zealand. J. Volcanol. Geotherm. Res. 148, 372–406. Smith, V.C., Shane, P.A.R., Nairn, I.A., Williams, C.M., 2006. Geochemistry and magmatic properties of eruption episodes from Haroharo linear vent zone, Okataina Volcanic Centre, New Zealand during the last 10 kyr. Bull. Volcanol. 69, 57–88. Smith, V.C., Shane, P.A.R., Nairn, I.A., 2010. Insights into silicic melt generation using pla- gioclase, quartz and melt inclusions from the caldera-forming Rotoiti eruption, Taupo volcanic zone, New Zealand. Contrib. Mineral. Petrol. 160, 951–971. Spörli, K.B., 1978. Mesozoic tectonics, North Island, New Zealand. Geol. Soc. Am. Bull. 89, 415–425. Stern, T., Benson, A., 2011. Wide-angle seismic imaging beneath an andesitic arc: Central North Island, New Zealand. J. Geophys. Res. 116 (B09308), 1–26. Storm, S., Shane, P.A.R., Schmitt, A.K., Lindsay, J.M., 2011. Contrasting punctuated zircon growth in two syn-erupted rhyolitemagmas from Tarawera volcano: Insights to crys- tal diversity in magmatic systems. Earth Planet. Sci. Lett. 301, 511–520. Storm, S., Shane, P.A.R., Schmitt, A.K., Lindsay, J.M., 2012. Decoupled crystallization and eruption histories of the rhyolite magmatic system at Tarawera volcano revealed by zircon ages and growth rates. Contrib. Mineral. Petrol. 163, 505–519. Storm, S., Schmitt, A.K., Shane, P.A.R., Lindsay, J.M., 2014. Zircon trace element chemistry at sub-micrometer resolution for Tarawera volcano, New Zealand, and implications for rhyolite magma evolution. Contrib. Mineral. Petrol. 167, 1000. Taylor, C.B., Fractions, H.J., Nairn, I.A., 1977. Preliminary measurements of tritium, deute- rium and oxygen-18 in lakes and groundwater of volcanic Rotorua region, New Zealand. Inst. Geol. Nucl. Sci. INS-R–227. Troll, V.R., Deegan, F.M., Jolis, E.M., Harris, C., Chadwick, J.P., Gertisser, R., Schwarzkopf, L.M., Borisova, A.Y., Bindeman, I.N., Sumarti, S., Preece, K., 2013. Magmatic differenti- ation processes at Merapi Volcano: inclusion petrology and oxygen isotopes. J. Volcanol. Geotherm. Res. 261, 38–49. Waight, T.E., Troll, V.R., Gamble, J.A., Price, R.C., Chadwick, J.P., 2017. Hf isotope evidence for variable slab input and crustal addition in basalts and andesites of the Taupo Vol- canic Zone, New Zealand. Lithos 284–285, 222–236. Wallace, L.M., Beavan, J., McCaffrey, R., Darby, D., 2004. Subduction zone coupling and tec- tonic block rotations in the North Island, New Zealand. J. Geophys. Res. 109, B12406. Wallace, L.M., Reyners, M., Cochran, U., Bannister, S., Barnes, P.M., Berryman, K., Downes, G., Eberhart-Phillips, D., Fagereng, A., Ellis, S., Nicol, A., McCaffrey, R., Beavan, R.J., Henrys, S., Sutherland, R., Barker, D.H.N., Litchfield, N., Townend, J., Robinson, R., Bell, R., Wilson, K., Power, W., 2009. Characterizing the seismogenic zone of a major plate boundary subduction thrust: Hikurangi Margin, New Zealand. Geochem. Geophys. Geosyst. 10, Q10006. Watts, K.E., Bindeman, I.N., Schmitt, A.K., 2011. Large-volume Rhyolite Genesis in Caldera Complexes of the Snake River Plain: insights from the Kilgore Tuff of the Heise Volca- nic Field, Idaho, with Comparison to Yellowstone and Bruneau–Jarbidge Rhyolites. J. Petrol. 52, 857–890. Wilcock, J., Goff, F., Minarik, W.G., Stix, J., 2013. Magmatic recharge during the formation and resurgence of the Valles Caldera, New Mexico, USA: evidence from quartz com- positional zoning and geothermometry. J. Petrol. 54, 635–664. Wilson, C.J.N., Rowland, J.V., 2016. The volcanic, magmatic and tectonic setting of the Taupo Volcanic Zone, New Zealand, reviewed from a geothermal perspective. Geothermics 59, 168–187. Wilson, C.J.N., Houghton, B.F., McWilliams, M.O., Landphere, M.A., Weaver, S.D., Briggs, R.M., 1995. Volcanic and structural evolution of Taupo Volcanic Zone, New Zealand: a review. J. Volcanol. Geotherm. Res. 68, 1–28. Wolff, J.A., Balsley, S.D., Gregory, R.T., 2002. Oxygen isotope disequilibrium between quartz and sanidine from the Bandelier Tuff, NewMexico, consistent with a short res- idence time of phenocrysts in rhyolitic magma. J. Volcanol. Geotherm. Res. 116, 119–135. Zellmer, G.F., Kimura, J.-I., Stirling, C.H., Lube, G., Shane, P.A., Iizuka, Y., 2020. Genesis of re- cent mafic magmatism in the Taupo Volcanic Zone, New Zealand: insights into the birth and death of very large volume rhyolitic systems? J. Petrol. 61, egaa027. Zoback, M.L., Anderson, R.E., Thompson, G.A., 1981. Cainozoic evolution of the state of stress and style of tectonism of the basin and range province of the Western United States. Philos. Trans. R. Soc. A Math. Phys. Eng. Sci. 300, 407–434. http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0250 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0250 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0250 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0255 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0255 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0255 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0260 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0260 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0260 http://refhub.elsevier.com/S0377-0273(21)00259-6/rf0265 http://refhub.elsevier.com/S0377-0273(2