Pleistocene to Holocene Growth of a Large

Upper Crustal Rhyolitic Magma Reservoir

beneath the Active Laguna del Maule Volcanic

Field, Central Chile

Nathan L. Andersen1*, Brad S. Singer1, Brian R. Jicha1, Brian L. Beard1,

Clark M. Johnson1 and Joseph M. Licciardi2

1Department of Geoscience, University of Wisconsin–Madison, Madison, WI 53706, USA; 2Department of Earth

Sciences, University of New Hampshire, Durham, NH 03824, USA

*Corresponding author. E-mail: nandersen@wisc.edu

Received June 17, 2016; Accepted January 26, 2017

ABSTRACT

The rear-arc Laguna del Maule volcanic field (LdM) in the Andean Southern Volcanic Zone, 36�S, is

among the most active latest Pleistocene–Holocene rhyolitic centers globally and has been inflating

at a rate of> 20 cm a–1 since 2007. At least 50 eruptions during the last 26 kyr allow for a thorough

interrogation of changes in the physical and chemical state of this large, �20 km diameter, silicic

system. Trace element concentrations and Sr, Pb and Th isotope ratios indicate that the mafic pre-

cursors to the LdM rhyolites result from mixing between partial melts of garnet-bearing mantleand crust in Th-excess and partial melts of garnet-free crust in U-excess. The 238U/230Th ratios of

the LdM lavas are decoupled from the slab fluid signature, similar to several recently studied fron-

tal arc volcanic centers in the Southern Volcanic Zone. A narrow range of radiogenic isotope com-

positions and increasing isotopic homogeneity with differentiation indicate that silicic magma is

generated by magma hybridization and crystallization in the upper crust with limited involvement

of older, radiogenic material. New 40Ar/39Ar and 36Cl ages reveal a wide footprint of silicic volcan-ism during the early post-glacial (25–19 ka) and Holocene (c. 8–2 ka) periods, but focused within a

single eruptive center during the interim period. Subtle temporal variations in trace element com-

positions and two-oxide temperatures indicate that these eruptions, issued from vents distributed

within a similar area, tapped at least two physically discrete rhyolite reservoirs. This compositional

distinction favors punctuated extraction and ephemeral storage of the erupted magma batches.

Frequent mafic recharge incubates this long-lived, growing shallow silicic magma reservoir above

the granite eutectic, which favors magma interactions over rejuvenation of near- to sub-solidussilicic cumulates. A long-term rate of mass addition—extrapolated from surface deformation accu-

mulated over the past decade—is comparable with those that have produced moderate- to large-

volume caldera-forming eruptions elsewhere.

Key words: rhyolite; Andes Southern Volcanic Zone; magma chamber; geochronology; radiogenicisotopes

INTRODUCTION

Large silicic volcanic systems are of great interest because

they generate caldera-forming eruptions that disperse

enormous quantities of ash over a vast area.

Heterogeneities in the resulting pyroclastic fall and flow

(ignimbrite) deposits are often interpreted to reflect the

structure of the pre-eruption magma reservoir (e.g.Hildreth, 1981). The composition and ages of major and

accessory phases can provide records of magma accumu-

lation, crystallization, and mixing on both short (100–102

VC The Author 2017. Published by Oxford University Press. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com 85

J O U R N A L O F

P E T R O L O G Y

Journal of Petrology, 2017, Vol. 58, No. 1, 85–114

doi: 10.1093/petrology/egx006

Advance Access Publication Date: 15 March 2017

Original Article

year) and long (104–106 year) timescales (e.g. Vazquez &

Reid, 2004; Charlier et al., 2005, 2008; Wark et al., 2007;

Costa, 2008; Reid, 2008; Reid et al., 2011; Wotzlaw et al.,

2013, 2015; Chamberlain et al., 2014a, 2014b).

Complementing these records are studies of smaller pre-

and post-caldera silicic eruptions that record the longerthermochemical context that produced the caldera-

forming system, particularly when the earlier or subse-

quently erupted material is physically distinct from the

caldera-forming system or the caldera-collapse event pro-

duces a structural realignment of the shallow magma sys-

tem (Metz & Mahood, 1985, 1991; Sutton et al., 2000;

Charlier et al., 2005; Smith et al., 2005, 2010; Simon et al.,

2007; Wilson & Charlier, 2009; Bachmann et al., 2012;

Barker et al., 2015).The archetype model of voluminous silicic magma

systems involves crystallization of mafic to intermediate

forerunners in the middle to upper crust, yielding an

intermediate to silicic crystal mush—an extensive

crystal-rich (>60% solid) reservoir containing evolved

interstitial melt. Crystal-poor eruptible magma bodies

are assembled by progressive extraction and accumula-

tion of melt from these crystal-rich domains (Bachmann& Bergantz, 2004; Hildreth, 2004) or remelting of silicic

cumulate during magma recharge events (Mahood,

1990; Wolff et al., 2015; Evans et al., 2016). The relative

importance of these mechanisms varies between

caldera-forming systems as well as within zoned ignim-

brites produced during individual events (e.g. Vazquez

& Reid, 2004; Charlier et al., 2005; Bindeman et al.,

2008; Wotzlaw et al., 2013, 2015; Chamberlain et al.,2014a, 2014b; Evans et al., 2016).

Departures from the model of progressive rhyolite

extraction have been noted at large silicic systems such

as Taupo Volcano and Yellowstone involving a greater

proportion of remelting of silicic forerunners and the

amalgamation of distinct rhyolite melts, potentially cat-

alyzed by extensional tectonics (Smith et al., 2004, 2010;

Charlier et al., 2005, 2008; Wilson et al., 2006; Shaneet al., 2007, 2008; Bindeman et al., 2008; Wilson &

Charlier, 2009; Allan et al., 2013; Begue et al., 2014;

Storm et al., 2014). Brief repose periods following the

eruption of compositionally distinct pre-caldera rhyo-

lites, durations of zircon crystallization, and crystal resi-

dence based on solid-state diffusion kinetics indicate

that the assembly of 102–103 km3 eruptible rhyolite

magma bodies in these systems occurred more rapidly

than predicted by models of progressive melt extraction(Charlier et al., 2008; Allan et al., 2013; Bindeman &

Simakin, 2014; Wotzlaw et al., 2015). Thus, understand-

ing the mechanisms of rhyolite genesis in a particular

system can inform predictions of the processes and

timescales of the formation of a future, potentially large

eruptible silicic magma body.

The importance of lower crustal differentiation in

producing basalts and andesites in arc settings is wellrecognized (e.g. Hildreth & Moorbath, 1988; Ownby

et al., 2011); it has also been proposed that silicic

magma is generated in the lower crust by partial

melting of the deep crust (up to 70% depending on the

magma flux and lithology of the crust), fractional crys-

tallization of hydrous basalt, and mixing of the resulting

differentiates and crustal melts. Shallow systems are

assembled incrementally from these lower crustal ‘hot

zones’ (Annen et al., 2006), but undergo limited chem-ical differentiation following shallow magma emplace-

ment. Thus, the volume of eruptible magma is primarily

a function of the magma flux to the upper crust (e.g.

Glazner et al., 2004; Annen et al., 2006; Annen, 2009;

Gelman et al., 2013).

The investigation of pre-caldera silicic eruptions can

provide clues to the physical and thermal evolution that

sets the stage for the assembly and eruption of a volu-minous silicic magma reservoir. Pre-caldera eruptive re-

cords can be limited owing to infrequent eruptions,

poorly resolved geochronology, burial or destruction by

subsequent caldera-forming events (Metz & Mahood,

1991; Stix & Gorton, 1993; Wilson et al., 2009).

Nevertheless, such records have proven useful in iden-

tifying changes in the mafic flux to the upper crust, the

amalgamation of previously discrete magma reservoirs,

and placing limits on the longevity of the subsequentcaldera-forming reservoir (e.g. Metz & Mahood, 1991;

Simon et al., 2007; Bindeman et al., 2008; Wilson &

Charlier, 2009; Chamberlain et al., 2014b).

Understanding the recent magmatism at historically ac-

tive rhyolitic volcanic centers (e.g. Miller, 1985; Hildreth,

2004; Smith et al., 2005; Castro & Dingwell, 2009;

Hildreth & Fierstein, 2012; Rawson et al., 2015) allows

for the interrogation of the structure of the magma res-ervoir, the petrogenesis of rhyolites, the physical and

thermal processes preceding the recent eruptions, and

their evolution through time. Such systems are poten-

tial sites of caldera-forming eruptions and, taken to-

gether, this information is valuable in evaluating the

possible style of future eruptions and establishing a

context in which to better interpret seismic, magnetotel-

luric, geodetic, and gravity observations (e.g. Singeret al., 2014).

The rear-arc Laguna del Maule (LdM) volcanic field

(Fig. 1) produced two dacitic to rhyodacitic caldera-

forming eruptions during the mid-Pleistocene. A recent

concentration of silicic volcanism has yielded at least 50

rhyolitic eruptions in the last 26 kyr; thus LdM is among

the most frequently erupting active rhyolitic volcanic

centers globally (Hildreth et al., 2010; Fierstein et al.,

2012; Sruoga, 2015). This remarkable spatial and tem-poral concentration of rhyolite eruptions since the last

glacial maximum, locally dated at c. 24 ka based on the

age of glaciated and unglaciated lava flows at LdM

(Singer et al., 2000), has encircled the lake in the central

LdM basin and is unprecedented in the southern Andes

(Fig. 2; see also Table 1; Hildreth et al., 2010; Singer

et al., 2014). Hildreth et al. (2010) presented several lines

of evidence suggesting that these eruptions are derivedfrom an integrated silicic magma system, most promin-

ently: (1) rhyolite lavas erupted 10–12 km apart have

nearly identical major and trace element compositions,

86 Journal of Petrology, 2017, Vol. 58, No. 1

suggesting that they are derived from a single homoge-

neous reservoir; (2) inclusions of mafic magma in rhyo-

dacite lavas are common, whereas mafic eruptions

have been rare and peripheral since the beginning of

post-glacial rhyolite volcanism, indicating that a broad,

low-density magma body is blocking the ascent of

mafic magma. Consequently, the numerous post-

glacial silicic eruptions at LdM may represent a high

temporal resolution sampling of the evolution of a

large, shallow magma system.

Several geophysical methods document continuing

volcanic unrest within the LdM basin that remains ac-

tive at the time of this writing. Geodetic data since 2007,

obtained by continuous global positioning system

(GPS) and interferometric synthetic aperture radar

(InSAR), record uplift at a rate in excess of 20 cm a–1,

among the fastest measured at a volcano not actively

erupting (Fournier et al., 2010; Feigl et al., 2014; Le

Mevel et al., 2015). A model of an inflating sill at 5 km

depth produces the best fit of the measured deform-

ation pattern, with an estimated volume increase of

0�03–0�05 km3 a–1 between 2007 and 2014 (Le Mevel

et al., 2016). This probably transient rate is one to two

orders of magnitude greater than the late Pleistocene to

Holocene eruptive fluxes at the Southern Volcanic Zone

(SVZ) frontal arc centers Mocho–Choshuenco and

Puyehue–Cord�on Caulle (Singer et al., 2008; Rawson

et al., 2015) and the average eruptive flux at LdM over

the last 1�5 Myr (Hildreth et al., 2010). During the same

period of time, frequent seismic swarms have occurred

at similarly shallow depths near the Nieblas (rln) and

Barrancas (rcb) rhyolite flows, which are among the

youngest in the volcanic field (Fig. 2; Singer et al.,

2014). Initial gravity and magnetotelluric studies also

suggest the presence of a shallow, possibly growing,

magma system beneath the area of deformation at LdM

(Singer et al., 2014; Miller et al., 2016). More recent geo-

detic and geomorphological observations indicate that

the rate of uplift and inflation slowed slightly in 2013 (Le

Mevel et al., 2015) and that dozens of similar inflation

episodes have probably occurred throughout the

Holocene (Singer et al., 2015).

The post-glacial eruptive chronology at LdM is cur-

rently defined by only four 40Ar/39Ar ages obtained

nearly two decades ago (Singer et al., 2000) and the

positions of lava flows relative to a paleoshoreline

marking the highstand of the lake produced when the

outlet gorge was dammed by the early rle rhyolite flow.

Consequently, the age relations of eruptions occurring

on opposite sides of the lake have been inferred based

only on geomorphological features such as the extent

of weathering and degree of pumice cover, hindering

the interpretation of the temporal record. New 40Ar/39Ar

and 36Cl surface exposure ages for late Pleistocene and

post-glacial LdM lavas that refine the eruptive sequence

are presented in this study. New whole-rock trace elem-

ent compositions, Sr, Pb, and Th isotope ratios, and

mineral thermobarometry are evaluated in the frame-

work of this new geochronology to examine the tem-

poral evolution of the rhyolite and rhyodacite magma

compositions. Models of magma evolution spanning

the last 150 kyr in the central LdM basin (earlier erup-

tions are sparse) are used to interrogate the continuity

and integration of the LdM magma system, the nature

and depth of the processes contributing to its evolution

through time, and the implications for the continuing

volcanic unrest.

GEOLOGICAL SETTING

The Quaternary LdM volcanic field is situated on the

crest of the Andes at 36�S in the Southern Volcanic

Zone (SVZ) of central Chile (Fig. 1). Between 32 and

37�S, the arc is characterized by a gradient in crustal

thickness from 35 km in the south to 60 km in the north

(Gilbert et al., 2006; Tassara et al., 2006; Tassara &

Echaurren, 2012). This near doubling in thickness cor-

relates with a transition from dominantly basaltic an-

desite to amphibole-bearing intermediate products and

well-documented gradients in trace element and radio-

genic isotope composition (Hildreth & Moorbath, 1988).

Distinctively, the segment of the arc between 34 and

37�S hosts several large Quaternary silicic volcanic cen-

ters in addition to LdM: the Maipo–Diamante Caldera

(Sruoga et al., 2012), the Calabozos Caldera (Hildreth

et al., 1984; Grunder & Mahood, 1988), Puelche

Volcanic Field (Hildreth et al., 1999), and Domuyo

Osorno

Puyehue-Cordón Caulle

Lagunadel Maule

Talca

Santiago

Cerro Azul-Quizapu

CalabozosCaldera

Maipo-Diamante

San Jose

Tupungato

PuelcheVolcanicField

Tatara-San Pedro

Antuco

Llaima

Villarrica

Mocho-Choschuenco

Nevados de Chillan

Chi

le

Arg

enti

na

100 km

Lonquimay

-34º

-36º

-38º

-40º

-74º -72º -70º -68º

7.4 cm/yr

Per

u-C

hile

Tre

nch

Domuyo

Concepción

Fig. 1. Regional map of the SVZ between 33 and 41�S showingthe location of Laguna del Maule. Selected frontal arc volcanos(triangles) and caldera systems and silicic volcanic centers(dark gray fields) are labeled for reference. The velocity of theNazca plate relative to the South American plate is calculatedusing MORVEL (DeMets et al., 2010).

Journal of Petrology, 2017, Vol. 58, No. 1 87

2800

2700

2600

2200

230024

00

2500

2600

2300

2400

25002600

2700

28002900

28002700

2600

26002700

2000

2100

22002300

2400

250026002700

2800

2800

2900

3000

28002700

2600

2500

24002300

23002400

25002600

2700

28002900

3000

2900 3000

23002400

25002600

27002800

27002600

2500

240023002200

2100

2000

1900

170018001600

30002900

2800

27002600

25002400

2300

3000

2900

2500

2600

270027002700270028002800

27027

21002100

2300230024002400

25002500

2300

2300

00

2200220023002300

24002400

2500250026002600

27027

2222

27002700

26002600

25002500

2400240023002300

2

2600260025005025002

2400240

25002500

26002600

27002700

2800280028009009000

000000

230023002300

20002000

210021002100

2200

220025

0025

0024

0024

00

000000

00

30003000300000030003000030000300030290029002900290029002900290090029002

28002800280028002802800028002800282222

00000

22290029002900

300300333333

27002700

2800800280080

28002800280028000280000280080800802828282828

290029009002900900290990929000002900

300030003000300030003000300003000

2800800800082800

000000000000

Cajón Grande de Bobadilla

Cajón Chico de Bobadilla

rcl

rsl

rcn

igcb

igcb

igcb

rle rca

PasoPehuenche

rsl

bbc

3031

19.0 ka

25.7 ka

rdnorep

3.3 ka

rcd2.1 ka

mvc

ras22.4 ka rap

rlnrdep

rdct

mct

mcp

dlp

apj

acnanc

aandlp

rdcd

mnprdnp

asmigsp

asprdsp

rdcn3.5 ka

igsp bec

apvaam

mpl

rddmrdop

rdam

rle ig rdne

rdac5.6 ka

20.0 ka

apo

rcb-py

rng

rcb

rcb-d

rcb-d

Aroyo de PalaciosA.

Sepú

lveda

A. de la Calle

Laguna del

Maule2162

Laguna

L.Negra

L.Cari

Launa

N

Cajón d

e Tron

coso

Fea

14.5 ka

ram

Rio Maule

8.0 ka

1.9 ka

11.4 ka

Central Laguna del Maule Volcanic FieldPost Glacial Eruptions

RhyoliteRhyodaciteAndesite

Pleistocene EruptionsRhyoliteDacite/RhyodaciteAndesiteBasalt/Mafic AndesitePleistocene ignimbritesigcb - 990 ka; igsp - 1.5 Ma

PumiceHighway 115

Sample Locations3.5 Eruption age [ka]

International Boundary

Center of Deformation

Lava Flow DirectionVolcanic Vent

0 5km

Contour interval 50 mall elevations masl

-36.1º

-36.2º

º5.07-º6.07--70.4º

-36.0º

Arroyo Curamilio

Arroyo Puente de Tierra

Arro

yo de

la P

arva

CHILE

ARGENTINA

115

115

3175

31222828

3037

2994

2888

3080

2767

2883

2855

3056

2874

2889

2162

2486

2680

Fig. 2. Simplified geological map of the central basin of the LdM volcanic field [after Hildreth et al. (2010)] showing sample loca-tions; unit names and abbreviations are listed in Table 1. Eruption ages are determined by 40Ar/39Ar except for the 36Cl age of unitrdcd; uncertainties associated with the 40Ar/39Ar ages are given in Table 2 and 36Cl data are given in the Supplementary Data. Thecenter of uplift near the southwestern lake shore is an approximate location based on the InSAR model of Feigl et al. (2014).

88 Journal of Petrology, 2017, Vol. 58, No. 1

Volcanic Complex (Miranda et al., 2006; Chiodini et al.,

2014), each situated in the rear-arc relative to the basalt-to andesite-dominated frontal arc volcanoes (Fig. 1).

Owing to repeated glaciation and the remote, rugged

terrain, it is not well appreciated that the productivity of

Pliocene to Holocene silicic volcanism in this northern

sector of the SVZ is comparable with that of the Andean

Central Volcanic Zone (Hildreth et al., 1984, 1999).Hildreth et al. (2010) documented the most recent 1�5

Myr of volcanic activity at LdM, which comprises more

than 350 km3 of lava, tephra, and pyroclastic deposits

ranging in composition from basalt to high-silica rhyo-

lite erupted from at least 130 vents. The Quaternary

eruptions overlie Paleogene to Neogene volcanic and

volcaniclastic rocks and Pliocene to Mesozoic plutonsand sedimentary strata (Nelson et al., 1999; Hildreth

et al., 2010). LdM volcanic products are of tholeiitic to

calc-alkaline, medium- to high-K compositions typical

of SVZ frontal arc volcanoes. Hildreth et al. (2010) foundevidence for neither systematic variation in the slab sig-

nature across the volcanic field nor any significant con-

tribution of back-arc, alkaline compositions. Basaltic

andesite to andesite dominates much of the preserved

eruptive history of LdM, but silicic (dacite–rhyolite)

eruptions have occurred throughout the volcanic fieldduring the Pliocene and Pleistocene (Hildreth et al.,

2010). Two silicic ignimbrites are preserved in the LdM

lake basin (Fig. 2), the 1�5 Ma two-pyroxene dacite Sin

Puerto Ignimbrite (igsp) and the 990 ka biotite rhyoda-

cite Bobadilla Ignimbrite (igcb) (Birsic, 2015). Of these,

only the Bobadilla caldera structure partially survived

the subsequent glaciation and erosion. Two middlePleistocene rhyolitic lavas are preserved near the north-

eastern shore of the lake, the 710 6 13 ka Rhyolite of

Table 1: Laguna del Maule eruptive units mapped in Fig. 2

Abbreviation* Unit name* Eruption age†

aam Andesite of Arroyo Los Mellicos 25�4 6 1�5 kaacn Andesite of Crater Negro post-glacialanc Andesite north of Crater Negro post-glacialapj Younger andesite of West Peninsula 21�1 6 3�4 kaapv Older andesite of West Penisula pre-glacialasm Andesite south of Arroyo Los Mellicos post-glacialasp Andesite of Laguna Sin Puerto <3�5 kabbc Basalt of Volc�an Bobadilla Chica 153 6 7 kabec Basalt of El Candado 61�8 6 3�6 kadlp Dacite of Laguna del Piojo pre-glacialigcb Ignimbrite of Cajones de Bobadilla (rhyodacite) 990 6 13 kaigsp Ignimbrite of Laguna Sin Puerto (dacite) 1484 6 15 kamcp Andesite of Crater 2657 post-glacialmct Andesite of Arroyo Cabeceras de Troncoso post-glacialmnp Andesite north of Estero Piojo post-glacialmpl Andesite of Volc�an Puente de la Laguna 54 6 21 kamvc Andesite of Volc�an de la Calle 152�1 6 6�5 karam Rhyolite of Arroyo Los Mellicos post-glacial; >19 karap Rhyolite of Arroyo de Palacios 22�4 6 2�0 karas Rhyolite of Arroyo de Sep�ulveda 19–20 karca Rhyolite of Caj�on Atravesado 710 6 13 karcb Rhyolite of Cerro Barrancas multiple flows; 11�4–1�9 karcb-d Cerro Barrancas Dome Complex (rhyolite) 14�5 6 1�5 karcb-py Cerro Barrancas Pyroclastic Flow (rhyolite) 11�4 6 1�1 karcd Rhyolite of Colada Divisoria 2�1 6 1�3 karcl Rhyolite of Cari Launa <3�3 karcn Rhyolite of Cerro Negro 466�0 6 5�6 kardac Rhyolite of Arroyo de la Calle 20�0 6 1�2 kardam Rhyodacite of Arroyo Los Mellicos post-glacial; >19 kardcd Rhyodacite of Colada Dendriforme 8�0 6 0�8 kardcn Rhyodacite of Northwest Coulee 3�5 6 2�3 kardct Rhyodacite of Arroyo Cabeceras de Troncoso 202 6 41 karddm Rhyodacite of Domo del Maule 114 6 14 kardne Rhyodacite NE of Loma de Los Espejos post-glacial; >19 kardno Rhyodacite NW of Loma de Los Espejos post-glacial; >19 kardnp Rhyodacite north of Estero Piojo post-glacialrdop Rhyodacite west of Presa Laguna del Maule pre-glacialrdsp Rhyodacite of Laguna Sin Puerto <3�5 karep Rhyolite east of Presa Laguna del Maule 25�7 6 1�2 karle Rholite of Loma de Los Espejos 19�0 6 0�7 karle-ig Espejos ignimbrite (rhyolite) post-glacial; >19 karln Rhyolite of Colada Las Nieblas Late Holocenersl Rhyolite south of Laguna Cari Launa 3�3 6 1�2 ka

*Abbreviations and unit names after Hildreth et al. (2010).†Ages are from Singer et al. (2000), Hildreth et al. (2010), Birsic (2015), and this study; all 40Ar/39Ar ages are calculated relative tothe 1�1864 Ma Alder Creek Sanidine (Jicha et al., 2016).

Journal of Petrology, 2017, Vol. 58, No. 1 89

Cajon Atravesado (rca) and the 466�0 6 5�6 ka Rhyolite

of Cerro Negro (rcn). The latter contains the most

evolved compositions in the volcanic field (Hildreth

et al., 2010).

Singer et al. (2000) determined the timing of the last

glacial retreat to be between 25�4 6 1�2 ka and 23�2 6 0�6ka based on 40Ar/39Ar age determinations (recalculated

to an Alder Creek Sanidine age of 1�1864 Ma; Jicha

et al., 2016) for four eruptions, including one glaciated

and three unglaciated lavas at approximately equal ele-

vation in the LdM basin. This age is consistent with the

moraine records east of the Andes between 47 and

46�S based on 3He, 10Be, and 26Al cosmogenic expos-

ure, 40Ar/39Ar, and 14C ages indicating that the last gla-cial maximum occurred prior to 23 ka with deglaciation

well under way by 16�5 ka (Kaplan et al., 2004; Hubbard

et al., 2005; Clark et al., 2009; Hein et al., 2010). The

post-glacial volcanism is concentrated in the LdM lake

basin, producing 36 silicic domes and coulees and doz-

ens of explosive eruptions from at least 24 vents encir-

cling the lake (Fig. 2; Hildreth et al., 2010; Fierstein et al.,

2012; Sruoga, 2015). Ten andesite flows emplaced since

the glacial retreat, primarily along the western lake-shore, are of subordinate volume. Basaltic andesite is

rare since the most recent deglaciation and the young-

est true basalt is the 61�8 6 3�6 ka basalt of El Candado

(bec) erupted north of LdM (Fig. 2; Hildreth et al., 2010,

recalculated to an Alder Creek Sanidine age of

1�1864 Ma; Jicha et al., 2016).

Silicic eruptions at LdM were explosive and effusive

and generally of modest volume (<1�3 km3; Hildrethet al., 2010; Fierstein et al., 2012). Continuing tephrostra-

tigraphic investigations (Fierstein et al., 2012; Sruoga,

2015) both within the LdM basin and of distal deposits

in Argentina, are not discussed in detail here. However,

of particular note, Fierstein et al. (2012) have identified a

voluminous explosive eruption that produced flow and

fall deposits up to 6 m thick in Argentina 30 km south

and east of LdM accounting for an order of magnitudegreater volume than any single event mapped in the

central basin by Hildreth et al. (2010). This explosive

event pre-dates the rle lava flow that dammed the lake

and thus is among the earliest post-glacial rhyolite

eruptions. However, its vent location and eruption age

remain uncertain.

Rhyolite flows preserved in the LdM basin are vitro-

phyric and carry �5% modal phenocrysts; the rhyoliteof Arroyo Palacios (rap) and all but the latest of the

Barrancas complex (rcb) flows are notably aphyric.

Phenocrysts, when present, are dominantly plagioclase,

subordinate biotite, Fe–Ti oxide, sparse quartz, acces-

sory zircon, apatite, and very rare FeS inclusions in

magnetite; several rhyolites also contain scarce amphi-

bole. With the exception of the rhyodacite of Arroyo de

la Calle (rdac) the rhyodacite lavas are concentrated in

the western and northwestern basin. They are vitrophy-ric to micro-pumiceous and nearly all carry a pheno-

cryst load of 10–25%, greater than any of the rhyolites;

only the rhyodacites of the Northwest Coulee (rdcn) and

Laguna Sin Puerto (rdsp) are crystal poor. The pheno-

cryst assemblage is similar to that of the rhyolites but

all lack quartz and contain amphibole. Most rhyodacite

lavas contain fine-grained, partly glassy, basaltic andes-

ite inclusions, frequently with quench textures, up to

40 cm in diameter in the rhyodacites of Colada

Dendriforme (rdcd) and NW of Loma de Los Espejos

(rdno), but more commonly 1–10 cm in diameter.

Similar inclusions are rare in the Rhyolite of Arroyo Los

Mellicos (ram) mini-dome but have not been found in

any other rhyolite.

NEW 40Ar/39Ar AND 36Cl AGES AND REVISEDERUPTION SEQUENCE

An effort to document the LdM eruptive sequence based

on the tephra stratigraphy and soil 14C ages is currently

under way (Fierstein et al., 2012; Sruoga, 2015).

However, the construction of a 14C chronology at LdM is

challenging owing to a dearth of organic material.

Whereas 14C ages typically have lower uncertainties,

where suitable material is lacking, 40Ar/39Ar and 36Cl

ages offer alternative methods to date young volcanic

eruptions. Twenty-six 40Ar/39Ar incremental heating ex-

periments, performed at the WiscAr Geochronology Lab

(see Supplementary Data for details; supplementary

data are available for downloading at http://www.pet

rology.oxfordjournals.org) yield plateau ages, all but one

containing more than 75% of the released 39Ar, and sup-

port 12 eruption ages (Fig. 3; Table 2). We attempted to

determine 40Ar/39Ar ages for nearly all post-glacial lavas.

However, owing to their youth and high atmospheric Ar

contents, LdM products commonly yield small fractions

of radiogenic 40Ar (40Ar*). Micropumiceous rhyodacites

and commonly vesiculated and glassy andesite flows

nearly all produced high 36Ar signals from which 40Ar*

could not be resolved. Dense rhyolitic obsidian more

commonly yields plateau ages; however, only approxi-

mately 50% of such samples produced resolvable ages.

Recoil of 39Ar during irradiation of volcanic glass can re-

sult in spurious ages. This effect is mitigated for the LdM

lavas by a short irradiation time; age plateaux character-

istic of recoil (i.e. decreasing apparent age with increas-

ing step heating temperature) are only sporadically

observed for sample aliquots subjected to longer dur-

ation irradiation (see Supplementary Data). Several ex-

periments display anomalously high ages in the low or

high temperature steps. However, this behavior is con-

sistent neither throughout the LdM sample suite, nor be-

tween aliquots prepared from single samples. The cause

of these discordant steps is not clear, but they account

for less than 5% of the gas in single experiments and do

not bias the reported ages. Inverse isochrons for all sam-

ples yield 36Ar/40Ar intercepts within uncertainty of the

atmospheric ratio of Lee et al. (2006), indicating that ex-

cess Ar is not significant. The isochron and plateau ages

for each experiment are indistinguishable at 2r uncer-

tainty; thus the more precise plateau ages are preferred.

90 Journal of Petrology, 2017, Vol. 58, No. 1

The eruption of the rle flow dammed the northern

outlet of the lake, causing the lake level to rise to

�200 m above its modern level and cutting a prominent

shoreline into all low-lying older rocks (Hildreth et al.,

2010; Singer et al., 2015). To constrain better the dur-

ation of the lake highstand, we undertook 36Cl surface

exposure age determinations of the rdcd lava flow,

which overruns the paleoshoreline is several places but

did not produce a resolvable 40Ar/39Ar age, and the

shoreline itself where it is notched into the igcb ignim-

brite along the north shore of the lake (36Cl methods

and results are in the Supplementary Data).

The new age determinations are discussed in con-

junction with the observations made during fieldwork

in support of the present work and by Hildreth et al.

(2010) to improve the chronology of the post-glacial

eruptions. Whereas LdM erupted regularly following

the last glacial maximum, the rhyolitic volcanism is

clustered in two periods of high eruption frequency.

An early post-glacial (EPG) group erupted prior to the

damming of the outlet gorge at 19 ka. This was fol-

lowed by a period of relative calm in much of the lake

basin during the latest Pleistocene, with rhyolitic activ-

ity limited to the Barrancas complex in the SE basin.

Finally, silicic eruptions encircled the lake during the

Holocene (Fig. 4).

Early post-glacial eruptionsThe earliest of the recent silicic units erupted shortly

prior to deglaciation, forming the rhyolite east of the

Presa (dam) (rep) at 25�7 6 1�2 ka in the northwestern

LdM basin. All subsequently erupted silicic units are

unglaciated, including the early voluminous pyroclastic

event (Fierstein et al., 2012) and numerous andesite and

rhyodacite flows and domes concentrated in the west-

ern and northwestern LdM basin. A single unglaciated

andesite flow (apo) erupted in the south; this glassy,

vesiculated lava did not produce resolvable 40Ar*. The

flow is largely buried by lake deposits and a pumice fan,

but apparently was erupted prior to the damming of the

lake. Rhyolite flows erupted on three sides of the lake in

a relatively short time interval at the end of the EPG

period: the Sep�ulveda rhyolite (ras) in the SE [which dir-

ectly overlies the 20�0 6 1�2 ka rhyodacite of Arroyo del

la Calle (rdac)], the 22�4 6 2�0 ka Palacios rhyolite (rap),

and the 19�0 6 0�7 ka Espejos rhyolite (rle).

Latest Pleistocene to Holocene eruptionsVolcanic activity waned throughout much of the LdM

basin following the end-EPG eruptions. The latest

Pleistocene eruptions were restricted to the Barrancas

center (rcb) on the southeastern rim of the lake basin.

An early episode of dome building is dated at 14�5 6 1�5

-10

0

10

20

30

40

0 0.2 0.4 0.6 0.8 1.0

Southern Cari Launa Rhyolite (rsl)3.3±1.2 ka

3.7±2.1 ka3.0±1.6 ka

0 0.2 0.4 0.6 0.8 1.0Cumulative 39Ar Fraction

0

20

30

40

50

60

10

Rhyolite East of Presa Laguna del Maule (rep)

25.7±1.2 ka

1.0

1.5

2.0

2.5

3.0

3.5

4.0

0.0 0.5 1.0 1.5 2.0 2.5 3.039Ar/40Ar

26.2±2.6 ka40Ar/39Ar0 = 296.5±9.8

n = 13

25.8±1.3 ka

25.0±3.0 ka

2.0

2.5

3.0

3.5

4.0

0.0 0.4 0.8 1.2 1.6 2.0 2.4

5.3±2.9 ka40Ar/39Ar0 = 293.8±8.6

n = 10

0.0

1.0

2.0

3.0

4.0

5.0

0 3 6 9

19.1±0.8 ka40Ar/39Ar0=296.4±2.6

n=18

0

10

20

30

40

50

0 0.2 0.4 0.6 0.8 1.0

Ag

e [k

a]

Espejos Rhyolite (rle)19.0±0.7 ka

17.7±2.5 ka

19.5±0.9 ka18.4±1.1 ka

36A

r/40

Ar

x 10

3

Fig. 3. Example 40Ar/39Ar age spectra and inverse isochrons for units rep, rle and rsl; values for all samples are available in theSupplementary Data. Plateau steps are colored boxes and ellipses; discordant, excluded steps are light gray. All uncertainties are-6 2r and include the analytical and J uncertainties.

Journal of Petrology, 2017, Vol. 58, No. 1 91

ka and is followed by an explosive event that produced

pyroclastic flow deposits extending SE away from the

lake into Argentina (Fig. 2). A dense vitric clast from thispyroclastic deposit gave an age of 11�4 6 1�1 ka. These

earliest products of the Barrancas complex are exposed

on its southern and eastern flanks and, therefore, are

not subject to shoreline erosion. Continued activity at

the Barrancas complex produced a series of rhyolite

flows, the northernmost of which, along with the rdcd

rhyodacite flow, erupted near the western lake shore;

these are the youngest units at sufficiently low elevation

within the lake basin to be subject to, but not affectedby, shoreline erosion. The youngest of the three north-

ern rcb flows yields an 40Ar/39Ar age of 5�6 6 1�1 ka;40Ar* could not be resolved from either of the underly-

ing flows. The rdcd flow yields a whole-rock 36Cl surface

exposure age of 8�0 6 0�8 ka. The ages of the rdcd and

northern rcb flows are consistent with a whole-rock 36Cl

Table 2: Summary of 40Ar/39Ar experiments

Sample no. K/Ca total Total fusionage [ka] 62r

40Ar/36Ari 6 2r MSWD Isochronage [ka] 62r

n 39Ar % MSWD Plateauage [ka] 62r

Rhyolite of Cerro Barrancas, eat summit flow (rcb)AR-267 5�81 2�4 6 2�1 301�5 6 6�1 0�99 8�7 6 5�8 6 of 9 84�2 1�03 1�6 6 0�7AR-267 5�80 4�4 6 3�1 323 6 33 0�24 -22�4 6 9�9 5 of 10 76�8 0�85 2�3 6 0�9

Combined isochron n¼19: 301�0 6 5�7 1�00 1�3 6 8�0 Weighted mean n 5 2: 1�7 1�9 6 0�6Rhyolite of Colada Divisoria (rcd)LdM-249* 5�06 7�7 6 3�2 305�2 6 8�1 0�12 -1�8 6 3�9 6 of 10 83�2 0�65 2�8 6 2�3LdM-249* 5�18 3�2 6 2�7 300�7 6 5�1 0�19 -0�5 6 1�0 7 of 8 98�4 0�27 1�3 6 2�5LdM-249 5�00 2�5 6 2�1 300�2 6 5�6 0�08 0�9 6 2�7 6 of 6 100�0 0�13 2�2 6 1�9

Combined isochron n¼19: 301�1 6 3�4 0�28 0�3 6 0�3 Weighted mean n 5 3: 0�35 2�1 6 1�3Rhyolite of South Cari Launa (rsl)ALDM-13-17 6�70 1�7 6 1�8 285�6 6 16�9 1�16 8�6 6 5�7 4 of 7 80�1 1�60 3�7 6 2�1ALDM-13-17 6�60 3�0 6 1�6 298�2 6 11�4 0�25 3�2 6 4�1 6 of 8 96�2 0�20 3�0 6 1�6

Combined isochron n¼10: 293�8 6 8�6 0�63 5�3 6 2�9 Weighted mean n 5 2: 0�65 3�3 6 1�2Rhyodacite of the Northwest Coulee (rdcn)LdM-12-27 1�46 -1�2 6 2�4 294�7 6 8�4 1�54 5�7 6 4�5 5 of 7 89�7 1�38 3�5 6 2�3Rhyolite of Cerro Barrancas, northern flow (rcb)LdM-210† 5�31 7�3 6 1�9 308�8 6 10�5 1�41 -2�2 6 2�4 5 of 6 98�2 2�43 5�2 6 2�7LdM-210 5�25 3�9 6 2�4 298�7 6 16�2 2�95 2�6 6 2�9 5 of 6 97�9 2�18 2�7 6 3�1LdM-210* 5�20 10�6 6 2�8 299�1 6 8�5 0�40 8�5 6 5�9 8 of 9 97�6 0�34 9�0 6 2�4LdM-210* 5�26 8�2 6 3�2 301�5 6 12�7 0�90 3�1 6 2�7 7 of 9 86�1 0�77 4�9 6 3�0LdM-210 5�12 5�6 6 1�4 331�0 6 37�9 0�63 1�9 6 1�9 7 of 7 100�0 1�14 5�7 6 1�2

Combined isochron n¼27: 298�7 6 3�2 1�57 5�6 6 1�3 Weighted mean n 5 4: 1�47 5�6 6 1�1Cerro Barrancas Pyroclastic Flow (rcb-py)CB-Curamilo A 6�18 12�6 6 4�9 318 6 8�6 0�47 -8�8 6 8�6 12 of 15 84�1 1�2 11�5 61�3CB-Curamilo A 5�88 10�0 6 5�1 294�9 6 5�8 1�3 16 6 10 5 of 10 67�3 1�4 11�361�9

Combined isochron n¼18: 298�7 6 4�8 1�2 11�4 6 7�1 Weighted mean n 5 2: 0�03 11�4 6 1�1Cerro Barrancas Dome Complex (rcb-d)CB-2 4�90 13�7 6 1�6 297�4 6 2�7 0�80 15�8 6 3�6 9 of 9 100�0 0�80 14�5 6 1�5Rhyolite of Loma de Los Espejos (rle)LdM-60 6�90 20�7 6 2�5 295�1 6 10�0 1�31 20�3 6 7�6 6 of 8 90�9 1�17 17�7 6 2�5LdM-60 7�10 17�8 6 1�2 204�5 6 116�9 0�55 20�7 6 1�6 4 of 6 91�2 0�93 19�5 6 0�9LdM-60 2�00 18�2 6 1�7 299�2 6 5�0 0�23 18�3 6 1�6 8 of 9 99�3 0�21 18�4 6 1�1

Combined isochron n¼18: 296�4 6 2�6 0�73 19�1 6 0�8 Weighted mean n 5 3: 0�84 19�0 6 0�7Rhyodacite of Arroyo de la Calle (rdac)LdM-213 2�05 20�9 6 1�8 294�1 6 9�7 0�59 22�6 6 3�3 7 of 7 100�0 0�62 21�2 6 1�5LdM-213 2�03 19�9 6 1�8 286�1 6 12�0 0�46 22�4 6 3�7 5 of 7 93�9 1�34 18�8 6 1�8

Combined isochron n¼13: 292�0 6 8�0 1�13 22�0 6 2�6 Weighted mean n 5 2: 1�28 20�0 6 1�2Rhyolite of Arroyo Palacios (rap)LdM-12-23 6�20 22�4 6 2�0 293�4 6 12�9 0�47 23�6 6 3�6 7 of 7 100�0 0�49 22�4 6 2�0Andesite of Arroyo Mellicos (aam)LdM-194 0�34 32�3 6 14�0 303�3 6 8�6 0�87 13�0 6 9�9 9 of 9 100�0 0�91 28�8 6 12�5LdM-194 0�31 31�6 6 7�0 302�7 6 4�7 0�99 10�5 6 9�6 5 of 6 82�9 1�52 23�3 6 8�1

Combined isochron n¼14: 303�0 6 4�1 0�76 10�8 6 6�3 Weighted mean n 5 2: 1�08 24�5 6 6�1Rhyolite East of Presa Laguna del Maule (rep)LdM-12-32 7�50 24�7 6 3�4 291�9 6 28�0 0�82 26�3 6 5�8 8 of 8 100�0 0�73 25�0 6 3�0LdM-12-32 7�00 27�0 6 1�3 296�6 6 10�6 0�42 26�3 6 2�8 5 of 6 83�4 0�35 25�8 6 1�3

Combined isochron n¼13: 296�5 6 9�8 0�60 26�2 6 2�6 Weighted mean n 5 2: 0�56 25�7 6 1�2

Weighted mean plateau ages in bold are preferred; 2r uncertainties include the analytical and J uncertainties.*Monitored with the 28.201 Fish Canyon Sanidine (Kuiper et al., 2008); all other experiments were monitored with the 1.1864 MaAlder Creek Sanidine (Jicha et al., 2016);†high MSWD; not included in weighted mean

92 Journal of Petrology, 2017, Vol. 58, No. 1

surface exposure age of 9�5 6 0�1 ka for the shoreline

cut into igcb in the northern lake basin.

The middle to late Holocene saw rhyolite eruptions

from four centers in the southern and eastern lakebasin (Fig. 4). A significant explosive eruption from the

Cari Launa complex (Fierstein et al., 2012) was fol-

lowed by the older of two Cari Launa rhyolite flows

(rsl) at 3�3 6 1�2 ka, the Rhyolite of Colada Divisoria

(rcd) at 2�1 6 1�3 ka, and the small rcb flow east of the

Barrancas summit at 1�9 6 0�6 ka. Neither the upper-most western rcb flow nor the rhyolite of Colada

Las Nieblas (rln) produced resolvable 40Ar*, but on the

basis of their similar lack of pumice cover and

uneroded morphology, they are probably of compar-

able age to the rcd and eastern summit rcb lavas and

thus are among the most recent eruptions in the vol-canic field.

Outside the Holocene south–SE rhyolite focus, the

rhyodacite of the Northwest Coulee (rdcn) erupted from

a vent near the crest of the NW basin wall and extends

nearly down to the current lake level 350 m below. Thisprominent flow is dated at 3�5 6 2�3 ka and is mantled

by the andesitic cinder ring of Laguna Sin Puerto (asp),

which was subsequently intruded by the rhyodacite of

Sin Puerto (rdsp). These eruptions likewise emanated

from a vent on the crest of the NW basin wall. Two

small andesitic fissure eruptions, the andesite of Crater2657 (mcp) and the andesite of Arroyo Cabeceras de

Troncoso (mct), occurred 6 km west of the SW lake-

shore. The ages of these eruptions are not well con-

strained; however, mcp scoria blankets the post-glacial

rhyodacite south of Estero Piojo (rdep) mini-domes to

the north, but not the mct craters, indicating that bothare younger than rdep and, although they are at a

rcb

rln

rcd

rcl

rsl

rng

rdcn

rdcd

asp

mcp mct

rdsp

Latest Pleistoceneto Holocene14.5 - ≤ 1.9 ka

rle

ras

rap

rep

rdam

rdnordne

rdep

rdac

apjacn

mnp

asm

anc

aan

aam

rdnpapo

Early Post glacial25.7 - 19.0 ka

ram

5 km(a)

(b)

rcb-d

deglaciation

lakehigh stand

30 25 20 15 10 5 0Age [ka]

North

West

South

East

apo

rcdrclrsl

rdcd

rln

rcb

rng

rasrdac

raprdep

aam

apjasm

rdamram

acnancaan

rdcnasp

rdsp

rle

rep

rdoprdne

rdnpmnp

(c)

rcb-drcb-py

rcb summit

rcb-d rcb-pyrcb-py

Fig. 4. Post-glacial eruptive sequence of central LdM basin lavas. Fill colors are the same as in Fig. 2. (a) The distribution of EPGeruptions—those erupted prior to and including the rle flow that dammed the outlet gorge producing the highstand of the lake. (b)The distribution of latest Pleistocene to Holocene eruptions. (c) The relative eruptive sequence constrained by 40Ar/39Ar ages fromSinger et al. (2000) and this study; the timing of the drawdown of the lake highstand is constrained by a 9�5 6 0�1 ka 36Cl surface ex-posure age of the highstand shoreline cut into igcb tuff. Black outlined boxes are 40Ar/39Ar and 36Cl ages, with the width corres-ponding to the 2r uncertainty. Gray outlined boxes are inferred eruption age ranges based on field relationships; the widths are setrelative to the nominal ages of the constraining events.

Journal of Petrology, 2017, Vol. 58, No. 1 93

higher elevation than the high strandline, possibly post-

date the rle eruption as well (Hildreth et al., 2010).

WHOLE-ROCK GEOCHEMICAL RESULTS

Major and trace elementsLavas erupted during the last 150 kyr in central LdM

range from basalt to high-silica rhyolite. Primitive lavas,

rare throughout the SVZ, are absent from central LdM

as indicated by the modest Mg# (�53) and low K/Rb

ratios (369–242) of the basalt and mafic andesite sam-

ples. The major and trace element evolution of central

LdM generally mirrors that of the entire 1�5 Myr erup-

tive history of the larger volcanic field (Hildreth et al.,

2010) and the frontal arc Tatara–San Pedro complex (T–

SP; Dungan et al., 2001). Central LdM trace element

compositions form narrow arrays in elemental variation

plots compared with the range observed in the volcanic

field as a whole (Fig. 5). The Pleistocene LdM ignim-

brites igcb and igsp are notably enriched in rare earth

elements (REE), particularly middle REE (MREE), Y, and

Zr compared with the post-glacial silicic lavas.

Whereas many major and trace elements, such as

K2O, MgO, Th, U, Rb, and Pb, evolve monotonically

with increasing SiO2, several display prominent inflec-

tions in variation diagrams (Fig. 5 and supplemen-

tary figures). Between 52 and 60% SiO2, high field

strength elements (HFSE) (except Ti), large ion litho-

phile elements (LILE) (except Sr), light REE (LREE),

heavy REE (HREE), and Y increase with increasing SiO2.

Between 60 and 68% SiO2, Zr and LREE level off and

TiO2, MREE, Y, and P2O5 begin to decrease. Ba concen-

trations increase to 65% SiO2 but vary little in the more

evolved lavas. Between 68 and 70% SiO2, Zr concentra-

tions begin to decrease and the depletion of Sr with

increasing SiO2 becomes greater.

Sr and Pb isotope ratiosThe Sr and Pb isotope compositions of the central LdM

units, measured at the University of Wisconsin–Madison

ICP–TIMS Isotope Laboratory [Sr by thermal ionization

mass spectrometry (TIMS) and Pb by multicollector in-

ductively coupled plasma mass spectrometry (MC-ICP-

MS); see the Supplementary Data for details], display

T-SP

T-SP

T-SP

Greater LdMCentral LdM basin

Pleistocene ignimbrites

0

5

10

15

20

25

30

Th [p

pm]

0

200

400

600

800

1000

Sr

[ppm

]

0

200

400

600

800

1000

K/R

b

0

3

6

9

12

15

Rb/

Y

5

10

15

20

25

La/Y

b

SiO2 OiS]% .tw[ 2 [wt. %]45 55 65 7550 60 70 8045 55 65 7550 60 70 80

(a)

(c)

(d)

(e)

(f)

T-SP

0

100

200

300

400

500

Zr [p

pm]

(b)

Fig. 5. Variation of selected trace elements with SiO2 for lava and pumice erupted in the central LdM basin during approximatelythe last 150 kyr. Data for the 1�5 Myr history of the entire volcanic field, including the Pleistocene igcb and igsp ignimbrites (Hildrethet al., 2010; Birsic, 2015) and T-SP (Dungan et al., 2001) are plotted for comparison. The typical 2r uncertainties associated with thecentral LdM data are smaller than the symbols. The central LdM data show less dispersed ranges and trends relative to the largerLdM volcanic field and T-SP. The REE and Y ratios of the igcb and igsp ignimbrites notably diverge from those of the post-glacial si-licic lavas. Plots of major element variation are available in the Supplementary Data.

94 Journal of Petrology, 2017, Vol. 58, No. 1

limited variation. Ratios of 87Sr/86Sr range from 0�70407

to 0�70422, 206Pb/204Pb from 18�615 to 18�646, 207Pb/204Pb

from 15�606 to 15�622, and 208Pb/204Pb from 38�521 to

38�565 (Fig. 6; Table 3; Supplementary Data Figs A6 and

A7). Whereas the 207Pb/204Pb ratio does not vary coher-

ently with major or trace element composition, higher87Sr/86Sr, 206Pb/204Pb, and 208Pb/204Pb ratios are corre-

lated with increasing SiO2 (Supplementary Data Fig. A7).

The late Pleistocene to early post-glacial andesites apj

and aam have elevated 87Sr/86Sr ratios similar to those

of the silicic eruptions, but slightly less radiogenic206Pb/204Pb and 208Pb/204Pb ratios compared with the

more mafic units. In contrast, quenched mafic inclusions

in the northern rhyodacite domes rdno and rdne have87Sr/86Sr ratios similar to those of the basalts and mafic

andesite lavas, but higher 206Pb/204Pb and 208Pb/204Pb

ratios. The 87Sr/86Sr ratio of the modest-volume andesite

scoria eruption asp is similar to that of apj, aam, and the

silicic eruptions, but also has the most radiogenic206Pb/204Pb and 208Pb/204Pb ratios of this sample suite.

The range of the central LdM 87Sr/86Sr ratios is not-

ably narrow compared with regional volcanic centers

(Fig. 7). The LdM volcanic field as a whole has a wider

range of 87Sr/86Sr ratios of 0�70388–0�70435 and one

high outlying ratio, 0�70483, from the 430 ka rhyolite of

Cerro Negro (rcn; Hildreth et al., 2010). The Miocene

Risco Bayo–Huemul plutonic complex exposed beneath

the Tatara San Pedro volcanic complex contains volu-

metrically minor domains with 87Sr/86Sr ratios signifi-

cantly greater (>0�7050) than those of juvenile lavas in

the SVZ (Nelson et al., 1999). No lava with a comparably

radiogenic Sr isotope ratio has erupted in the central

LdM since the middle Pleistocene. The range of central

LdM is also similar to, but slightly narrower than, those

found at the nearby Pleistocene silicic centers including

the Puelche Volcanic Field (0�70386–0�70440; Hildreth

et al., 1999) and the Loma Seca Tuff and associated

lavas (0�70380–0�70433; Grunder, 1987).

Th isotopesThe Th isotopic compositions, measured by MC-ICP-MS

at the University of Wisconsin–Madison ICP–TIMS

Isotope Laboratory (see Supplementary Data for de-

tails), span a narrow range with modest disequilibrium

in both U- and Th-excess (Fig. 8; Table 4). The age-

corrected (230Th/232Th)0 activity ratios of the LdM lavas

range from 0�773 to 0�808, among the lowest yet meas-

ured in the SVZ. The rhyolites and rhyodacites display a

modest U-excess, up to 5%, and a narrow range of

(230Th/232Th)0 ratios, 0�793–0�808. The mafic lavas show

a greater diversity of Th isotopic compositions. The

(230Th/232Th)0 ratios of mafic lavas are nearly all lower

and have a 50% larger range, 0�773–0�800, than those of

the silicic eruptions. Most are in 2–5% Th-excess.

Quenched mafic inclusions hosted in units rdno and

rdne and the basaltic andesite lava mpl are in 3–4%

U-excess and have low (230Th/232Th)0 ratios spanning a

similar range to the Th-excess lavas (Fig. 8).

THERMOMETRY AND BAROMETRY

Two-oxide thermometryThe compositions of Fe–Ti oxides were determined by

electron microprobe at the University of Wisconsin–

Madison (see the Supplementary Data for details).

In the LdM rhyolites and rhyodacites, the ulvospinel

content of magnetite ranges from Ulv13 to Ulv25 and

hematite content of ilmenite from Hm25 to Hm31. The

average Fe–Ti oxide compositions of the apj andesite

flow are Ulv50 and Hm15. Oxides in both the rhyolites

and rhyodacites span the compositional range

observed in the silicic units; however, the highest

ulvospinel contents found in rhyolites are limited to the

products of the Cari Launa (rcl, rsl) center.

Fe–Ti oxide temperatures calculated using the cali-

bration of Ghiorso & Evans (2008) are 760–850�C for the

rhyolites, 796–854�C for the post-glacial rhyodacites,

and 760�C for the late Pleistocene rhyodacite rddm

(Fig. 9). Silicic units yield an fO2 1�19–1�32 log units

above the Ni–NiO buffer (NNO). Oxides from the

Younger Andesite of the Western Peninsula (apj) gave a

temperature of 1017�C and fO2 0�3 log units greater

than NNO. The range of temperatures produced for

multiple oxide pairs from each sample is �35�C for all

but three samples, commensurate with the 630�C un-

certainty typically ascribed to the two-oxide thermom-

eter (Ghiorso & Evans, 2008). The later erupted Cari

Launa rhyolites and unit rdcd produced temperature

15.45

15.55

15.65

15.7520

7 Pb/

204 P

b

Quaternary frontal arc

NHRL

Mz intrusions

Pz basement

Pε-N intrusions

sed

SA-NMORB

18.2 18.3 18.4 18.5 18.6 18.7 18.8206Pb/204Pb

Fig. 6. The 206Pb/204Pb and 207Pb/204Pb ratios of central LdMbasin lavas (red squares); data are given in Table 3. Alsoshown is the Northern Hemisphere Reference Line (NHRL;Hart, 1984), the composition of South Atlantic N-MORB (SA-NMORB; Douglass et al., 1999), Mesozoic (Mz) and Paleogeneto Neogene (Pe–N) intrusive rocks (Lucassen et al., 2004),Paleozoic (Pz) intrusive and metamorphic basement (Lucassenet al., 2004), SVZ sediments (Hildreth & Moorbath, 1988;Lucassen et al., 2010; Jacques et al., 2013), and QuaternarySVZ frontal arc lavas (Davidson et al., 1987; Gerlach et al.,1988; Hildreth & Moorbath, 1988; Hickey-Vargas et al., 1989;McMillan et al., 1989; Jacques et al., 2013; Holm et al., 2014).The LdM lavas yield a narrow range of 207Pb/204Pb isotopicratios compared with the frontal arc edifices and are distinctfrom those of the Paleozoic to Mesozoic basement, indicatingthat any assimilation was of younger, more primitive crust.

Journal of Petrology, 2017, Vol. 58, No. 1 95

spreads greater than 60�C. The younger Cari Launa lavaflow (rcl) and associated pumice cone produced a simi-

lar range of temperatures that in aggregate is 812–

884�C; the lowest temperature in this range is more

than 2r from the mean. Excluding this temperature nar-

rows the range to 845–884�C. Unit rdcd produced a

similarly wide range of 823–889�C; all calculated tem-

peratures are within two standard deviations of the

mean.

Amphibole thermobarometryAmphibole and plagioclase crystals in five rhyodacite

lavas (rdac, rdne, rdno, rdcd, and rdcn) were analyzed

by electron microprobe at the University of Wisconsin–

Madison. The plagioclase compositions are utilized toestimate the magma water content required for the

amphibole barometer calibration of Putirka (2016); a

more thorough interrogation of the plagioclase com-

positions will be the subject of a future contribution.

The anorthite content of plagioclase rims ranges from

An19 to An43. Using the hygrometer of Waters & Lange

(2015), the plagioclase rim and rhyodacite whole-rock

compositions yield a mean water content for each unit

ranging from 4�5 to 5�0 wt % at 850�C and 250 MPa; agrand mean of 4�8 wt % is adopted for the amphibole

calculations. The Waters & Lange (2015) hygrometer re-

quires an estimate of the crystallization pressure, but is

Table 3: Whole-rock Sr and Pb isotopic compositions

Sample Unit 87Sr/86Sr 2SE 206Pb/204Pb 2SE % 207Pb/204Pb 2SE % 208Pb/204Pb 2SE % n

LDM-12-25 aam 0�70419 0�00001 18�618 0�00005 15�613 0�00005 38�532 0�00004 2LDM-12-19 apj 0�70419 0�00001 18�623 0�00004 15�612 0�00004 38�540 0�00004 2ALDM-13-09 asp 0�70419 0�00001 18�648 0�00004 15�614 0�00004 38�570 0�00004 2LDM-12-34 bec 0�70412 0�00001 18�623 0�00006 15�611 0�00006 38�533 0�00004 1LDM-12-31 mnp 0�70409 0�00001 18�622 0�00005 15�613 0�00005 38�538 0�00004 1LDM-12-15 mpl 0�70408 0�00001 18�623 0�00008 15�621 0�00007 38�550 0�00004 1LDM-12-23 rap 0�70418 0�00001 18�638 0�00005 15�614 0�00006 38�557 0�00004 2LDM-13-13 rcb 0�70419 0�00001 18�636 0�00004 15�615 0�00004 38�557 0�00004 2ALDM-13-14 rcb 0�70420 0�00001 18�633 0�00005 15�612 0�00005 38�549 0�00004 2LDM-12-07 rcd 0�70422 0�00001 18�632 0�00006 15�612 0�00006 38�549 0�00004 2LDM-12-08 rcl 0�70419 0�00001 18�634 0�00005 15�613 0�00005 38�552 0�00004 1LDM-12-11 rdac 0�70420 0�00001 18�638 0�00005 15�614 0�00005 38�553 0�00004 1LDM-12-17 rdcd 0�70418 0�00001 18�640 0�00004 15�613 0�00005 38�557 0�00004 3LDM-12-17i rdcdi 0�70410 0�00001 18�621 0�00006 15�614 0�00007 38�534 0�00004 1LDM-12-27 rdcn 0�70413 0�00001 18�630 0�00004 15�612 0�00005 38�538 0�00004 2ALDM-13-10 rddm 0�70420 0�00001 18�633 0�00006 15�613 0�00005 38�547 0�00004 1LDM-12-03 rdne 0�70421 0�00001 18�636 0�00006 15�616 0�00005 38�551 0�00004 1ALDM-13-01 rdnei 0�70407 0�00001 18�630 0�00010 15�606 0�00011 38�535 0�00004 1LDM-12-33 rdno 0�70420 0�00001 18�635 0�00003 15�613 0�00003 38�551 0�00004 2LDM-12-33i rdnoi 0�70412 0�00001 18�637 0�00003 15�617 0�00003 38�561 0�00004 1LDM-12-16 rdnp 0�70411 0�00001 18�631 0�00006 15�612 0�00007 38�546 0�00004 1ALDM-13-08 rdsp 0�70414 0�00001 18�636 0�00004 15�614 0�00005 38�537 0�00004 1LDM-12-32 rep 0�70420 0�00001 18�637 0�00004 15�612 0�00005 38�550 0�00004 1LDM-12-04 rle 0�70419 0�00001 18�637 0�00004 15�613 0�00004 38�550 0�00004 3LDM-12-30b rle p 0�70420 0�00001 18�637 0�00004 15�613 0�00004 38�554 0�00004 1LDM-12-21 rln 0�70420 0�00001 18�636 0�00005 15�615 0�00005 38�560 0�00004 1ALDM-13-17 rsl 0�70419 0�00001 18�634 0�00004 15�613 0�00004 38�554 0�00004 2Standard analyses 2SD 2SD 2SD 2SDNIST SRM-987 0�71028 0�00001 30NBS-981 16�940 0�004 15�496 0�004 36�720 0�011 22NBS-982 36�754 0�026 17�161 0�006 36�749 0�015 11BCR-2 0�70505 0�00002 18�756 0�029 15�628 0�009 38�720 0�100 5 (Sr), 7(Pb)AGV-2 0�70402 0�00002 18�861 0�041 15�619 0�005 38�535 0�044 6 (Sr and Pb)

87Sr/86Sr ratios are reported as measured; age correction is inconsequential for these young samples.

0.7036

0.7038

0.7040

0.7042

0.7044

0.7046

87S

r/86

Sr

0 200 400 600 800Sr [ppm]

RhyoliteRhyodaciteBasalt - AndesiteMafic Inclusions

Greater LdMLoma Seca TuffTatara - San PedroPuelche Volcanic Field

Central LdMBasin SVZ 36º S

Fig. 7. Comparison of the central LdM basin 87Sr/86Sr as a func-tion of Sr content with those of nearby volcanic centers includ-ing T-SP (Davidson et al., 1987), the rear-arc Puelche volcanicfield (Hildreth et al., 1999), the Calabozos Caldera complex–Loma Seca Tuff (Grunder, 1987), and older eruptions through-out the LdM volcanic field (Hildreth et al., 2010). The regionaldata are plotted age corrected; the age correction is insignifi-cant for the central LdM lavas and these ratios are plotted asmeasured values. The central LdM lavas show a notably nar-row range compared with these nearby systems.

96 Journal of Petrology, 2017, Vol. 58, No. 1

relatively insensitive to this parameter. Over a range of

100–900 MPa, the calculated water content varies by

only 0�15 wt %. Thus, the inclusion of a pressure esti-

mate in the hygrometry calculation does not bias the

amphibole barometry.

The LdM amphiboles are pargasite to magnesio-

hornblende based on the classification scheme of

Hawthorne et al. (2012). Amphibole formulae based on 23

oxygen atoms, pressures, and temperatures are calcu-

lated using the method of Putirka (2016). The equilibrium

melt SiO2 is calculated to assess equilibrium with the host

magma; amphiboles that deviate by more than 4 wt %

from the host composition, the uncertainty associated

with the equilibrium SiO2 estimate, are not included in the

pressure calculations (Putirka, 2016). The resulting dataset

comprises 12–38 amphibole analyses for each unit and

yields average crystallization pressures of 190–250 MPa

with uncertainties of 30–50 MPa (Fig. 9). These pressures

are consistent with those calculated by the less precise,

but magma composition-independent, barometer calibra-

tions of Ridolfi et al. (2010) and Ridolfi & Renzulli (2012).

Pressure- and magma composition-independent amphi-

bole thermometry produces a range of 828–933�C, which

overlaps the two oxide temperatures from the rhyodacite

lavas, but also extends to higher temperatures.

DISCUSSION

The narrow compositional arrays of the central LdM

basin lavas suggest a common magmatic origin

(Hildreth et al., 2010). However, divergent correlations

among radiogenic isotope ratios and inflections in the

trajectory of trace element variation diagrams suggest

distinct differentiation pathways involving diverse crus-

tal assimilants and crystallizing assemblages. In the fol-

lowing sections we explore the following: (1) the

processes that have contributed to the geochemical

characteristics of the LdM lavas, particularly the sources

of U- and Th-excess; (2) whether these processes devi-

ate significantly from those inferred at frontal arc volca-

noes; (3) the processes promoting the more

homogeneous isotopic compositions of the rhyolites

compared with the mafic samples; (4) the temporal co-

herence of the thermo-chemical evolution of the LdM

magma system; (5) the implications for the structure

and state of the modern magma reservoir.

Crustal contributions to mafic magmasFrontal arc centers in the central and southern SVZ

commonly show relatively narrow ranges of radiogenic

isotope ratios, despite trace element evidence for sig-

nificant crustal interaction, owing to limited isotopic

contrast between the primary mafic magmas and the ju-

venile crust (e.g. Davidson et al., 1987; Dungan et al.,

2001). Uranium-series isotopes are a sensitive tracer of

magma evolution in arc systems as they provide infor-

mation about the nature of mantle and crustal compo-

nents, the processes leading to their mixing, and in

some cases the timescales of these processes (e.g.

Hickey-Vargas et al., 2002; Turner et al., 2003, 2010;

Jicha et al., 2007, 2009; Reubi et al., 2011; Ankney et al.,

2013). Mafic lavas in U-excess are common in arc set-

tings and are often attributed to the flux of slab fluids to

the mantle wedge (e.g. Turner et al., 2003). Less com-

mon Th-excess continental arc magmas are generally

thought to reflect a garnet signature inherited from the

0.7

0.8

0.9

1.0

0.7 0.9 1.1 1.30.750

0.775

0.800

0.825

0.850

0.70 0.75 0.80 0.85 0.90(238U/232Th)

(230 T

h/23

2 Th)

o

(238U/232Th)

Mafic lavaMafic inclusionRhyodaciteRhyoliteRhyolite Glass

Puyehue - Cordon Caulle

33º - 41º S historic mafic eruptions

LlaimaQuizapu

Laguna del Maule

Osorno and small Puyehue centers)b()a(

equi

line

equi

line

Fig. 8. Equiline plots of age-corrected Th isotope activity ratios for central LdM lavas and pumice erupted in the last 150 kyr. (a) TheLdM data compared with those measured at other SVZ volcanic systems (Hickey-Vargas et al., 2002; Sigmarsson et al., 2002; Jichaet al., 2007; Reubi et al., 2011; Ruprecht & Cooper, 2012). Central LdM lavas have among the lowest (230Th/232Th)0 activity ratios yetmeasured in the SVZ. (b) Detail equiline plot of the LdM data including the Th-excess mafic lavas and U-excess silicic products andrhyodacite-hosted mafic enclaves. The uncertainties in the (230Th/232Th)0 data include those of the ages used to correct the meas-ured ratios for decay since eruption. Dashed tie-lines connect mafic inclusions to their host rhyodacite.

Journal of Petrology, 2017, Vol. 58, No. 1 97

mantle or lower crust owing to its affinity for U over Th

(DU/DTh¼2�3–12�9; e.g. Rubatto & Hermann, 2007; Qian

& Hermann, 2013). In the SVZ, correlations among

fluid-mobile trace elements, 10Be/9Be, and U-excess in

frontal arc basalts have been interpreted as a slab fluid

control of the primary Th isotope signature (Hickey-

Vargas et al., 2002; Sigmarsson et al., 2002).

However, subsequent U-series studies of several SVZ

centers, including LdM, call into question the ubiquity of

this relationship. The enrichment of fluid-mobile elem-

ents in the SVZ is modest compared with volcanic arcs

globally (e.g. Ba/Th<300) and is only weakly correlated

with U-excess (Fig. 10; Supplementary Data Fig. A8).

Moreover, correlations between fluid-mobile element en-

richment and U-excess can result from crustal assimila-

tion rather than variations in the slab fluid signature

(Reubi et al., 2011). Whereas the addition of slab fluids to

the mantle wedge plays an important role in promoting

U-excess at some frontal arc centers, several mechan-

isms could contribute to their decoupling in the SVZ: (1)

long magma residence (>350 kyr) following the addition

of the fluid component to the mantle wedge allows the

U-excess signature to decay away (Hickey-Vargas et al.,

2002); (2) the addition of a Th-enriched sediment melt to

the mantle wedge would mitigate the fluid-derived U en-

richment (Jacques et al., 2013).

In the absence of significant fluid-derived U enrich-

ment, 3–6% partial melting of garnet lherzolite mantle

(e.g. Ottonello et al., 1984), with a composition

estimated as the average of that of Palme & O’Neill

(2003), will yield Th-excess similar to that measured in

the LdM lavas (Fig. 10). However, these low extents of

melting favor silica-undersaturated melts inconsistent

with the silica-saturated to -oversaturated lavas erupted

at LdM. Thus, the Th excess at LdM most probably re-

flects a greater extent of mantle melting and a contribu-

tion from garnet-bearing crust (GBC). The 207Pb/204Pb

ratios of the LdM lavas are distinct from those of the

more radiogenic Paleozoic to Mesozoic basement, indi-

cating that this crustal component must be relatively

primitive (Fig. 6; Luccassen et al., 2004).

Models of lower crust melting are calculated using ex-

perimental phase equilibria and partition coefficients

from the literature (see the Supplementary Data for

model parameters). The composition of the lower crust

is estimated using the global average of Rudnick & Gao

(2003); the narrow range of the (230Th/232Th)0 ratios of

the LdM lavas suggest that the initial U/Th ratio of the

crustal component is similar to that observed at LdM,

and thus the estimated crustal composition is adjusted

accordingly. Batch melting of GBC (e.g. Berlo et al., 2004;

Hora et al., 2009) and the formation of garnet during de-

hydration melting of initially garnet-free amphibolite

(Wolf & Wyllie, 1993; Ankney et al., 2013) have been pro-

posed to explain Th-excess in continental arc settings.

The latter, although appropriate for the large Th-excess

observed in Cascade lavas (Jicha et al., 2009; Ankney

et al., 2013; Wende et al., 2015), yields large Th-excess

Table 4: Whole-rock and glass 230Th–238U compositions

Sample Unit Age (ka) Th (ppm) U (ppm) (238U/232Th) 2SE (230Th/232Th) 2SE (230Th/232Th)0 2SE* (230Th/238U)0 n

LDM-12-25 aam 25�4 6 1�5 9�20 2�27 0�748 0�004 0�781 0�005 0�790 0�008 1�057 1LDM-12-19 apj 21�1 6 3�4 8�04 2�03 0�765 0�005 0�778 0�005 0�781 0�007 1�021 1ALDM-13-09 asp <3�5 7�94 2�07 0�792 0�005 0�790 0�005 0�790 0�005 0�998 2LDM-12-34 bec 61�8 6 3�6 3�09 0�77 0�754 0�005 0�765 0�005 0�773 0�012 1�025 1LDM-12-15 mnp <24 4�06 1�09 0�814 0�005 0�800 0�005 0�798 0�008 0�981 1LDM-12-31 mpl 54 6 21 6�76 1�65 0�742 0�004 0�770 0�005 0�788 0�020 1�062 1LDM-12-23 rap 22�4 6 2�0 22�05 5�95 0�819 0�005 0�800 0�005 0�795 0�007 0�970 1LDM-13-13 rcb <3 18�80 5�05 0�815 0�005 0�799 0�005 0�798 0�005 0�980 1ALDM-13-14 rcb 14�5–5�6 20�97 5�61 0�812 0�005 0�799 0�005 0�798 0�006 0�982 1LDM-12-07 rcd 2�2 6 1�2 20�59 5�49 0�810 0�005 0�798 0�005 0�798 0�005 0�986 2LDM-12-08 rcl <3�3 20�14 5�41 0�815 0�005 0�799 0�005 0�798 0�005 0�979 1LDM-12-11 rdac 20�0 6 1�2 19�99 5�42 0�822 0�005 0�798 0�005 0�793 0�007 0�964 1LDM-12-17 rdcd 8�00 6 0�84 19�60 5�39 0�834 0�005 0�798 0�005 0�797 0�005 0�956 2LDM-12-17i rdcd i 8�00 6 0�84 3�39 0�88 0�784 0�005 0�798 0�005 0�798 0�005 1�017 1LDM-12-27 rdcn 3�5 6 2�3 15�20 4�12 0�822 0�005 0�799 0�005 0�798 0�006 0�971 1ALDM-13-10 rddm 114 6 14 18�22 4�94 0�823 0�005 0�815 0�005 0�802 0�027 0�974 1LDM-12-03 rdne 25�7–19�0 16�36 4�36 0�809 0�005 0�799 0�005 0�797 0�007 0�985 1ALDM-13-01 rdne i 25�7–19�0 5�39 1�42 0�798 0�005 0�778 0�005 0�773 0�008 0�968 1LDM-12-33i rdno i 25�7–19�0 6�16 1�67 0�823 0�005 0�783 0�005 0�774 0�008 0�941 1LDM-12-16 rdnp <24 15�34 6�18 0�828 0�005 0�802 0�005 0�799 0�009 0�965 1ALDM-13-08 rdsp <3�5 16�97 4�61 0�824 0�005 0�804 0�005 0�804 0�005 0�976 1LDM-12-32 rep 25�7 6 1�2 23�42 6�32 0�819 0�005 0�802 0�005 0�797 0�008 0�973 1LDM-12-04 rle 19�0 6 0�7 23�50 6�32 0�816 0�005 0�803 0�005 0�800 0�007 0�980 2LDM-12-30b rle pum 19�0 6 0�7 23�25 6�20 0�810 0�005 0�808 0�005 0�808 0�007 0�998 1LDM-12-21 rln <3 19�08 5�14 0�817 0�005 0�800 0�005 0�800 0�005 0�980 1ALDM-13-17 rsl 3�3 6 1�2 20�89 5�56 0�808 0�005 0�797 0�005 0�796 0�005 0�985 2LDM-12-07 rcd glass 2�1 6 1�2 21�76 5�81 0�811 0�005 0�794 0�005 0�794 0�005 0�979 1LDM-12-04 rle glass 19�0 6 0�7 23�67 6�36 0�815 0�005 0�803 0�005 0�801 0�007 0�982 1LDM-12-21 rln glass <3 20�25 5�42 0�812 0�005 0�798 0�005 0�798 0�005 0�982 1ALDM-13-14 rcb glass 14�5–5�6 21�16 5�66 0�812 0�005 0�794 0�005 0�793 0�006 0�977 1LDM-12-08 rcl glass <3�5 20�78 5�60 0�817 0�005 0�801 0�005 0�801 0�005 0�980 1Standard analyses 2SD 2SDBCR-2 5�87 1�68 0�869 0�002 0�876 0�006 6AGV-2 6�01 1�86 0�938 0�002 0�946 0�005 5

*(230Th/232Th)0 uncertainty includes that of the eruption age.

98 Journal of Petrology, 2017, Vol. 58, No. 1

and HREE depletions inappropriate for the SVZ

(Supplementary Data Fig. A3). Mixing of 10% partial

melts of garnet-bearing crust and mantle reasonably re-

produces the range of Th-excess and REE compositions

found at LdM; however, the presence of U-excess mafic

lavas requires an additional explanation (Figs 10 and 11).

LdM mafic lavas in U-excess could be interpreted as

reflecting the slab fluid signature only partially over-

printed in the lower crust. However, these samples are

enriched in incompatible elements relative to the basalts

and mafic andesites in Th-excess, indicating that the

U-excess mafic lavas have experienced greater inter-

action with a crustal component. In contrast to garnet

production by amphibolite dehydration, the formation of

clinopyroxene during the melting of plagioclase- and

amphibole-bearing crust (garnet-free crust; GFC) can

produce U-excess (Fig. 11; Beard & Lofgren, 1991; Berlo

et al., 2004). Holocene intermediate lavas at T-SP were

produced, in part, by the melting of hornblende-bearing

mafic intrusions similar to T-SP xenoliths (Costa &

Singer, 2002). A 10% dehydration melt of this material

yields 6% U-excess, commensurate with the range

observed in the LdM lavas (Fig. 10).

Mixing among the mantle, GBC, and GFC end-

members, each produced by 10% partial melting, can

explain the Th isotope and trace element diversity of

the LdM mafic lavas. Variation of the Th isotope ratios

with the Zr/Th and La/Yb ratios forms offset arrays with

the largest Th-excess, found in units mpl and aam,

associated with higher La/Yb and lower Zr/Th ratios.

This offset is consistent with variable mixing, 5–30%, of

the GBC and mantle melts. Additional mixing with a

10% partial melt of GFC yields the range of U-series dis-

equilibrium observed in the LdM mafic lavas (Fig. 11).

Thus, despite a relatively limited range in isotopic com-

positions, the LdM lavas reflect extensive interactions

between mantle-derived melts and the continental

crust. Moreover, these processes vary little from those

inferred at frontal arc centers throughout the SVZ (e.g.

Davidson et al., 1987; Hildreth & Moorbath, 1988;

McMillan et al., 1989; Dungan et al., 2001; Costa &

Singer, 2002; Jicha et al., 2007). Thus, whereas the con-

centration of rhyolite at LdM is exceptional, the underly-

ing mafic magmatic processes are not.

Shallow vs deep origin of rhyoliteThe LdM silicic lavas are depleted in Ti, P, Sr, and Y,

have negative Eu anomalies, and have lower Dy/Yb

ratios relative to the andesites (Fig. 5; Supplementary

Data Fig. A5). These trends indicate a shift in the differ-

entiation regime from that of mafic magmas primarily

influenced by assimilation of crustal melts. Annen et al.

(2006) suggested that the majority of compositional di-

versity of volcanic rocks is imparted by lower crustal

processes. This model is inconsistent with the relatively

shallow crystallization pressures determined by amphi-

bole barometry at LdM. However, it is possible that the

amphibole is late crystallized and does not capture the

high-pressure differentiation history of the silicic mag-

mas. Differences in phase equilibria and the compos-

ition of potentially assimilated rocks between the deep

and shallow crust would impart predictable, divergent

geochemical trends during the generation of silicic

magma that are compared with the LdM compositions

to judge the plausibility of differentiation in the lower vs

upper crust.

We utilize Rhyolite-MELTS (Gualda et al., 2012)

to simulate fractional crystallization of an andesitic

LdM parental magma at a range of pressures

(150–1050 MPa), initial water contents (1–6 wt %), and

fO2 buffers (QFM to QFMþ2, where QFM is quartz–

-17

-16

-15

-14

-13

-12

-11

-10

650 700 750 800 850 900

QFM +2 Bishop

Tuff

Gla

ss M

t.

Loma Seca

Tuff

Log

fO2

T [ºC]

LdM

Rhyolite

LdM

Rhyodacite

NNO

VTTS

post-Oruanui

(a)0

100

200

300

400800 850 900 950

P [M

Pa]

T [ºC]

(b)

rdacrdcdrdcn

rdnerdno

Rhyodacite lavas

Holocene

EPG

Fig. 9. Results of mineral thermobarometry for central LdM eruptions. (a) T–fO2 plot for central LdM silicic eruptions. Fields showthe range of temperatures and oxygen fugacities for the Loma Seca Tuff (Grunder & Mahood, 1988), Bishop Tuff (Hildreth &Wilson, 2007), Glass Mountain rhyolites (Metz & Mahood, 1991), post-Oruanui rhyolites (Sutton et al., 2000) and the Valley of TenThousand Smokes rhyolites (VTTS; Hildreth, 1983). Reference T–fO2 curves for the nickel–nickel oxide buffer (NNO) and 2 log unitsabove the quartz–fayalite–magnetite buffer (QFMþ2) are shown, illustrating the highly consistent T–fO2 buffering of the LdM erup-tions. (b) Temperatures and pressures derived from amphibole compositions for LdM rhyodacite lavas. The pressure calculationassumes a magma with 4�8 wt % H2O based on plagioclase hygrometry (Waters & Lange, 2015). Each point is a single spot analysisand has uncertainties of 630�C and 6160 MPa (Putirka, 2016). The bars on the left of the plot are the average pressure and associ-ated uncertainty for each unit. The pressures of the Holocene lavas are nominally 50–60 MPa less than, but within uncertainty of,those of the EPG units.

Journal of Petrology, 2017, Vol. 58, No. 1 99

fayalite–magnetite) to evaluate the conditions in which

the LdM rhyolite magma formed. Each model is cooled

from the calculated liquidus to c. 700�C, depending on

model convergence at low melt fractions. The variation

of SiO2 and MgO of the LdM lavas is best reproduced

by shallow, oxidizing conditions and a moderate initial

water content. High pressures, water contents and

reducing conditions promote the early stabilization of

pyroxene at the expense of plagioclase and magnetite,

producing large depletions in MgO over a narrow range

of SiO2, inconsistent with the LdM compositions (Fig.

12). Moreover, Gaulda & Ghiorso (2013) argued that the

increasing stability of quartz with depth precludes the

generation of rhyolite by high-pressure fractional

crystallization.

MELTS is not well calibrated for hydrous intermedi-

ate to silicic compositions saturated in amphibole.

However, in this case, the SiO2/MgO ratio of LdM

amphibole (2�6–3�4) is between those of orthopyroxene

(2–3) and clinopyroxene (3�3–4�4) predicted by MELTS

such that the crystallization of either two pyroxenes or

amphibole would have a similar impact on the magma

SiO2/MgO ratio. Whereas some model misfit may result

from the prediction of pyroxene rather than amphibole

crystallization, the agreement between the MELTS mod-

eling and amphibole barometry indicates that the sup-

pression of plagioclase and magnetite crystallization is

the more important factor. Thus, MELTS simulations of

hydrous systems must be interpreted with caution, but

can yield useful first-order phase equilibrium con-

straints even when amphibole is present.

The physical plausibility of a viscous rhyolite magma

ascending >30 km through the crust is questionable

(e.g. Rubin, 1995). Even if it were possible, the similarity

of the rhyolite 87Sr/86Sr ratios to those of the mafic

and rhyodacite lavas (Fig. 7) weigh against a deep crust

origin. Following differentiation in the lower crust, Sr-

depleted rhyolite would then traverse the crustal col-

umn that includes highly radiogenic Paleozoic to

Mesozoic rocks (Lucassen et al., 2004; Supplementary

Data Fig. A6). The inevitable assimilation of even small

amounts (<5%) of this material would produce higher

and more variable 87Sr/86Sr ratios in the rhyolites than

observed. The more radiogenic 87Sr/86Sr ratios,

>0�7046, of the mid-Pleistocene rcn rhyolite erupted in

the eastern LdM basin and the most-evolved domains

of the Miocene plutonic complex beneath T-SP (Nelson

et al., 1999; Hildreth et al., 2010) potentially reflect as-

similation of this material; however, the modestly radio-

genic, homogeneous 87Sr/86Sr ratios of the post-glacial

rhyolites do not. Taken together, the isotope ratios of si-

licic LdM lavas, the incongruity between the predicted

phase equilibrium and the LdM major element compos-

itions, and shallow crystallization pressures recorded

by amphibole barometry rule out generation of the LdM

rhyolites in the lower crust.

Shallow hybridization and fractionalcrystallizationThe narrow range of Th isotope ratios and uniform

U-excess of the silicic lavas contrast with the more var-

ied mafic compositions (Fig. 8). Fractionating Th from U

in the upper crust to produce the silicic compositions

from a parental melt in Th-excess is not

0.74

0.76

0.78

0.80

0.82

0.84

0.65 0.70 0.75 0.80 0.85 0.90

equi

line

2 3 4 6 1015

51015

(238U/232Th)

(230 T

h/23

2 Th)

0

Garnet-freecrust melting

Garnetlherzolitemelting

LdMsiliciclavas

Garnet-bearingcrust melting

5 10 1520

30

(a)

(b)

50

100

150

200

250

300

350

0.6 0.8 1.0 1.2 1.4 1.6 1.8

Ba/

Th

(238U/230Th)0

LlaimaLdM Puyehue

Puyehue SECOsorno

Quizapu

Villarica

Villarica SEC

equi

line

Fig. 10. Sources of U-series disequilibrium in central LdMlavas. (a) Plot of SVZ U-series disequilibrium data for maficlavas compared with the Ba/Th ratio, an indicator of fluid en-richment. Volcanic centers, including small eruptive centers(SEC) associated with larger edifices, are listed in the legend ingeographical order from north (Quizapu) to south (Osorno)along the arc (Fig. 1). Some centers display evidence of cou-pling between fluid enrichment and U-excess; however, thiscorrelation may also result from crustal overprinting of theslab signature (Reubi et al., 2011). The range of Ba/Th in theTh-excess lavas is similar to that in U-excess and thus a strongcoupling between fluid enrichment and U-series disequilibriumis not evident in the SVZ. Data sources: Sun (2001), Hickey-Vargas et al. (2002), Jicha et al. (2007), Reubi et al. (2011) andRuprecht & Cooper (2012). (b) The U-series disequilibrium ex-pected during melting of the garnet-bearing mantle, garnet-bearing lower crust, and garnet-free crust (see theSupplementary Data for model parameters). The Th-excess ap-parent in the mafic LdM samples (red squares) can be pro-duced by melting with residual garnet in either the mantle orlower crust. U-excess in several mafic andesites and the siliciclavas reflects the overprinting of the garnet signature by partialmelting of garnet-free crust rather than U enrichment impartedby a subduction fluid (see text).

100 Journal of Petrology, 2017, Vol. 58, No. 1

straightforward. Crystallization of major phases will not

significantly increase the U/Th ratio, but accessory

phases such as apatite, titanite, allanite, and monazite

have greater leverage (Berlo et al., 2004). Of these, only

apatite is common at LdM. Rare, possibly xenocrystic,

titanite has been recovered by heavy liquid separation

from the large, early tephra eruption, but not from any

other LdM rhyolite; neither allanite nor monazite are

present. The crystallization of sufficient apatite or titan-

ite to produce U-excess from a Th-excess mafic magma

is not consistent with the P2O5 and MREE compositions

of the LdM lavas: fractionation of 0�3% titanite

(DTh¼18�7, DU¼7, DDy¼ 935, DYb¼ 393; Bachmann

et al., 2005) or 3�2% apatite (DTh¼ 2�82, DU¼ 1�9;

Condomines, 1997) is required to produce the observed

change in the U/Th ratio. The crystallization of these

phases in this quantity would decrease the Dy/Yb ratio

by a factor of seven and the P2O5 composition by 1�4 wt

%, respectively; both are approximately four times

greater than the variation observed in the central LdM

lava compositions. Thus, the crystallization of accessory

phases cannot account for the U-excess observed in the

silicic lavas.

The eruption of mafic magma in Th-excess and

evolved magma in U-excess has been observed at

several volcanoes in the Andes, Cascades, and Alaska

(Garrison et al., 2006; Jicha et al., 2007; Turner et al.,

2010; Ankney et al., 2013). This transition has variously

been ascribed to mixing with a U-excess endmember

derived from small degrees of partial melting with re-

sidual accessory phases, hydrothermal alteration of

assimilated wallrock, and variation in the contribution

of a subduction component through time. The garnet-

free crustal component evident at LdM offers an alter-

native explanation. The requirement of garnet in the

production of Th-excess limits this process to the lower-

most crust. Thus, only rapidly ascending magmas

would preserve a garnet-derived Th isotope signature.

Those that stall in the middle to upper crust and further

differentiate will have greater opportunity to interact

with GFC and acquire U-excess. Amphibole, common in

arc crust, is produced both by direct crystallization and

by reaction between clinopyroxene and ascending hy-

drous melt. Costa et al. (2002) advocated the latter

mechanism for the generation of amphibole beneath

T-SP and it also probably occurs at LdM. The subse-

quent melting of amphibole-bearing crust has been pro-

posed as an important source of melt and volatiles in

volcanic arcs more generally (e.g. Davidson et al., 2007,

2013); thus, the production of clinopyroxene during

0.85

0.90

0.95

1.00

1.05

1.10

1.15

0 10 20 30La/Yb

510

20 30 50

LdMsiliciclavas

(23

8U

/23

0T

h) 0

10% GFCmelt

10% mantle

melt 10

30

50

10

30

50

(b)

10 20 30 40 50Zr/Th

10% GBCmelt

10% GFCmelt

10% mantle

melt

LdMsiliciclavas

510

203050

10% GBCmelt

10

3050 10

3050

1

10

100

La Ce Nd Sm Eu Dy Yb La Ce Nd SmEu Dy Yb

sam

ple

/ch

on

dri

te

103050

% GFC melt

(a)

(c) (d)Mantle melt + 30% GBC melt Mantle melt + 5% GBC melt

equiline equiline

Fig. 11. A mixing model to explain the variation of U-series disequilibrium and the trace element composition of the mafic LdMlavas. The mixing endmembers are 10% melts of garnet lherzolite mantle, garnet-bearing crust (GBC), and garnet-free crust (GFC)(see the text and Supplementary Data). (a) and (b) show the variation of U-series disequilibrium with the Zr/Th and La/Yb ratios pro-duced by first mixing mantle and GBC melts. Subsequent mixing with a 10% melt of garnet-free crust produces the range of Th-and U-excess observed in the LdM mafic samples (red squares). The offset arrays of LdM data are consistent with varying mixingproportions of the mantle and GBC end-members. (c) and (d) show chondrite-normalized (Sun & McDonough, 1989) REE patternsproduced by 10%, 30%, and 50% mixing of the GFC endmember with a melt composed of 5% or 30% mixing of GBC with the man-tle melt, compared with those of the mafic LdM lavas (gray field).

Journal of Petrology, 2017, Vol. 58, No. 1 101

amphibole dehydration may be an under-appreciated

source of U-excess in intermediate to evolved continen-

tal arc magmas.

The evolution of the major and many trace element

compositions from the andesitic to silicic magmas is con-

sistent with the fractionation of the plagioclaseþamphiboleþbiotiteþFe–Ti oxideþ apatite 6 zircon as-

semblage observed in the rhyodacite and rhyolite lavas.

The saturation of zircon yields prominent inflections in the

evolution of the Zr concentration (Fig. 5); deviations from

the expected closed-system evolution would favor more

extensive open-system processes. Zr and Th are similarly

incompatible in major phases and thus, prior to zircon sat-

uration, fractional crystallization would produce compar-able enrichments in both elements. In central LdM, the

modest difference in the Zr concentrations of the rhyoda-

cite and andesite lavas is incongruent with the two-fold

difference in the Th concentrations.

We first consider a model of zircon-free fractional

crystallization of an andesite parental magma utilizing a

range of Zr partition coefficients, the anhydrous mineral

assemblage predicted by the best-fit MELTS model (Fig.

12), and a hydrous mineral assemblage in which amphi-bole crystallizes in place of pyroxene (Table 5). None of

these fractional crystallization pathways are able to pro-

duce the variation in Zr composition of the intermediate

LdM lavas (Fig. 13). The zircon saturation temperature

of most of the post-glacial rhyodacites is less than but

within uncertainty of the two-oxide temperature, indi-

cating they may have been zircon saturated—based on

the zircon saturation model of Watson & Harrison(1983); none are zircon saturated using the model of

Boehnke et al. (2013). Thus, the Zr contents of the rho-

dacite lavas could be produced by fractional crystalliza-

tion including a small but increasing modal per cent

zircon or could reflect open-system processes.

The two-oxide temperature of the andesite apj is

1017�C, several hundred degrees higher than the zircon

saturation temperature of this lava (Watson & Harrison,1983; Boehnke et al., 2013). The onset of zircon satur-

ation during cooling is evaluated by combining the

major element composition–crystallinity–temperature

relationship predicted by MELTS with a zircon-free frac-

tional crystallization model of the Zr content. The crys-

tallizing andesite magma saturates zircon after cooling

c. 260�C, resulting in 47% crystallization and reaching a

maximum Zr concentration of 305 ppm. This Zr contentis 15% greater than that of the central LdM rhyodacites,

indicating that they evolved under dominantly zircon-

undersaturated conditions (Fig. 13).

Moreover, the conclusions of this model are consist-

ent with the amphibole thermometry. The amphibole

temperatures and equilibrium melt SiO2 compositions

define an SiO2–temperature evolution that deviates

somewhat from the relationship predicted by MELTS.Nevertheless, the comparison of the zircon saturation

temperatures of the LdM lavas and the amphibole crys-

tallization temperatures indicates that zircon was not

saturated in the LdM magma until it reached c. 70%

0

1

2

3

4

55 60 65 70 75 80

MgO

[w

t. %

]

SiO2 [wt. %]

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

7508509501050

phas

e fr

actio

n

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

7508509501050

phas

e fr

actio

n

T [C]

210 MPa, 3% H2O, QFM+2

210 MPa, 3% H2O, QFM

210 MPa, 5% H2O, QFM+2600 MPa, 3% H2O, QFM+2

Model Conditions

melt

mtcpx

plag

opx

bt+qtz

opx

cpx

plag

mt

gt+qtzmelt

210 MPa, 3% H2O, QFM+2

600 MPa, 3% H2O, QFM+2

(a)

(b)

(c)

Fig. 12. Comparison of rhyolite-MELTS (Gualda et al., 2012) frac-tional crystallization simulations with the SiO2–MgO variation ofcentral LdM lavas to evaluate the effect of crystallization at arange of pressure, H2O, and fO2 conditions. (a) Four representa-tive MELTS simulations. These models are not exhaustive of therange of conditions considered but rather were selected to illus-trate the effect of changes to each parameter (see text). The best-fit model (thick black line) involves a low-pressure, moderatewater content, and oxidizing conditions, consistent with the min-eral thermobarometry (Fig. 9). Higher pressures, water content,and more reducing conditions produce significant depletions inMgO at intermediate SiO2 contents that strongly contrast withthe LdM data. (b) and (c) illustrate the contrasting crystallizing as-semblage produced at 210 and 600 MPa. Higher pressures, aswell as high H2O and more reducing conditions, stabilize pyrox-ene early at the expense of magnetite and plagioclase and pro-duce MgO-depleted magmas. Mineral abbreviations: mt,magnetite; cpx, clinopyroxene; opx, orthopyroxene; plag, plagio-clase; bt, biotite; gt, garnet; qtz, quartz.

102 Journal of Petrology, 2017, Vol. 58, No. 1

SiO2 (Fig. 13). Thus, whereas some LdM rhyodacites

may have saturated zircon prior to eruption, it was a

late-crystallizing phase, and the rhyodacite Zr contents

are primarily the result of open-system processes rather

than zircon fractionation. The rhyolite compositions are

consistent with an additional 20–35% crystallization of

an intermediate hybridized magma, assuming fraction-

ation of the mineral assemblage observed in the silicic

LdM lavas (Fig. 13; Table 5).

The array of rhyodacite Th–Zr compositions does not

readily implicate an LdM lava composition as the silicic

mixing endmember. It is relatively enriched in most in-

compatible trace elements, similar to the LdM rhyolites,

but not depleted in Zr (Fig. 13). The isotopic compos-

ition of the LdM lavas weighs against significant remo-

bilization of existing silicic crust, such as the plutons

beneath T-SP or the Pleistocene LdM ignimbrites.

Whereas the mafic LdM lavas span the entire range of

Sr and Pb isotopic compositions measured in the cen-

tral basin, reflecting the diversity imparted by lower

crustal interactions, and the rhyodacites nearly so, the

rhyolites exhibit more homogeneous isotope ratios des-

pite the wide spatial distribution of vents (Fig. 14).

Significant contributions from the modestly more radio-

genic and isotopically diverse upper crust would yield

higher, and probably more variable, 87Sr/86Sr ratios in

the LdM silicic lavas than observed.

The silicic end-member could be the product of rela-

tively closed-system differentiation (Fig. 13). However,

no magma with a composition consistent with this evo-

lution has erupted in the LdM since the 990 ka Bobadilla

ignimbrite. Moreover, a silicic mixing endmember

derived from closed-system differentiation would in-

herit the more varied isotopic ratios of the mafic lavas

and would not promote the increasingly homogeneous

isotopic ratios in the more evolved magmas.

Instead, the high temperature of the intruding mafic

magma could promote the resorption of zircon during

magma hybridization, thereby enriching the Zr content

relative to the rhyolitic magma. Zircon has not been

identified in thin sections of the LdM rhyodacites so the

presence of rare, partially resorbed zircon cannot be

confirmed. However, the abundance of mafic inclusions

in the rhyodacite lavas records the shallow mixing and

mingling of mafic and silicic magmas. Moreover,

plagioclase phenocrysts display a range of textures

ranging from relatively homogeneous to complexly

zoned, including resorption surfaces with overgrowths

and mottled and sieved cores reflecting varied and

complex thermal histories (Fig. 15). In contrast, plagio-

clase in the rhyolite lavas is only weakly zoned; this is

probably the result of efficient melt extraction from the

zones of magma hybridization, followed by a limited de-

gree of cooling and crystallization prior to eruption.

Taken together, the inferred trace element and isotopic

composition of the silicic endmember and the outcrop-

to mineral-scale textures of the rhyodacite lavas favor

self-assimilation—hybridization of intruding mafic to

intermediate magma with the post-glacial silicic reser-

voir including the resorption of zircon, rather than as-

similation of the upper crust (Figs 13 and 14).

Extensive rhyolitic magma was probably not avail-

able during the early growth of the LdM system. Thus,

the initial stages of magma reservoir development may

have involved remobilizing remnants of mid-

Pleistocene episodes of silicic magmatism and shallow

silicic intrusions or production of silicic magma by

closed-system processes (Fig. 13). Geochemical evi-

dence of such magma has not yet been identified and

they may have never produced an eruption. As the

post-glacial silicic system grew progressively larger, the

assimilation of young, hybridized rhyolite overtook any

contribution of the older material or highly fractionated

magma. Through self-assimilation, the increasing size

and homogeneity of the LdM magma system is a

coupled and self-reinforcing process.

Temporal evolution of the LdM magma systemThe temporal and spatial distribution of LdM eruptions

favors a laterally integrated shallow silicic magma sys-

tem (Hildreth et al., 2010) and offers clues to its struc-

ture and variations in the magmatic focus through time.

Eruptions in the southern and eastern LdM basin were

dominantly rhyolitic, excepting the apo andesite,

throughout post-glacial times, whereas volcanism in

the NW is characterized by a wider range of

Table 5: Partition coefficients and phase proportions used in fractional crystallization models

Phase Partition coefficients Fractionated phase proportions (%)

Zr Th Intermediate Intermediate Silicicanhydrous hydrous

plagioclase 0�001–0�01 0�0006 59 59 69orthopyroxene 0�026–0�14 0�04–0�22 5clinopyroxene 0�13–0�41 0�04–0�29 20amphibole 0�23–0�93 0�01–0�25 25 10magnetite 0�025–0�35 0�05–0�42 16 16 5biotite 0�05 0�01–0�5 15apatite 0�01 1�08–2�82 0�9 0�9 0�9zircon* 6–18 0�07 0�07 0�08

*The Zr content of zircon is assumed to be stoichiometric.Data sources: Luhr & Carmichael (1980); Bacon & Druitt (1988); Dunn & Sen (1994); Ewart & Griffin (1994); Sisson (1994); Brenanet al. (1995); Bindeman et al. (1998); Villemant (1988); Sano et al. (2002); Blundy & Wood (2003); Bachmann et al. (2005).

Journal of Petrology, 2017, Vol. 58, No. 1 103

compositions (Fig. 4). The common andesite eruptions

in the west and NW during EPG time suggest that the

upper crustal magma system was thinner there relative

to the south. Similarly, the numerous rhyodacite erup-

tions in the NW carry abundant, large mafic inclusions,

whereas they are rare and small in the lone post-glacial

rhyodacite eruption in the south, rdac. Taken together,

magmatism in the central LdM has been focused in the

southern basin since before the last glacial maximum,

resulting in a well-developed mush and a preponder-

ance of rhyolitic eruptions.

In the NW, the most recently erupted rhyolite is the

19�0 ka rle flow; subsequent eruptions of any compos-

ition are scarce until the late Holocene (Fig. 4). The most

recent northern eruptions, units rdcn and rdsp,

occurred after a local hiatus of as much as 15 kyr. These

geographical differences in the eruption frequency and

physical and compositional characteristics of the erup-

tive products indicate that the crystal mush, well de-

veloped to the south, either thins or is discontinuous

beneath the NW portion of the lake basin. Renewed vol-

canism in the NW during the Holocene produced units

rdcn and rdsp, suggesting a recent expansion of the

magmatic footprint at LdM and potentially lateral

growth of the active silicic magma system. The amphi-

bole crystallization pressures of the Holocene rdcn and

rdcd lavas are nominally 50–60 MPa less than those of

the EPG rhyodacites, although this difference is within

the uncertainty of the barometer calibration (Fig. 9).

These results suggest, but cannot prove, that the lateral

growth of the LdM system may have been accompa-

nied by the shallowing of active magmatism.Whereas spatial distinctions in the distribution of

mafic and silicic eruptions are readily apparent, com-

positional differences among the post-glacial rhyolites

are subtle, but coherent in time rather than with vent lo-

cation. Holocene rhyolites are enriched in Y and MREE

compared with the EPG rhyolites (Fig. 16). Two-oxide

temperatures vary similarly. The eruption temperature

ranges of the EPG (737–801�C) and Holocene (781–

850�C) rhyolites overlap; however, the Holocene tem-

peratures are consistently at the higher end of the total

range, suggesting an increase in magma reservoir tem-

perature with time (Fig. 16). That the earlier erupted

rhyolite is cooler and more evolved precludes linking

the EPG and Holocene compositions by a progressive

differentiation or mixing process.

The variation of most trace elements in the rhyolites

defines a single liquid line of decent; in contrast, the

Holocene enrichments in Y and MREE define opposing

trends with SiO2 compared with the earlier erupted

rhyolites. These trace elements show flat or decreasing

trends with SiO2 in the EPG rhyolites but increasing

trends in the Holocene rhyolites (Fig. 16) and thus are

0

100

200

300

400Zr

[pp

m] zrc-

free

FC

FC, 0.07% zrc

mantle melt

GBC melt

Miocene plutons

2060

80

zrcresorption10

30

20

0 10 20 30Th [ppm]

LdMig.

600

700

800

900

1000

1100

1200

58 63 68 73

T [

°C]

SiO2 [wt. %]

MELTS model

B(2013)

W&H(1983)

zrc under-saturated zrc saturated

amphibolethermometryLdM lava zrcsaturation T

zircon undersaturatedzircon saturatedsaturation not determined

GFC melt

FC, 0.08% zrc

(a) (b)

Fig. 13. Comparison of fractional crystallization and magma mixing contributions to the upper crustal genesis of the LdM siliciclavas and evaluation of zircon saturation during magma differentiation. (a) The variation of Th and Zr concentrations in central LdMlavas. The trace element compositions of the mafic lavas are dominated by mixing between partial melts of the mantle, GBC, andGFC (Fig. 11). Zircon-free fractional crystallization (zrc-free FC) of a parental andesitic magma produces an enrichment in Zr greaterthan observed in the LdM rhyodacites; the dashed lines show the range of models produced using the low and high partition coeffi-cients reported in the literature and hydrous vs anhydrous fractionating assemblages (Table 5). The temperature evolution of thebest-fit MELTS simulation in Fig. 12 predicts that the magma system will saturate in zircon at 305 ppm Zr, indicating that the flat Zrevolution of the rhyodacites is not due to the fractionation of a small modal fraction of zircon. Instead, the rhyodacite compositionsare most consistent with mixing between intermediate and silicic LdM magmas (green line), the latter enriched in Zr by the resorp-tion of zircon (green diamond). The rhyolite compositions are consistent with an additional 20–35% crystallization of a hybridizedintermediate magma. (b) The SiO2–temperature evolution of the MELTS model calculated from the amphibole compositions com-pared with the SiO2–zircon saturation temperature relationship of the LdM lavas—calculated using the calibration of Watson &Harrison (1983)—and predicted by the MELTS fractional crystallization model using the zircon saturation calibration of bothWatson & Harrison (1983) [W&H(1983)] and Boehnke et al. (2013) [B(2013)]. Both the model and mineral data predict that themagma saturates in zircon at c. 70% SiO2, consistent with the inflection in the whole-rock SiO2–Zr variation (Fig. 5).

104 Journal of Petrology, 2017, Vol. 58, No. 1

also inconsistent with progressive eruption from a

zoned magma reservoir. Instead, the compositional dif-

ferences reflect discrete magma bodies that, remark-

ably, produced eruptions over a comparably wide area,

similar to those inferred for the Mamaku and Ohakuri

ignimbrites and rhyolites following the 25�4 ka caldera-

forming Oruanui eruption in the Taupo Volcanic Zone

(Sutton et al., 2000; Vandergoes et al., 2013; Begue

et al., 2014; Barker et al., 2014, 2015). Rhyolites of dis-

tinct composition were erupted c. 20 kyr apart, from

vents separated by only 2 km (e.g. rap and rln). In con-

trast, coeval rhyolites nearly identical in composition

erupted more than 10 km apart during both the EPG

(e.g. rap and rle) and Holocene (e.g. rln and rcd). Rather

than being the products of small, short-lived, isolated

magmatic systems, the temporally coherent, spatially

extensive rhyolitic eruptions imply the extraction of

chemically distinct magma from a long-lived, compos-

itionally evolving, upper crustal source region.Long-term variations in rhyolite composition, tem-

perature, and mineralogy can be driven by variations in

the lower crust temperature in response to the basalt

flux from the mantle and changes in the supply of slab

fluids (Deering et al., 2008, 2010). However, at LdM, the

relatively short duration of rhyolitic volcanism and

nearly invariant fO2 buffering indicate that the subtle

differences in trace element composition and tempera-

ture are more probably related to the upper crust proc-

esses of rhyolite differentiation and extraction. Hildreth

(2004) proposed that trace element variations among

broadly homogeneous rhyolites can reflect the variable

stability of accessory phases. Similarly, Barker et al.

(2014, 2015) attributed the diversity of post-Oruani sili-

cic magma compositions at Taupo volcano, in part, to

the resorption of amphibole, clinopyroxene, and zircon.

At LdM, extraction of a volatile-rich rhyolite would

leave behind a relatively water-poor cumulate mush

(Wolff et al., 2015). The repeated intrusion of mafic

magma would promote the resorption of amphibole or

late crystallized, cryptic titanite, resulting in MREE- and

Y-enriched magma (e.g. Deering et al., 2011). Thus, the

eruption of compositionally distinct rhyolites over time

may reflect long-term changes in the phase equilibrium

and temperature of the plutonic mush induced by the

aggregate effect of at least 26 kyr of mafic intrusions

into the upper crust. Alternatively, the composition of

each rhyolite could reflect the ephemeral effect of each

most recent magma recharge episode. In this case,

compositional differences between one set of coeval

rhyolites and the next could be a record of the response

to and size of the mafic incursions, but not necessarily

of the long-term dynamics and thermo-chemical state

of the magma reservoir. Protracted extraction or resi-

dence in the crust would tend to average out subtle

compositional differences; thus, the preservation of

compositional distinctions among the LdM rhyolites

favors rapid melt segregation and only brief storage.

Whereas there is scarce evidence for physical inter-

action between the erupted rhyolite and intruding mafic

to intermediate magma, the extraction of crystal-poor

rhyolite could nevertheless be catalyzed by magma re-

charge in the lower reaches of the magma reservoir.

Increasing temperatures would raise the porosity of the

crystal mush and, along with the exsolution of volatiles

from the mafic magma, increase the buoyancy of the

rhyolitic liquid (e.g. Barker et al., 2016).

Structure and dynamics of the magma reservoirThe combination of the basin-wide progression generally

from andesite to rhyolite, the importance of magma hy-

bridization in rhyolite petrogenesis, and the temporal co-

herence of variable rhyolite compositions suggests the

physical configuration of the LdM magma system

0.7040

0.7041

0.7042

0.7043

18.61 18.62 18.63 18.64

87S

r/86

Sr

206Pb/204Pb

0.7040

0.7041

0.7042

0.7043

0.7044

0 100 200 300 400 500 600

87S

r/86

Sr

Sr [ppm]

Mioceneplutons

LdM ignimbrites

(a)

(b)

20 40 60

20 40 60

80

204060

80

20406080

Fig. 14. The effect of magma hybridization on Sr and Pb iso-tope ratios; symbols are the same as in Fig. 7. Curves illustratemixing between high and low 87Sr/86Sr mafic magma and anaverage rhyolite composition. The isotopic diversity of themafic magmas, inherited from lower crust interactions, islargely preserved by the rhyodacites. The comparatively nar-row range of the rhyolite isotope ratios is produced by hybrid-ization and homogenization within an integrated magmasystem. The fields in (a) show the range of 87Sr/86Sr ratios forigcb, igsp, and the Risco Bayo–Huemul plutons, plotted asmeasured (Nelson et al., 1999; Hildreth et al., 2010).Assimilation of this material would yield higher and more var-ied 87Sr/86Sr ratios in the post-glacial rhyolites than observed,favoring a model of self-assimilation within the post-glacialmagma reservoir.

Journal of Petrology, 2017, Vol. 58, No. 1 105

illustrated in Fig. 17: it comprises an integrated magma

source zone, sustained during at least the last 26 kyr. This

region is spatially extensive and intercepts the ascent of

diverse mafic magmas that promote magma hybridiza-

tion, resorption of accessory phases, and the segregation

of crystal-poor melt batches. In the south, this magma

mingling and mixing is limited to the base of the crystal

mush, resulting in little physical interaction between the

recharge magma and the erupted rhyolite batches.

Thinning of the system to the north allows for penetration

of mafic magma to shallower levels, thereby promoting

the eruption of mingled and hybridized magma. Crystal-

poor rhyolite is periodically extracted and stored only

briefly prior to eruption. The composition of these erupted

magma batches reflects the longer-term homogenization

in the upper crust by magma hybridization, temporal vari-

ation in the thermochemical state of the magma reservoir,and possibly compositional characteristics imparted dur-

ing melt extraction.

The repeated generation of compositionally and iso-

topically distinct rhyolite magma batches is an increas-

ingly recognized feature of long-lived silicic magma

systems. These systems have produced a range of

eruptive behavior including the sequential eruption of

diverse rhyolites, the coeval eruption of spatially and

compositionally distinct magmas, and the pre-eruption

amalgamation of several magma bodies, yielding volu-

minous ignimbrites characterized by isotopically and

compositionally diverse phenocrysts (Bindeman et al.,

2008; Deering et al., 2008; Charlier & Wilson, 2010;

Klemetti et al., 2011; Storm et al., 2011, 2014; Barker

et al., 2014, 2015; Begue et al., 2014; Wotzlaw et al.,

2015; Evans et al., 2016; Myers et al., 2016; Rubin et al.,

2016). The compositional continuity of the distributed

rhyolite eruptions through time observed at LdM and

Taupo, post-Oruanui, (Sutton et al., 2000; Barker et al.,

2014) demonstrates the remarkably lateral continuity

possible in silicic systems and the short timescales over

which compositional distinctions can be produced.Neither LdM nor Taupo have erupted high-SiO2 rhyo-

lite with extreme depletions in Sr and Ba, large negative

Eu anomalies, and low temperatures that characterize,

for example, the Glass Mountain rhyolites erupted at

Long Valley (Metz & Mahood, 1991; Hildreth & Wilson,

(b)(a) (c)

(e)(d) (f)

(h)(g) (i) (j)

rdcdrdne rdcn

Fig. 15. Textural evidence of open-system processes in LdM rhyodacites. (a–c) Outcrop photographs of rdne, rdcd, and rdcn show-ing representative examples of chilled mafic inclusions, highlighted by the arrows. (d–j) BSE images of representative rhyodaciteplagioclase textures including sieved and mottled cores, resorption surfaces, and oscillatory zoning—all indicative of varied andcomplex thermal histories. In contrast, rhyolite plagioclase crystals, not shown, are dominantly homogeneous. The scale bar ineach image represents 100 lm.

106 Journal of Petrology, 2017, Vol. 58, No. 1

2007). Such compositions are indicative of a eutectic

mineral assemblage saturated in two feldspars and

quartz. The crystallinity of eutectic systems is more sen-

sitive to temperature than those saturated in plagioclase

and quartz, resulting in more variable trace element

compositions in response to large changes in crystallin-

ity during both cooling and remelting of crystal mushes

(e.g. Mahood, 1990; Sutton et al., 2000; Bindeman &

Simakin, 2014). Recharge by hotter magma is implicated

in the generation of eruptible rhyolite reservoirs in both

sanidine-bearing and sanidine-free systems but possibly

by different physical mechanisms. The melting of fertile,

sanidine-bearing mush or hydrothermally altered silicic

precursors contributed to the caldera-forming magma

reservoirs in Long Valley (Chamberlain et al., 2014a;

Evans et al., 2016), San Juan (Bachmann et al., 2005;

Wotzlaw et al., 2013), and Yellowstone (Bindeman et al.,

2008; Bindeman & Simakin, 2014; Wotzlaw et al., 2015)

systems. In sanidine-free systems, the thermal input of

magma recharge catalyzes the extraction of crystal-poor

rhyolite and resorption of some minerals, but not remelt-

ing on the scale observed in eutectic systems (Barker

et al., 2014, 2015, 2016; Singer et al., 2016). A number of

factors such as the local and regional tectonics, the crus-

tal lithology, the depth of the magma system, and its

volatile contents contribute to the dynamics of rhyolite

generation. However, the minerology-dependent re-

sponse of the shallow reservoir to magma recharge may

also have significant implications for the varied mechan-

isms and timescales of the generation of eruptible rhyo-

litic magma bodies and the growth of their source

regions; this is worthy of further investigation.

The similarity between the rhyolite volcanism at

LdM and following the Oruanui eruption in Taupo is

striking and suggests similar underlying dynamics.

Owing to active rifting and a high flux of mantle-

derived melt, the rhyolite productivity of the TVZ is re-

markable globally (Wilson et al., 2009). Tectonic exten-

sion is often suggested as a catalyst for rhyolite

volcanism (e.g. Hughes & Mahood, 2011) and the con-

centration of silicic volcanism behind the frontal arc in

the SVZ, and at LdM in particular, may be related to

back-arc extension (Folguera et al., 2012). However,

widespread extensional structures are not observed at

LdM and thus the effect of local to regional extension

cannot be confirmed.

2

4

6

8

Sm

[ppm

]

14

16

18

20

22

24

65 70 75

Y [p

pm]

SiO2 [wt. %]

LdM ignimbrites

EPG trend

Latest Pleistocene

& Holocene trend

LdM ignimbrites23 - 46 ppm Y

EPGL. Pleistocene& Holocene

Silicic Lavas

700

725

750

775

800

825

850

875

900

0102030

Tw

o-o

xid

e T

[ºC

]

Eruption age [ka]

rhyoliterhyodacite

(a)

(b)

5 km

(c)

(d)

EPGLatest Pleistocene

& Holocene

Fig. 16. (a, b) Comparison of Sm and Y concentrations for EPG and Holocene silicic eruptions and central LdM ignimbrites igcb andigsp (Hildreth et al., 2010; Birsic, 2015) indicating that two compositionally distinct post-glacial rhyolite bodies were erupted in centralLdM. The enrichment of the Holocene rhyolites in MREE and Y is consistent with the resorption of cryptic titanite and/or amphibole.The destabilization of these phases could be in response to either repeated mafic intrusion or the ephemeral effect of each most recentrecharge event. Error bars corresponding to the 2r analytical uncertainty are smaller than the symbol size. (c) Temporal variation intwo-oxide temperatures. The Holocene rhyolites are subtly hotter than those erupted in the EPG, whereas rhyodacite temperaturesvary little during post-glacial times. Eruption ages were determined by 40Ar/39Ar or 36Cl, or were estimated from stratigraphic relation-ships (Table 1; Fig. 4). Pink and orange symbols are rhyolites and rhyodacites, respectively. Vertical error bars are the range of tem-peratures produced by touching pairs or the minimum and maximum determined by combinations of isolated titanomagnetite andilmenite crystals, with the tick indicating the average; the uncertainty in the thermometer calibration is 630�C (Ghiorso & Evans,2008). (d) Map showing the distribution of silicic lavas erupted during the EPG and latest Pleistocene to Holocene.

Journal of Petrology, 2017, Vol. 58, No. 1 107

IMPLICATIONS FOR THE CONTINUING UNREST

The continuing inflation at LdM is interpreted as a re-

sponse to magma emplacement in the shallow crust

(Feigl et al., 2014; Le Mevel et al., 2016; Miller et al.,

2016). The uplift of the southern lake highstand paleo-

shoreline of >60 m implies repeated similar deform-

ation episodes throughout the Holocene, consistent

with the emplacement of a significant volume of

magma into the shallow crust (Singer et al., 2015).

Zircon crystallization ages suggest that the 600 km3,

rhyolitic Bishop Tuff magma body accumulated at a

rate of 7�5 km3 ka–1 for 80 kyr prior to its eruption

(Chamberlain et al., 2014b). At LdM, if the rate of vol-

ume addition modeled to explain the modern uplift is

taken as the growth rate of the silicic magma system

and the average length of a deformation episode is

taken to be 10 years, the integrated volume increase

would be 0�5 km3 per inflation episode (Le Mevel et al.,

2016). The physical significance of magma emplace-

ment rates derived from zircon crystallization intervals

is a matter of debate. At a minimum, they probably rep-

resent an average of many punctuated, high-flux peri-

ods rather than protracted steady-state mass addition.

To achieve a long-term average flux at LdM of similar

magnitude would require 15 magmatic episodes, simi-

lar to the one occurring today, every thousand years.

Frequent shallow intrusion of magma at LdM, with an

average recurrence interval of decades to centuries, is

consistent with the repeated eruption of rhyolite since

the last glacial maximum and the dramatic deformation

of the highstand paleoshoreline during the last 9�5 kyr.

Wilson & Charlier (2009) suggested that long zircon

crystallization histories record inheritance of antecrysts

during the growth of magmatic mush and not the accu-

mulation of eruptible magma. Rates of melt extraction

leading up to rhyolite eruptions can reach several km3

a–1 as inferred for the Oruanui and a post-caldera erup-

tion at Taupo (Allan et al., 2013; Barker et al., 2016). The

rate of volume addition inferred from geodesy at LdM is

not of this remarkable magnitude, but is similar to the

more modest rates of rhyolite extraction of other

Holocene Taupo rhyolites (Barker et al., 2016). Thus,

whereas the rate of uplift today at LdM is globally re-

markable (Le Mevel et al., 2015), the potential rates of

(a) Early Post Glacial

meltingminglingmixing

crystal-poor rhyoliteholding zone

intermediate forerunnersto rhyolite flare-up

Laguna del Maule

NW SE

Mafic magma from lower crustmantle melt + crust melt

ongoing uplift >20 cm/yr

xtl-rich

xtl-poor

Laguna del Maule

growth of magma mushaccommodated by surface deformation

rejuvenation ofnorthern source zone

crystal-poor meltextraction

(b) Holocene

crystal-poor meltextraction

2 km

2 km

shallow, eruptiblemingled melt

Laguna del Maule (c) Modern configuration

shallow seismicity

crystal-poor meltextraction?

continued magma intrusionpromotes surface deformation

? ?

Mafic magma from lower crustmantle melt + crust melt

Mafic magma from lower crustmantle melt + crust melt

Fig. 17. Conceptual cross-sections of the structure and tem-poral evolution of the LdM magma system. The three panelsdo not represent specific moments in time, but rather summar-ize important facets of the magma system during each eruptiveepisode. The shallow LdM magma system comprises an exten-sive crystal-rich magma source zone that extends beneathmost, if not all, of the lake basin. Throughout post-glacial time,mafic magmas ascending from deeper in the crust are inter-cepted, providing a source of mass, heat, and volatiles prevent-ing the system from cooling to the granite eutectic.Hybridization and crystallization yield isotopically homoge-neous rhyolite (Fig. 14) that is segregated into eruptible, crys-tal-poor bodies that fed the post-glacial rhyolite eruptions. (a)During the EPG, the abundant eruption of mafic and mafic in-clusion-bearing rhyodacite lavas in the NW suggests that themushy rhyolite source zone thins compared with the south-eastern basin where similar products are not observed. Thehighly consistent trace element compositions of rhyoliteserupted in the north and south suggest, but do not require, thatthe erupted reservoir was integrated throughout the LdMbasin. (b) During the latest Pleistocene to Holocene, less com-mon mafic eruptions suggest that growth of the northernmagma system increased its capability to intercept ascending

magma. The lack of Holocene rhyolite eruptions in northernLdM suggests that the segregation of melt was limited to thesouthern basin. (c) The focus of magma intrusion may havemigrated during the Holocene as the modern inflation center isNW of the most productive Holocene rhyolite center,Barrancas, and the areas of maximum shoreline deformation(Singer et al., 2015). The continuing crustal deformation andshallow seismicity concentrated near the rcb and rln rhyolites(Feigl et al., 2014; Le Mevel et al., 2015; Singer et al., 2015) re-flects magma intrusion and the movement of melt and fluid,consistent with the magmatic processes inferred throughoutpost-glacial times. Consequently, the future segregation oferuptible, crystal-poor rhyolite appears likely. However, thatsuch a magma body currently exists, and if so, its extent andvolume, is the subject of a continuing geophysics investigation(Singer et al., 2014).

108 Journal of Petrology, 2017, Vol. 58, No. 1

long-term reservoir growth and continuing melt extrac-

tion are comparable with those inferred beneath pro-

ductive rhyolitic systems that produced calderas

elsewhere.

Mixing between existing reservoirs and intruding

magma has been found to precede silicic eruptions by

as little as weeks to years (e.g. Druitt et al., 2012; Till

et al., 2015; Singer et al., 2016). The duration of extraor-

dinary inflation at LdM has already exceeded these

shortest temporal estimates. Volcanic inflation episodes

usually conclude without eruption, and the most recent

geodetic observations suggest that the rate of uplift at

LdM is beginning to decrease (Le Mevel et al., 2015).

There is no evidence to suggest that the current unrest

is anything but a continuation of the longer-term proc-

esses operating at LdM that produced significant de-

formation of the highstand paleoshoreline during the

Holocene and frequent eruptions since 26 ka. Future

rhyolite eruptions are likely; however, that an eruption

is imminent is not at all clear. Whether these future

events will continue to be of modest volume or if the

system is building towards a larger eruption remains an

open question and is the subject of continuing geophys-

ical surveys and numerical modeling investigations

(Singer et al., 2014).

SUMMARY AND CONCLUSIONS

The post-glacial concentration of rhyolite volcanism at

LdM is fundamentally the product of magmatism

throughout the thickness of the crust little different

from that inferred at SVZ frontal arc volcanoes.

Mantle-derived basalt mixes with two lower crustal

components prior to shallow emplacement. Whereas

the eruptive expression of this magmatism along the

frontal arc is dominantly mafic to intermediate in the

SVZ, rear-arc systems such as LdM yield more silicic

compositions—possibly catalyzed by regional back-arc

extension. Upon ascending to the upper crust this

mafic magma mingles and mixes with pre-existing

silicic material followed by fractional crystallization

yielding the rhyolitic compositions. The combination

of self-assimilation and the plagioclaseþquartz-satu-

rated silicic magma buffers the modestly evolved,

broadly homogeneous compositions.

The rhyolite eruptions are clustered in two pulses,

both of which produced activity throughout the LdM

basin. Distinct trace element compositions and two-

oxide temperatures indicate that at least two crystal-

poor magma bodies were extracted during post-glacial

times. The temporal correlation of increasing tempera-

ture and enrichment in trace elements compatible in

titanite and amphibole suggests that melt extraction is

catalyzed by mafic recharge that promotes resorption

of accessory and hydrous phases within a crystal-rich

reservoir.

The petrological model supported by the geochem-

ical data is consistent with shallow magmatism that

probably produced episodes similar to the continuing

unrest throughout post-glacial time. Extrapolating the

volume change estimated for the modern inflation

yields a rate of mass addition consistent with that which

produced the rhyolitic Long Valley caldera-forming

eruption. However, that this unrest is foretelling either a

future caldera-forming event at LdM or an imminent

eruption of any particular style is not clear.

SUPPLEMENTARY DATA

Supplementary data for this paper are available at

Journal of Petrology online.

ACKNOWLEDGEMENTS

Wes Hildreth generously contributed samples and hasbeen a source of insight since the outset of this project.

Luis Torres Jara is thanked for invaluable guidance and

logistical support for navigating the Laguna. Meagan

Ankney, Allison Wende, and John Fournelle are

thanked for analytical assistance in obtaining the radio-

genic isotope and electron microprobe data. Amanda

Houts is thanked for laboratory assistance with chlorine

extraction. Robert Finkel and Susan Zimmerman are

thanked for careful 36Cl accelerator mass spectrometry

measurements and data reduction at CMAS-LLNL. Thiswork greatly benefited from many fruitful discussions

with Helene Le Mevel, Judy Fierstein, Paty Sruoga, Wes

Hildreth, and the LdM research group. Simon Barker,

Jorn-Frederik Wotzlaw, and Chad Deering are thanked

for insightful reviews, and Gerhard Woerner for editor-

ial handling. This research is supported by the US NSF

(EAR-1322595, EAR-1411779 to B.S.S.), the Geological

Society of America (9791-12, 10016-13 to N.L.A.), the

Wisconsin Alumni Research Foundation (WARF), and

University of Wisconsin Department of Geoscience giftfunds.

REFERENCES

Allan, A. S. R., Morgan, D. J., Wilson, C. J. N. & Millet, M.-A.(2013). From mush to eruption in centuries: assembly of thesuper-sized Oruanui magma body. Contributions toMineralogy and Petrology 166, 143–164.

Ankney, M. E., Johnson, C. M., Bacon, C. R., Beard, B. L. &Jicha, B. R. (2013). Distinguishing lower and upper crustalprocesses in magmas erupted during the buildup to the 7�7ka climactic eruption of Mount Mazama, Crater Lake,Oregon, using 238U–230Th disequilibria. Contributions toMineralogy and Petrology 166, 563–585.

Annen, C. (2009). From plutons to magma chambers: Thermalconstraints on the accumulation of eruptible silicic magmain the upper crust. Earth and Planetary Science Letters 284,409–416.

Annen, C., Blundy, J. D. & Sparks, R. S. J. (2006). The genesis ofintermediate and silicic magmas in deep crustal hot zones.Journal of Petrology 47, 505–539.

Bachmann, O. & Bergantz, G. W. (2004). On the origin of crystal-poor rhyolites: extracted from batholithic crystal mushes.Journal of Petrology 45, 1565–1582.

Bachmann, O., Dungan, M. A. & Bussy, F. (2005). Insights intoshallow magmatic processes in large silicic magma bodies:

Journal of Petrology, 2017, Vol. 58, No. 1 109

the trace element record in the Fish Canyon magma body,

Colorado. Contributions to Mineralogy and Petrology 149,338–349.

Bachmann, O., Deering, C. D., Ruprecht, J. S., Huber, C.,Skopelitis, A. & Schnyder, C. (2012). Evolution of silicic mag-

mas in the Kos–Nisyros volcanic center, Greece: A petrologi-cal cycle associated with caldera collapse. Contributions toMineralogy and Petrology 163, 151–166.

Bacon, C. R. & Druitt, T. H. (1988). Compositional evolution ofthe zoned calcalkaline magma chamber of Mount Mazama,

Crater Lake, Oregon. Contributions to Mineralogy andPetrology 98, 224–256.

Barker, S. J., Wilson, C. J. N., Smith, E. G. C., Charlier, B. L. A.,Wooden, J. L., Hiess, J. & Ireland, T. R. (2014). Post-superer-

uption magmatic reconstruction of Taupo volcano (NewZealand), as reflected in zircon ages and trace elements.Journal of Petrology 55, 1511–1533.

Barker, S. J., Wilson, C. J. N., Allan, A. S. R. & Schipper, C. I.(2015). Fine-scale temporal recovery, reconstruction and

evolution of a post-supereruption magmatic system.Contributions to Mineralogy and Petrology 170, 2–40.

Barker, S. J., Wilson, C. J. N., Morgan, D. J. & Rowland, J. V.(2016). Rapid priming, accumulation, and recharge of

magma driving recent eruptions at a hyperactive calderavolcano. Geology. G37382.1.

Beard, J. S. & Lofgren, G. E. (1991). Dehydration melting andwater-saturated melting of basaltic and andesitic green-stones and amphibolites at 1, 3, and 6�9 kb. Journal of

Petrology 32, 365–401.Begue, F., Deering, C. D., Gravley, D. M., Kennedy, B. M.,

Chambefort, I., Gualda, G. A. R. & Bachmann, O. (2014).Extraction, storage and eruption of multiple isolated magma

batches in the paired Mamaku and Ohakuri eruption,Taupovolcanic zone, New Zealand. Journal of Petrology 55,1653–1684.

Berlo, K., Turner, S., Blundy, J. & Hawkesworth, C. (2004). Theextent of U-series disequilibria produced during partial melt-

ing of the lower crust with implications for the formation ofthe Mount St. Helens dacites. Contributions to Mineralogyand Petrology 148, 122–130.

Bindeman, I. N. & Simakin, A. G. (2014). Rhyolites—hard to pro-

duce, but easy to recycle and sequester: Integrating micro-geochemical observations and numerical models.Geosphere 10, 930–957.

Bindeman, I. N., Davis, A. M. & Drake, M. J. (1998). Ion micro-probe study of plagioclase–basalt partition experiments at

natural concentration levels of trace elements. Geochimicaet Cosmochimica Acta 62, 1175–1193.

Bindeman, I. N., Fu, B., Kita, N. T. & Valley, J. W. (2008). Originand evolution of silicic magmatism at Yellowstone based on

ion microprobe analysis of isotopically zoned zircons.Journal of Petrology 49, 163–193.

Birsic, E. M. (2015). Petrology and 40Ar/39Ar chronology of theLaguna Sin Puerto and Bobadilla ignimbrites, Laguna delMaule Volcanic Field, Chile. MS thesis, University of

Wisconsin–Madison.Blundy, J. D. & Wood, B. J. (2003). Mineral–melt partitioning of

uranium, thorium, and their daughters. In: Bourdon, B.,Henderson, G. M., Lundstrom, C. C. & Turner, S. P. (eds)

U-series Geochemistry. Mineralogical Society of Americaand Geochemical Society, Reviews in Mineralogy andGeochemistry 52, 59–123.

Boehnke, P., Watson, E. B., Trail, D., Harrison, T. M. & Schmitt,A. K. (2013). Zircon saturation re-revisited. Chemical

Geology 351, 324–334.Brenan, J. M., Shaw, H. F., Ryerson, F. J. & Phinney, D. L.

(1995). Experimental determination of trace-element

partitioning between pargasite and a synthetic hydrous

andesitic melt. Earth and Planetary Science Letters 135,1–11.

Castro, J. M. & Dingwell, D. B. (2009). Rapid ascent of rhyolitic

magma at Chaiten Volcano, Chile. Nature 461, 780–783.Chamberlain, K. J., Morgan, D. J. & Wilson, C. J. N. (2014a).

Timescales of mixing and mobilization in the Bishop Tuff

magma body: Perspectives from diffusion chronometry.Contributions to Mineralogy and Petrology 168, 1–24.

Chamberlain, K. J., Wilson, C. J. N., Wooden, J. L., Charlier, B.

L. A. & Ireland, T. R. (2014b). New perspectives on theBishop Tuff from zircon textures, ages and trace elements.Journal of Petrology 55, 395–426.

Charlier, B. L. A. & Wilson, C. J. N. (2010). Chronology and evo-lution of caldera-forming and post-caldera magma systemsat Okataina Volcano, New Zealand from zircon U–Th model-age spectra. Journal of Petrology 51, 1121–1141.

Charlier, B. L. A., Wilson, C. J. N. & Davidson, J. P. (2008). Rapidopen-system assembly of a large silicic magma body: time-resolved evidence from cored plagioclase crystals in the

Oruanui eruption deposits, New Zealand. Contributions toMineralogy and Petrology 156, 799–813.

Charlier, B. L. A., Wilson, C. J. N., Lowenstern, J. B., Blake, S.,

Van Calsteren, P. W. & Davidson, J. P. (2005). Magma gener-ation at a large, hyperactive silicic volcano (Taupo, NewZealand) revealed by U–Th and U–Pb systematics in zircons.

Journal of Petrology 46, 3–32.Chiodini, G., Liccioli, C., Vaselli, O., Calabrese, S., Tassi, F.,

Caliro, S., Caselli, A., Agusto, M. & D’Alessandro, W. (2014).

The Domuyo volcanic system: An enormous geothermal re-source in Argentine Patagonia. Journal of Volcanology andGeothermal Research 274, 71–77.

Clark, P. U., Dyke, A. S., Shakun, J. D., Carlson, A. E., Clark, J.,Wohlfarth, B., Mitrovica, J. X., Hostetler, S. W. & McCabe, A.M. (2009). The Last Glacial Maximum. Science 325, 710–714.

Condomines, M. (1997). Dating recent volcanic rocks through230Th–238U disequilibrium in accessory minerals: Exampleof the Puy de Dome (French Massif Central). Geology 25,375–378.

Costa, F. (2008). Residence times of silicic magma associatedwith calderas. Developments in Volcanology 10, 1–55.

Costa, F. & Singer, B. S. (2002). Evolution of Holocene dacite

and compositionally zoned magma, Volcan San Pedro,Southern Volcanic Zone, Chile. Journal of Petrology 43,1571–1593.

Costa, F., Dungan, M. A. & Singer, B. S. (2002). Hornblende- andphlogopite-bearing gabbroic xenoliths from Volc�an SanPedro (36�S), Chilean Andes: evidence for melt and fluid mi-

gration and reactions in subduction-related plutons. Journalof Petrology 43, 219–241.

Davidson, J., Dungan, M., Ferguson, K. M. & Colucci, M. T.

(1987). Crust–magma interactions and the evolution of arcmagmas: The San Pedro–Pellado volcanic complex, south-ern Chilean Andes. Geology 15, 443–446.

Davidson, J., Turner, S., Handley, H., Macpherson, C. &Dosseto, A. (2007). Amphibole ‘sponge’ in arc crust?.Geology 35, 787.

Davidson, J., Turner, S. & Plank, T. (2013). Dy/Dy*: variations

arising from mantle sources and petrogenetic processes.Journal of Petrology 54, 525–537.

Deering, C. D., Cole, J. W. & Vogel, T. A. (2008). A rhyolite com-

positional continuum governed by lower crustal source con-ditions in the Taupo Volcanic Zone, New Zealand. Journal ofPetrology 49, 2245–2276.

Deering, C. D., Cole, J. W. & Vogel, T. A. (2011). Extraction ofcrystal-poor rhyolite from a hornblende-bearing intermedi-ate mush: a case study of the caldera-forming Matahina

110 Journal of Petrology, 2017, Vol. 58, No. 1

eruption, Okataina volcanic complex. Contributions to

Mineralogy and Petrology 161, 129–151.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. Contributions to

Mineralogy and Petrology 160, 609–629.DeMets, C., Gordon, R. G. & Argus, D. F. (2010). Geologically

current plate motions. Geophysical Journal International

181, 1–80.Douglass, J., Schilling, J.-G. & Fontignie, D. (1999). Plume–

ridge interactions of the Discovery and Shona mantle

plumes with the southern Mid-Atlantic Ridge (40�–55�S).

Journal of Geophysical Research 104, 2941.Druitt, T. H., Costa, F., Deloule, E., Dungan, M. & Scaillet, B.

(2012). Decadal to monthly timescales of magma transfer

and reservoir growth at a caldera volcano. Nature 482,

77–80.Dungan, M., Wulff, A. & Thompson, R. (2001). Eruptive stratig-

raphy of the Tatara–San Pedro Complex, 36�S, Southern

Volcanic Zone, Chilean Andes: reconstruction method and

implications for magma evolution at long-lived arc volcanic

centers. Journal of Petrology 42, 555–626.Dunn, T. & Sen, C. (1994). Mineral/matrix partition coefficients

for orthopyroxene, plagioclase, and olivine in basaltic to an-

desitic systems: A combined analytical and experimental

study. Geochimica et Cosmochimica Acta 58, 717–733.Evans, B. W., Hildreth, W., Bachmann, O. & Scaillet, B. (2016). In

defense of magnetite–ilmenite thermometry in the Bishop

Tuff and its implication for gradients in silicic magma reser-

voirs. American Mineralogist 101, 469–482.Ewart, A. & Griffin, W. L. (1994). Application of proton-

microprobe data to trace-element partitioning in volcanic

rocks. Chemical Geology 117, 251–284.Feigl, K. L., Le Mevel, H., Tabrez Ali, S., C�ordova, L., Andersen,

N. L., DeMets, C. & Singer, B. S. (2014). Rapid uplift in

Laguna del Maule volcanic field of the Andean southern vol-

canic zone (Chile) 2007–2012. Geophysical Journal

International 196, 885–901.Fierstein, J., Sruoga, P., Amigo, �A., Elissondo, M. & Rosas, M.

(2012). Postglacial eruptive history of Laguna del Maule vol-

canic field in Chile, from fallout stratigraphy in Argentina.

AGU Fall Meeting V31F-03.Folguera, A., Alasonati Tasarova, Z., Gotze, H. J., Rojas Vera, E.,

Gimenez, M. & Ramos, V. A. (2012). Retroarc extension in

the last 6 Ma in the South–Central Andes (36�S–40�S) eval-

uated through a 3-D gravity modelling. Journal of South

American Earth Sciences 40, 23–37.Fournier, T. J., Pritchard, M. E. & Riddick, S. N. (2010). Duration,

magnitude, and frequency of subaerial volcano deformation

events: New results from Latin America using InSAR and a

global synthesis. Geochemistry, Geophysics, Geosystems

11, Q01003.Garrison, J., Davidson, J., Reid, M. R. & Turner, S. (2006).

Source versus differentiation controls on U-series disequili-

bria: Insights from Cotopaxi Volcano, Ecuador. Earth and

Planetary Science Letters 244, 548–565.Gelman, S. E., Gutierrez, F. J. & Bachmann, O. (2013). On the

longevity of large upper crustal silicic magma reservoirs.

Geology 41, 759–762.Gerlach, D. D. C., Frey, F. A. F., Moreno-Roa, H. & Lopez-

Escobar, L. (1988). Recent volcanism in the Puyehue–

Cordon Caulle region, southern Andes, Chile (40�5�S): petro-

genesis of evolved lavas. Journal of Petrology 29, 333–382.Ghiorso, M. S. & Evans, B. W. (2008). Thermodynamics of

rhombohedral oxide solid solutions and a revision of the

Fe–Ti two-oxide geothermometer and oxygen-barometer.

American Journal of Science 308, 957–1039.Gilbert, H., Beck, S. & Zandt, G. (2006). Lithospheric and upper

mantle structure of central Chile and Argentina.Geophysical Journal International 165, 383–398.

Glazner, A. F., Bartley, J. M., Coleman, D. S., Gray, W. & Taylor,

R. Z. (2004). Are plutons assembled over millions of years byamalgamation from small magma chambers?. GSA Today14, 4–11.

Grunder, A. L. (1987). Low d18O silicic volcanic rocks at theCalabozos Caldera Complex, Southern Andes, Evidence forupper-crustal contamination. Contributions to Mineralogy

and Petrology 95, 71–81.Grunder, A. L. & Mahood, G. A. (1988). Physical and chemical

models of zoned silicic magmas: the Loma Seca Tuff andCalabozos Caldera, Southern Andes. Journal of Petrology29, 831–867.

Gualda, G. A. R. & Ghiorso, M. S. (2013). Low-pressure origin ofhigh-silica rhyolites and granites. Journal of Geology 121,537–545.

Gualda, G. A. R., Ghiorso, M. S., Lemons, R. V. & Carley, T. L.(2012). Rhyolite-MELTS: a modified calibration of MELTSoptimized for silica-rich, fluid-bearing magmatic systems.

Journal of Petrology 53, 875–890.Hart, S. R. (1984). A large-scale isotope anomaly in the

Southern Hemisphere mantle. Nature 309, 753–757.Hawthorne, F. C., Oberti, R., Harlow, G. E., Maresch, W. V.,

Martin, R. F., Schumacher, J. C. & Welch, M. D. (2012). IMA

report: Nomenclature of the amphibole supergroup.American Mineralogist 97, 2031–2048.

Hein, A. S., Hulton, N. R. J., Dunai, T. J., Sugden, D. E., Kaplan,

M. R. & Xu, S. (2010). The chronology of the Last GlacialMaximum and deglacial events in central ArgentinePatagonia. Quaternary Science Reviews 29, 1212–1227.

Hickey-Vargas, R., Roa, H. M., Escobar, L. L. & Frey, F. a. (1989).Geochemical variations in Andean basaltic and silicic lavasfrom the Villarrica-Lanin volcanic chain (39.5� S): an evalua-

tion of source heterogeneity, fractional crystallization andcrustal assimilation. Contributions to Mineralogy and

Petrology 103, 361–386.Hickey-Vargas, R., Sun, M., L�opez-Escobar, L., Moreno-Roa, H.,

Reagan, M. K., Morris, J. D. & Ryan, J. G. (2002). Multiple

subduction components in the mantle wedge: Evidencefrom eruptive centers in the Central Southern volcanic zone,Chile. Geology 30, 199–202.

Hildreth, W. (1981). Gradients in silicic magma chambers: impli-cations for lithospheric magmatism. Journal of GeophysicalResearch: Solid Earth 86, 10153–10192.

Hildreth, W. (1983). The compositionally zoned eruption of 1912in the Valley of Ten Thousand Smokes, Katmai NationalPark, Alaska. Journal of Volcanology and Geothermal

Research 18, 1–56.Hildreth, W. (2004). Volcanological perspectives on Long

Valley, Mammoth Mountain, and Mono Craters: several con-tiguous but discrete systems. Journal of Volcanology andGeothermal Research 136, 169–198.

Hildreth, W. & Fierstein, J. (2012). The Novarupta–KatmaiEruption of 1912—Largest Eruption of the TwentiethCentury: Centennial Perspectives. US Geological Survey,

Professional Papers 259.Hildreth, W. & Moorbath, S. (1988). Crustal contributions to arc

magmatism in the Andes of Central Chile. Contributions to

Mineralogy and Petrology 98, 455–489.Hildreth, W. & Wilson, C. J. N. (2007). Compositional zoning of

the Bishop Tuff. Journal of Petrology 48, 951–999.Hildreth, W., Grunder, A. L. & Drake, R. E. (1984). The Loma

Seca Tuff and the Calabozos caldera: A major ash-flow and

Journal of Petrology, 2017, Vol. 58, No. 1 111

caldera complex in the southern Andes of central Chile.

Geological Society of America Bulletin 95, 45–54.Hildreth, W., Fierstein, J., Godoy, E., Drake, R. E. & Singer, B. S.

(1999). The Puelche Volcanic Field: extensive Pleistocenerhyolite lava flows in the Andes of central Chile. Revista

Geol�ogica de Chile 26, 1–31.Hildreth, W., Godoy, E., Fierstein, J. & Singer, B. S. (2010).

Laguna del Maule Volcanic Field: Eruptive History of aQuaternary basalt-to-rhyolite distributed volcanic field onthe Andean rangecrest in central Chile. Santiago: Servicio

Nacional de Geologia y Mineria.Holm, P. M., Søager, N., Dyhr, C. T. & Nielsen, M. R. (2014).

Enrichments of the mantle sources beneath the SouthernVolcanic Zone (Andes) by fluids and melts derived from

abraded upper continental crust. Contributions toMineralogy and Petrology 167, 1004.

Hora, J. M., Singer, B. S., Worner, G., Beard, B. L., Jicha, B. R. &Johnson, C. M. (2009). Shallow and deep crustal control ondifferentiation of calc-alkaline and tholeiitic magma. Earth

and Planetary Science Letters 285, 75–86.Hubbard, A., Hein, A. S., Kaplan, M. R., Hulton, N. R. J. &

Glasser, N. (2005). A modelling reconstruction of the lastglacial maximum ice sheet and its deglaciation in the vicinity

of the northern Patagonian icefield, South America.Geografiska Annaler 87, 375–391.

Hughes, G. R. & Mahood, G. A. (2011). Silicic calderas in arc set-tings: Characteristics, distribution, and tectonic controls.Geological Society of America Bulletin 123, 1577–1595.

Jacques, G., Hoernle, K., Gill, J., Hauff, F., Wehrmann, H.,Garbe-Schonberg, D., van den Bogaard, P., Bindeman, I. N.& Lara, L. E. (2013). Across-arc geochemical variations in theSouthern Volcanic Zone, Chile (34�5–38�0�S): Constraints on

mantle wedge and slab input compositions. Geochimica etCosmochimica Acta 123, 218–243.

Jicha, B. R., Singer, B. S., Beard, B. L., Johnson, C. M., Moreno-Roa, H. & Naranjo, J. A. (2007). Rapid magma ascent andgeneration of 230Th excesses in the lower crust at Puyehue–

Cord�on Caulle, Southern Volcanic Zone, Chile. Earth andPlanetary Science Letters 255, 229–242.

Jicha, B. R., Johnson, C. M., Hildreth, W., Beard, B. L., Hart, G.L., Shirey, S. B. & Singer, B. S. (2009). Discriminating assimi-

lants and decoupling deep- vs. shallow-level crystal recordsat Mount Adams using 238U–230Th disequilibria and Os iso-topes. Earth and Planetary Science Letters 277, 38–49.

Jicha, B. R., Singer, B. S. & Sobol, P. (2016). Re-evaluation ofthe ages of 40Ar/39Ar sanidine standards and supereruptions

in the western U.S. using a Noblesse multi-collector massspectrometer. Chemical Geology 431, 54–66.

Kaplan, M. R., Ackert, R. P., Singer, B. S., Douglass, D. C. &Kurz, M. D. (2004). Cosmogenic nuclide chronology of

millennial-scale glacial advances during O-isotope stage 2in Patagonia. Geological Society of America Bulletin 116,308–321.

Klemetti, E. W., Deering, C. D., Cooper, K. M. & Roeske, S. M.(2011). Magmatic perturbations in the Okataina Volcanic

Complex, New Zealand at thousand-year timescales re-corded in single zircon crystals. Earth and Planetary ScienceLetters 305, 185–194.

Kuiper, K. F., Deino, A., Hilgen, F. J., Krijgsman, W., Renne, P. R.

& Wijbrans, J. R. (2008). Synchronizing rock clocks of Earthhistory. Science 320, 500–504.

Lee, J. Y., Marti, K., Severinghaus, J. P., Kawamura, K., Yoo, H.S., Lee, J. B. & Kim, J. S. (2006). A redetermination of the iso-topic abundances of atmospheric Ar. Geochimica et

Cosmochimica Acta 70, 4507–4512.Le Mevel, H., Feigl, K. L., C�ordova, L., DeMets, C. & Lundgren, P.

(2015). Evolution of unrest at Laguna del Maule volcanic

field (Chile) from InSAR and GPS measurements, 2003 to

2014. Geophysical Research Letters 42, 6590–6598.Le Mevel, H., Gregg, P. M. & Feigl, K. L. (2016). Magma injection

into long-lived reservoir to explain geodetically measureduplift: Application to the 2004–2014 episode at Laguna delMaule volcanic field, Chile. Journal of GeophysicalResearch 121, 6092–6108.

Lucassen, F., Trumbull, R., Franz, G., Creixell, C., V�asquez, P.,Romer, R. L. & Figueroa, O. (2004). Distinguishing crustalrecycling and juvenile additions at active continental mar-gins: the Paleozoic to recent compositional evolution of theChilean Pacific margin (36–41�S). Journal of SouthAmerican Earth Sciences 17, 103–119.

Lucassen, F., Wiedicke, M. & Frantz, G. (2010). Complete recy-cling of a magmatic arc: evidence from chemical and iso-topic composition of Quaternary trench sediments.International Journal of Earth Science 99, 687–701.

Luhr, J. F. & Carmichael, I. S. E. (1980). The Colima Volcaniccomplex, Mexico. Contributions to Mineralogy andPetrology 71, 343–372.

Mahood, G. A. (1990). Second reply to comment of R. S. J.Sparks, H. E. Huppert, and C. J. N. Wilson on ‘Evidence forlong residence times of rhyolitic magma in the Long Valley

magmatic system: the isotopic record if precaldera lavas ofGlass Mountain’. Earth and Planetary Science Letters 99,395–399.

McMillan, N. J., Harmon, R. S. R., Moorbath, S., Lopez-Escobar,L. & Strong, D. F. (1989). Crustal sources involved in contin-ental arc magmatism: a case study of Volcan Mocho–

Choshuenco, southern Chile. Geology 17, 1152–1156.Metz, J. M. & Mahood, G. A. (1985). Precursors to the Bishop

Tuff Eruption: Glass Mountain, Long Valley, California.Journal of Geophysical Research 90, 11121.

Metz, J. M. & Mahood, G. A. (1991). Development of the LongValley, California, magma chamber recorded in precalderarhyolite lavas of Glass Mountain. Contributions toMineralogy and Petrology 106, 379–397.

Miller, C. A., Williams-Jones, G., Fournier, D. & Witter, J. (2016).3D gravity inversion and thermodynamic modelling reveal

properties of shallow silicic magma reservoir beneathLaguna del Maule, Chile. Earth and Planetary ScienceLetters 459, 14–27.

Miller, C. D. (1985). Holocene eruptions at the Inyo volcanicchain, California: Implications for possible eruptions in LongValley caldera. Geology 13, 14–17.

Miranda, F., Folguera, A., Leal, P., Naranjo, J. & Pesce, A.(2006). Upper Pliocene to Lower Pleistocene volcanic com-plexes and Upper Neogene deformation in the south–central Andes (36�30’–38�S). In: Kay, S. M. & Ramos, V. A.(eds) Evolution of an Andean Margin: A Tectonic andMagmatic View from the Andes to the Neuquen Basin (35�–

39�S lat). Geological Society of America Special Paper 407,287–298.

Myers, M. L., Wallace, P. J., Wilson, C. J. N., Morter, B. K. &Swallow, E. J. (2016). Prolonged ascent and episodic ventingof discrete magma batches at the onset of the HuckleberryRidge supereruption, Yellowstone. Earth and Planetary

Science Letters 451, 285–297.Nelson, S. T., Davidson, J. P., Heizler, M. T. & Kowallis, B. J.

(1999). Tertiary tectonic history of the southern Andes: Thesubvolcanic sequence to the Tatara–San Pedro volcaniccomplex, lat 36�S. Geological Society of America Bulletin111, 1387–1404.

Ottonello, G., Joron, J. L. & Piccardo, G. B. (1984). Rare earthand 3d transition element geochemistry of peridotitic rocks:II. Ligurian peridotites and associated basalts. Journal ofPetrology 25, 373–393.

112 Journal of Petrology, 2017, Vol. 58, No. 1

Ownby, S. E., Lange, R. A., Hall, C. M. & Delgado-Granados, H.

(2011). Origin of andesite in the deep crust and eruptionrates in the Tancitaro–Nueva Italia region of the CentralMexican Arc. Geological Society of America Bulletin 123,274–294.

Palme, H. & O’Neill, H. S. C. (2003). Cosmochemical estimatesof mantle composition. In: Carlson, R. W. (ed.) Treatise onGeochemistry, Vol. 2: The Mantle and Core. Amsterdam:Elsevier, pp. 1–38.

Putirka, K. (2016). Amphibole thermometers and barometersfor igneous systems, and some implications for eruptionmechanisms of felsic magmas at arc volcanoes. American

Mineralogist 101, 841–858.Qian, Q. & Hermann, J. (2013). Partial melting of lower crust at

10–15 kbar: constraints on adakite and TTG formation.Contributions to Mineralogy and Petrology 165, 1195–1224.

Rawson, H., Naranjo, J. A., Smith, V., Fontijn, K., Pyle, D. M.,Mather, T. A. & Moreno, H. (2015). The frequency and mag-nitude of post-glacial explosive eruptions at Volc�an Mocho–Choshuenco, southern Chile. Journal of Volcanology and

Geothermal Research 299, 103–129.Reid, M. R. (2008). How long does it take to supersize an erup-

tion. Elements 4, 23–28.Reid, M. R., Vazquez, J. A. & Schmitt, A. K. (2011). Zircon-scale

insights into the history of a supervolcano, Bishop Tuff,Long Valley, California, with implications for the Ti-in-zircongeothermometer. Contributions to Mineralogy and

Petrology 161, 293–311.Reubi, O., Bourdon, B., Dungan, M. A., Koornneef, J. M., Selles,

D., Langmuir, C. H. & Aciego, S. (2011). Assimilation of theplutonic roots of the Andean arc controls variations in U-ser-

ies disequilibria at Volcan Llaima, Chile. Earth and PlanetaryScience Letters 303, 37–47.

Ridolfi, F. & Renzulli, A. (2012). Calcic amphiboles in calc-

alkaline and alkaline magmas: Thermobarometric and che-mometric empirical equations valid up to 1,130�C and2�2 GPa. Contributions to Mineralogy and Petrology 163,877–895.

Ridolfi, F., Renzulli, A. & Puerini, M. (2010). Stability and chem-ical equilibrium of amphibole in calc-alkaline magmas: Anoverview, new thermobarometric formulations and applica-tion to subduction-related volcanoes. Contributions to

Mineralogy and Petrology 160, 45–66.Rubatto, D. & Hermann, J. (2007). Experimental zircon/melt and

zircon/garnet trace element partitioning and implications forthe geochronology of crustal rocks. Chemical Geology 241,

38–61.Rubin, A. M. (1995). Getting granite dikes out of the source re-

gion. Journal of Geophysical Research 100, 5911.Rubin, A., Cooper, K. M., Leever, M., Wimpenny, J., Deering, C.,

Rooney, T., Gravley, D. & Yin, Q. Z. (2016). Changes inmagma storage conditions following caldera collapse atOkataina Volcanic Center, New Zealand. Contributions to

Mineralogy and Petrology 171, 1–18.Rudnick, R. L. & Gao, S. (2003). Composition of the continental

crust. In: Rudnick, R. L. (ed.) Treatise on Geochemistry, Vol.3: The Crust. Elsevier, pp. 1–64.

Ruprecht, P. & Cooper, K. M. (2012). Integrating the uranium-series and elemental diffusion geochronometers in mixedmagmas from Volcan Quizapu, central Chile. Journal ofPetrology 53, 841–871.

Sano, Y., Terada, K. & Fukuoka, T. (2002). High mass resolutionion microprobe analysis of rare earth elements in silicateglass, apatite and zircon: lack of matrix dependency.

Chemical Geology 184, 217–230.Shane, P., Martin, S. B., Smith, V. C., Beggs, K. F., Darragh, M.

B., Cole, J. W. & Nairn, I. A. (2007). Multiple rhyolite magmas

and basalt injection in the 17�7 ka Rerewhakaaitu eruption

episode from Tarawera volcanic complex, New Zealand.

Journal of Volcanology and Geothermal Research 164,

1–26.Shane, P., Nairn, I. A., Smith, V. C., Darragh, M., Beggs, K. &

Cole, J. W. (2008). Silicic recharge of multiple rhyolite mag-

mas by basaltic intrusion during the 22�6 ka Okareka

Eruption Episode, New Zealand. Lithos 103, 527–549.Sigmarsson, O., Chmeleff, J., Morris, J. & Lopez-Escobar, L.

(2002). Origin of 226Ra–230Th disequilibria in arc lavas from

southern Chile and implications for magma transfer time.

Earth and Planetary Science Letters 196, 189–196.Simon, J. I., Reid, M. R. & Young, E. D. (2007). Lead isotopes by

LA-MC-ICPMS: Tracking the emergence of mantle signa-

tures in an evolving silicic magma system. Geochimica et

Cosmochimica Acta 71, 2014–2035.Singer, B. S., Hildreth, W. & Vincze, Y. (2000). 40Ar/39Ar evi-

dence for early deglaciation of the Central Chilean Andes.

Geophysical Research Letters 27, 1663–1666.Singer, B. S., Jicha, B. R., Harper, M. A., Naranjo, J. A., Lara, L.

E. & Moreno-Roa, H. (2008). Eruptive history, geochron-

ology, and magmatic evolution of the Puyehue–Cord�on

Caulle volcanic complex, Chile. Geological Society of

America Bulletin 120, 599–618.Singer, B. S., Andersen, N. L., Le Mevel, H. et al. (2014).

Dynamics of a large, restless, rhyolitic magma system at

Laguna del Maule, southern Andes, Chile. GSA Today 24,

4–10.Singer, B. S., Tikoff, B., Le Mevel, H., Andersen, N. L., Cordova,

L. & Licciardi, J. (2015). Linking modern, rapid, surface uplift,

at the Laguna del Maule volcanic filed, Chilean Andes, to

rhyolitic magma-driven uplift spanning the Holocene. AGU

Fall Meeting G31C–04.Singer, B. S., Costa, F., Herrin, J. S., Hildreth, W. & Fierstein, J.

(2016). The timing of compositionally-zoned magma reser-

voirs and mafic ‘priming’ weeks before the 1912 Novarupta–

Katmai rhyolite eruption. Earth and Planetary Science

Letters 451, 125–137.Sisson, T. W. (1994). Hornblende–melt trace-element partition-

ing measured by ion microprobe. Chemical Geology 117,

331–344.Smith, V. C., Shane, P. & Nairn, I. A. (2004). Reactivation of a

rhyolitic magma body by new rhyolitic intrusion before the

15�8 ka Rotorua eruptive episode: implications for magma

storage in the Okataina Volcanic Centre, New Zealand.

Journal of the Geological Society, London 161, 757–772.Smith, V., Shane, P. & Nairn, I. (2005). Trends in rhyolite geo-

chemistry, mineralogy, and magma storage during the last

50 kyr at Okataina and Taupo volcanic centres, Taupo

Volcanic Zone, New Zealand. Journal of Volcanology and

Geothermal Research 148, 372–406.Smith, V., Shane, P. & Nairn, I. (2010). Insights into silicic melt

generation using plagioclase, quartz and melt inclusions

from the caldera-forming Rotoiti eruption, Taupo volcanic

zone, New Zealand. Contributions to Mineralogy and

Petrology 160, 951–971.Sruoga, P. (2015). Actividad explosiva postglacial del centro

Barrancas, complejo volcanico Laguna del Maule (36�05’S,

70�30’O). Congreso Geologico Chileno.Sruoga, P., Etcheverr, M. P., Feineman, M., Rosas, M., Burkert,

C. & Iba, O. (2012). Complejo Caldera Diamante–Volc�an

Maipo (34�10’S, 69�50’O): evoluci�on volcanol�ogica y geoqu�ı-

mica e implicancias en su peligrosidad. Revista de la

Asociancion Geologica Argentina 69, 508–530.Stix, J. & Gorton, M. P. (1993). Replenishment and crystalliza-

tion in epicontinental silicic magma chambers: evidence

Journal of Petrology, 2017, Vol. 58, No. 1 113

from the Bandelier magmatic system. Journal ofVolcanology and Geothermal Research 55, 201–215.

Storm, S., Shane, P., Schmitt, A. K. & Lindsay, J. M. (2011).Contrasting punctuated zircon growth in two syn-eruptedrhyolite magmas from Tarawera volcano: Insights to crystaldiversity in magmatic systems. Earth and Planetary ScienceLetters 301, 511–520.

Storm, S., Schmitt, A. K., Shane, P. & Lindsay, J. M. (2014).Zircon trace element chemistry at sub-micrometer reso-lution for Tarawera volcano, New Zealand, and implicationsfor rhyolite magma evolution. Contributions to Mineralogyand Petrology 167. doi:10.1007/s00410-014-1000-z

Sun, M. (2001). Geochemical variation among small eruptivecenters in the central SVZ of the Andes: An evaluation ofsubduction, mantle, and crustal influences. PhD thesis,Florida International University, Miami.

Sun, S. & McDonough, W. F. (1989). Chemical and isotopic sys-tematics of oceanic basalts: implications for mantle com-position and processes. In: Saunders, A. D. & Norry, M. J.(eds) Magmatism in the Ocean Basins. Geological Society,London, Special Publications 42, 313–345.

Sutton, A. N., Blake, S., Wilson, C. J. N. & Charlier, B. L. A.(2000). Late Quaternary evolution of a hyperactive rhyolitemagmatic system: Taupo volcanic centre, New Zealand.Journal of the Geological Society, London 157, 537–552.

Tassara, A. & Echaurren, A. (2012). Anatomy of the Andeansubduction zone: Three-dimensional density modelupgraded and compared against global-scale models.Geophysical Journal International 189, 161–168.

Tassara, A., Gotze, H. J., Schmidt, S. & Hackney, R. (2006).Three-dimensional density model of the Nazca plate and theAndean continental margin. Journal of GeophysicalResearch: Solid Earth 111, 1–26.

Till, C. B., Vazquez, J. A. & Boyce, J. W. (2015). Months betweenrejuvenation and volcanic eruption at Yellowstone caldera,Wyoming. Geology 43, 695–698.

Turner, S., Bourdon, B. & Gill, J. (2003). Insights into magmagenesis at convergent margins from U-series isotopes. In:Bourdon, B., Henderson, G. M., Lundstrom, C. C. & Turner,S. P. (eds) U-series Geochemistry. Mineralogical Society ofAmerica and Geochemical Society, Reviews in Mineralogyand Geochemistry 52, 255–315.

Turner, S., Sandiford, M., Reagan, M., Hawkesworth, C. &Hildreth, W. (2010). Origins of large-volume, composition-ally zoned volcanic eruptions: New constraints from U-ser-ies isotopes and numerical thermal modeling for the 1912Katmai–Novarupta eruption. Journal of GeophysicalResearch 115, B12201.

Vandergoes, M. J., Hogg, A. G., Lowe, D. J., et al. (2013). Arevised age for the Kawakawa/Oruanui tephra, a key markerfor the Last Glacial Maximum in New Zealand. QuaternaryScience Reviews 74, 195–201.

Vazquez, J. A. & Reid, M. R. (2004). Probing the accumulationhistory of the voluminous Toba magma. Science 305,991–994.

Villemant, B. (1988). Trace-Element Evolution in the PhlegreanFields (Central-Italy) – Fractional Crystallization andSelective Enrichment. Contributions to Mineralogy andPetrology 98, 169–183.

Waters, L. E. & Lange, R. A. (2015). An updated calibration ofthe plagioclase–liquid hygrometer–thermometer applicableto basalts through rhyolites. American Mineralogist 100,2172–2184.

Wark, D. A., Hildreth, W., Spear, F. S., Cherniak, D. J. & Watson,E. B. (2007). Pre-eruption recharge of the Bishop magmasystem. Geology 35, 235–238.

Watson, E. B. & Harrison, T. M. (1983). Zircon saturation re-visited: temperature and composition effects in a variety ofcrustal magma types. Earth and Planetary Science Letters64, 295–304.

Wende, A. M., Johnson, C. M. & Beard, B. L. (2015). Tracingchanges in mantle and crustal influences in individual cone-building stages at Mt. Shasta using U–Th and Sr isotopes.Earth and Planetary Science Letters 428, 11–21.

Wilson, C. J. N. & Charlier, B. L. A. (2009). Rapid rates of magmageneration at contemporaneous magma systems, TaupoVolcano, New Zealand: insights from U–Th model-age spec-tra in zircons. Journal of Petrology 50, 875–907.

Wilson, C. J. N., Blake, S., Charlier, B. L. A. & Sutton, A. N.(2006). The 26�5 ka Oruanui eruption, Taupo Volcano, NewZealand: development, characteristics and evacuation of alarge rhyolitic magma body. Journal of Petrology 47, 35–69.

Wilson, C. J. N., Gravley, D. M., Leonard, G. S. & Rowland, J. V.(2009). Volcanism in the cental Taupo Volcanic Zone, NewZealand: tempo, styles, and controls. In: Thordarson, T.,Self, S., Larsen, G., Rowland, S. K. & Hoskuldsson, A. (eds)Studies in Volcanology: The Legacy of George Walker.Geological Society, pp. 225–247.

Wolf, M. B. & Wyllie, P. J. (1993). Garnet growth during am-phibolite anatexis: implications of a garnetiferous restite.Journal of Geology 101, 357–373.

Wolff, J. A., Ellis, B. S., Ramos, F. C., Starkel, W. A., Boroughs,S., Olin, P. H. & Bachmann, O. (2015). Remelting of cumu-lates as a process for producing chemical zoning in silicictuffs: A comparison of cool, wet and hot, dry rhyoliticmagma systems. Lithos 236–237, 275–286.

Wotzlaw, J.-F., Schaltegger, U., Frick, D. A., Dungan, M. A.,Gerdes, A. & Gunther, D. (2013). Tracking the evolution oflarge-volume silicic magma reservoirs from assembly tosupereruption. Geology 41, 867–870.

Wotzlaw, J.-F., Bindeman, I. N., Stern, R. A., D’Abzac, F.-X. &Schaltegger, U. (2015). Rapid heterogeneous assembly ofmultiple magma reservoirs prior to Yellowstone supererup-tions. Scientific Reports 5, 14026.

114 Journal of Petrology, 2017, Vol. 58, No. 1