Bachmann 2004
Bachmann 2004
Bachmann 2004
1093/petrology/egh019
The largest accumulations of rhyolitic melt in the upper crust occur in (hereafter referred to as ‘silicic’) remains a major
voluminous silicic crystal mushes, which sometimes erupt as challenge of igneous petrology. In particular, the origin
unzoned, crystal-rich ignimbrites, but are most frequently preserved of rhyolites, which are typically the most crystal-poor,
as granodioritic batholiths. After approximately 40–50% crystal- despite being the most viscous silicate liquids that the
lization, magmas of intermediate composition (andesite–dacite) typic- Earth produces, remains elusive. Combinations of two
ally contain high-SiO2 interstitial melt, similar to crystal-poor end-member mechanisms are commonly invoked to
rhyolites commonly erupted in mature arc and continental settings. explain such magmas (Fig. 1a): (1) partial melting of
This paper analyzes the feasibility of system-wide extraction of this crustal material; (2) fractional crystallization of a more
melt from the mush, a mechanism that can rationalize a number mafic parent. Although systems dominated by crustal
of observations in both the plutonic and volcanic record, such as: melting occur in some settings (e.g. peraluminous
(1) abrupt compositional gaps in ignimbrites; (2) the presence of volcanics: Clemens & Wall, 1984; Munksgaard, 1984;
chemically highly evolved bodies at the roof of subvolcanic batholiths; Pichavant et al., 1988a, 1988b; intra-oceanic rhyolites:
(3) the observed range of ages (up to 200–300 ka) recorded by Gunnarsson et al., 1998; Smith et al., 2003; igneous
zircons in silicic magmas; (4) extensive zones of low P-wave velocity provinces related to continental break-up: Green &
in the shallow crust under active silicic calderas. We argue that Fitz-Thomas, 1993; Milner et al., 1995; Riley et al.,
crystal–melt segregation occurs by a combination of several processes 2001), many geochemical studies suggest that fractional
(hindered settling, micro-settling, compaction) once convection is crystallization plays an important role in the generation
hampered as the rheological locking point of the crystal–melt mixture of silicic magmas (e.g. Michael, 1983; Bacon & Druitt,
(50 vol. % crystals) is attained. We constrain segregation rates by 1988; Mahood & Halliday, 1988; Hildreth et al., 1991;
using hindered settling velocities and compaction rates as end- DePaolo et al., 1992; Hildreth & Fierstein, 2000; Lindsay
members. Time scales estimated for the formation of >500 km3 of et al., 2001; Clemens, 2003). This is particularly true for
crystal-poor rhyolite range from 104 to 105 years, within the silicic volcanic rocks, which are known to have a more
estimated residence times of mushes in the upper crust (>105 pronounced crystal fractionation signature than their
years, largely based on U/Th and U/Pb dating). This model plutonic equivalents (Halliday et al., 1991; Fig. 1b).
provides an integrated picture of silicic magmatism, linking the For magmas to evolve by fractional crystallization,
evolution of plutonic and volcanic systems until storage in the upper differential motion must occur between crystals and
crust, where granitoids become the leftovers from rhyolitic eruptions. melt (two-phase flow, where all crystals are considered
to be one effective phase). Despite numerous attempts, no
consensus has emerged on the dominant process occur-
KEY WORDS: crustal evolution; igneous processes; rhyolite; silicic ring in the viscous, silicic systems. Three influential
magmatism; two-phase flow hypotheses, all assuming an initially near-liquidus batch
of silicate melt, are: (1) crystal settling, commonly advoc-
ated since Bowen (1928); (2) convective fractionation in a
INTRODUCTION crystallizing, double-diffusive boundary layer (Chen &
Establishing the processes and controls on the generation Turner, 1980; McBirney, 1980; Rice, 1981; McBirney
of magmas with SiO2 content in excess of 65 wt % et al., 1985; Spera et al., 1995); (3) the ‘solidification front
1
Volcanic rocks a
Plutonic rocks
NCI
0.5
100
10
=
Sr
10
b/
R
Basalts
1
00
01
0.
1
10
0.
0.
1
=
=
=
=
Sr
=
Sr
Sr
Sr
b/
Sr
b/
b/
b/
R
b/
R
1
R
R
Fig. 1. (a) Variation of the NCI value (Neodymium Crustal Index; DePaolo et al., 1992) for cogenetic volcanic and plutonic rocks from the
Atesina–Cima d’Asta volcano-plutonic complex (Barth et al., 1993), Latir magmatic center (Johnson et al., 1990), as well as some 20–40 Ma
rhyolites and granitoids from the western USA (DePaolo et al., 1992). (b) Rb vs Sr content for magmatic rocks from several magmatic provinces
(western North America, Andes, Great Britain, Australia; modified from Halliday et al., 1991). Dashed line represents a crystal fractionation trend,
using the concentration of Rb and Sr in the interstitial melt of the Fish Canyon Tuff (FCT; a typical monotonous intermediate) as initial value and
typical bulk partition coefficients for these elements in FCT (DSr ¼ 12, DRb ¼ 05; Bachmann, 2001). Tick marks indicate 15 and 30% of Rayleigh
fractional crystallization. The more restricted, but overlapping NCI range displayed by the volcanic sequences suggests that the erupted magmas
are a subset of a consanguineous petrological suite, with both crystal fractionation and crustal assimilation playing a role in the generation of these
magmas. However, the rhyolites record a stronger crystal fractionation signature (higher Rb/Sr ratios).
instabilities’ hypothesis of Marsh and coworkers [see et al., 2003). The genetic link between silicic volcanic
Marsh (2002) for a review]. As we discuss below, none units and shallow granitoids (e.g. Lipman, 1984; Halliday
of these processes, taken alone, appears to satisfy the et al., 1991; Fig. 1), as well as the number of spatially and
geological observations from voluminous rhyolites. temporally associated crystal-rich intermediate magmas
Another group of two-phase flow mechanisms, which and cogenetic rhyolites (e.g. Hildreth & Fierstein, 2000;
require a higher crystal fraction (50 vol. % of crystals in Lindsay et al., 2001), has led a number of workers to
the magma), involves the upward percolation of buoyant propose the following hypothesis for the generation of
interstitial melt from a mush, as a result of either the crystal-poor rhyolites (e.g. Bacon & Druitt, 1988; Sisson
exsolution of a gas phase (gas-driven filter-pressing, & Bacon, 1999; Hildreth & Fierstein, 2000): could silicic
Anderson et al., 1984; Sisson & Bacon, 1999), or compac- mushes in the upper crust, detected seismically under
tion phenomena (e.g. McKenzie, 1984; Shirley, 1986; active calderas and preserved as batholiths or crystal-
Philpotts et al., 1996; Rabinowicz et al., 2001; Jackson rich ignimbrites, be the sources for rhyolites? In this
1566
BACHMANN AND BERGANTZ ORIGIN OF CRYSTAL-POOR RHYOLITES
study, we employ field, chemical and physical constraints origin (Bergantz & Dawes, 1994; Evans & Hanson,
to quantitatively assess whether rhyolites can be pro- 1997; Martin et al., 1997; Castro et al., 1999; Pati~ no
duced by this process of melt expulsion from homo- Douce, 1999; Barboza & Bergantz, 2000; Petford et al.,
geneous silicic mushes once crystals start forming a 2000); however, an exhaustive discussion of this conun-
connected framework ( 45–50 vol. % crystal; Fig. 2). drum is beyond the scope of this study. We begin with the
Around this rheological transition from liquid to solid, the fact that they exist and focus on the documentation of
main mechanism for crystal–melt segregation changes their magmatic youth provided by the giant, unzoned,
from crystal settling to melt extraction by compaction. crystal-rich, dacitic ignimbrites, referred to as mono-
Thus, to cover a range of possible time scales, we used tonous intermediates (Hildreth, 1981). One of the critical
hindered settling velocities and compaction rate to show observations made on these ‘erupted batholiths’ and
that differential motion between crystals and melt in other highly crystalline units of intermediate composition
static mushes close to their rheological transition can is that at the observed crystallinity (40–45% crystals),
occur rapidly enough to form the most voluminous rhyo- their interstitial liquid phase is chemically very similar
lites (segregation rates of 101–103 km3/yr). We also to crystal-poor silicic magmas (Cashman & Blundy, 2000;
suggest that gas-driven filter pressing related to exsolution Bachmann et al., 2002; Schmitt et al., 2003), making these
and/or upward percolation of a low-density phase in crystal mushes a natural source and storage site for high-
these upper-crustal mushes (Sisson & Bacon, 1999; SiO2 melts in the upper crust. Considering the volume of
Bachmann & Bergantz, 2003) may enhance crystal– these units (103–104 km3), extraction of only 10% of
melt segregation but is not the dominant mechanism of this interstitial melt would be enough to account for the
melt expulsion from mushes. largest crystal-poor rhyolites (102–103 km3). But is there a
process capable of separating melt from crystals in these
silicic mushes at a fast enough rate to form a voluminous
THE MAKING OF RHYOLITES rhyolite cap in geologically reasonable time scales?
Silicic crystal mushes: rhyolite nurseries
Kilometer-sized granodioritic batholiths are found in the Two-phase flow at high crystal fraction
roots of all mature continental arcs. They are so common A key assumption of this melt expulsion model is that
that the average composition of the upper crust is essen- enough crystallization occurs to produce rhyolitic inter-
tially granodioritic (Taylor & McClennan, 1985). Despite stitial melt and that system-wide convection is hampered
this ubiquity, much controversy still surrounds their (Fig. 2b). Importantly, both of these criteria converge at
1567
JOURNAL OF PETROLOGY VOLUME 45 NUMBER 8 AUGUST 2004
crystal fractions >04. First, as observed in numerous Table 1: Units and variables
systems (e.g. Druitt & Bacon, 1989; Bachmann et al.,
2002; Schmitt et al., 2003), magmas with intermediate Symbol Description Value/units
whole-rock composition and >35–45 vol. % crystals con-
tain interstitial rhyolitic melt. Second, convection ceases
kf Permeability m2
once a rigid crystal framework is attained, which is
f Porosity 0.4—0.5
thought to occur at around 50–55% crystals (Rigid
r Radius m
Percolation Threshold, RPT; Vigneresse et al., 1996).
K Constant in permeability relationship 50—200
The onset of yield strength in crystal–melt suspensions,
m Dynamic viscosity of melt Pa s
thereby retarding convection, may even occur at lower
crystal fraction (for plagioclase laths, the critical crystal- rs Density of solid phase kg/m3
linity fc is <03; Philpotts et al., 1998; Saar et al., 2001), rl Density of liquid phase kg/m3
even though the homogeneity in whole-rock composition Uhs Hindered settling velocity m/s
of monotonous intermediates (crystal contents up to UStokes Stokes velocity m/s
1568
BACHMANN AND BERGANTZ ORIGIN OF CRYSTAL-POOR RHYOLITES
f(c)
as granodioritic plutons, commonly contain megacrysts 0.4
and glomerocrysts, up to several centimeters in dia-
meter (Bateman & Chappell, 1979; Lipman et al., 1997; 0.2
Seaman, 2000; Bachmann et al., 2002), suggesting that
the effective grain size of the mushes may be larger
(3–5 mm). Using r ¼ 3 mm, the permeability is of the 0.0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
order of 108 m2, similar to recent estimates for high-
104
with
ð1 cÞ2
f ðcÞ ¼ ð3Þ
ð1 þ c1=3 Þ½5c=3ð1 cÞ 103
where Uhs, the hindered settling velocity, equals UStokes,
the Stokes settling velocity,
2r 2 gDr 102
UStokes ¼ ð4Þ 0.001 0.002 0.003 0.004 0.005
9m Grain size (m)
where r is radius of the solid particle, g is the acceleration Fig. 4. Time scales required for crystals with a density of 2600 kg/m3
from gravity, Dr is the density contrast between particle (feldspar, quartz) and variable radii to sink 500 m in a silicic mush
and melt, and m is the dynamic viscosity of the melt, (crystal volume fraction is 05) with a melt density of 2300 kg/m3, and
melt viscosities of 1045, 105, and 1055 Pa s.
corrected by the factor f (c), c being the crystal fraction
(Barnea & Mizrahi, 1973). As seen in Fig. 3, Uhs drops
sharply as crystallinity increases, and tends to zero considered to be of the order of 1000 km2, and sinking
at crystallinities >06. Using settling rates based on distances of 05 km are sufficient to produce 500 km3 of
monodispersed particles will give an outer bound; rhyolite. As mush piles are several kilometers thick, we
polydispersed suspensions are more realistic for magmas, can assume that the porosity will not change significantly
but their settling behavior is poorly known (Davis & in the upper 05 km, and accept the time scale required
Acrivos, 1985). for a crystal to sink for 05 km as a reasonable estimate
The hindered settling velocities are calculated assum- (Fig. 4).
ing constant porosity. However, in silicic mush piles, the Grain size and viscosity will be the most important
porosity will tend to decrease downward as crystals settle. parameters for a given porosity. Taking the upper range
Thus, a question is whether this downward decreas- of melt viscosities as published by Scaillet et al. (1998;
ing porosity will greatly reduce settling rates. As a first 1045–1055 Pa s), Fig. 4 shows that, for crystallinities of
approximation, one can estimate the maximum vertical 50%, crystals with a density of 2600 kg/m3 (approx-
distance that crystals must sink from the top of the pile to imately feldspar, quartz) and radii of 2–3 mm will take
account for the volumes of the largest erupted rhyolites 103–104 years to sink 500 m in a rhyolite melt. Increasing
( 102–103 km3). On the basis of caldera surface, hori- the crystallinity to 60% increases the time scales by a
zontal extensions of large magma chambers are usually factor of 27.
1569
JOURNAL OF PETROLOGY VOLUME 45 NUMBER 8 AUGUST 2004
Compaction
Assuming that crystals form a deformable, purely viscous 104.5 Pa s
network (no account of viscoelastic behavior of the 106 105 Pa s
matrix, De 1; Vasilyev et al., 1998), time scales neces- 105.5 Pa s
Time (years)
sary for the separation of crystals and melt in compacting
magmatic systems can be calculated (McKenzie, 1984,
1985; Shirley, 1986). The compaction length, dc, defined 105
as the distance over which compaction occurs above an
impermeable boundary, is
1=2
z þ 43 h 104
dc ¼ kf ð5Þ
m
0.001 0.002 0.003 0.004 0.005
where z and h are the effective bulk and shear viscosities Grain size (m)
defined as follows:
Range of timescales necessary
th h dc to form a > 500 km3 rhyolite
¼ : ð7Þ 104
t0 dc h cap above a silicic mush
If h dc, then
t0 h h
th ¼ ¼ ð8Þ 103
dc w0 ð1 fÞ
1570
BACHMANN AND BERGANTZ ORIGIN OF CRYSTAL-POOR RHYOLITES
to 5 10 km3/yr (104–105 years for rhyolite caps of Such extended residence times would allow enough
500 km3); this range agrees with the slightly slower crystal–melt segregation to form the most voluminous
magma output rates in silicic systems estimated by Trial rhyolites, even by slow compaction alone (except for
& Spera (1990; 102–104 km3/yr). These magma out- very small grain sizes), if magmas are preferentially stored
put estimates of Trial & Spera (1990), calculated using as crystal mushes. As the model we are proposing
erupted volumes divided by the repose time in many requires that high-porosity mushes exist for 104–105
volcanic systems, must include every step in the genera- years, the evolution of magma crystallinity with time
tion of evolved magmas, and, thus, provide minimum must be determined. Several arguments, listed below,
values. suggest that crystal mush is the long-lived magma storage
These segregation rates do not take into account the state in the crust and provide some justification as to why
effect of shear-induced expansivity of porous media and we did not perform conductive cooling models to assess
the possible melt expulsion (or ‘seepage’) related to it the importance of cooling on segregation rates.
(Petford & Koenders, 2003). We consider that the mush (1) There is little evidence for the presence of large
as a whole is immobile with respect to its wall rocks, and bodies of crystal-poor intermediate to silicic magmas
1571
JOURNAL OF PETROLOGY VOLUME 45 NUMBER 8 AUGUST 2004
mushes. Occurrences of comagmatic crystal-poor and volumes of individual units suggest the absence of a
crystal-rich magmas are common in the volcanic record. large, continuous, crystal-poor body over the entire erupt-
Young examples in active volcanic arcs include: (1) the ive period. Rather, these ‘leaks’ tapped localized horizons
Atana–Toconao system in the Altiplano–Puna Ridge (or cupolas) or rhyolitic melt that formed by periodic
Complex, Central Andes, where the rhyolitic Toconao segregation events from an underlying, voluminous and
ignimbrite has been interpreted as the evolved cap of the homogeneous mush. In many such cases, however, the
crystal-rich dacitic Atana ignimbrite (Lindsay et al., 2001; crystal-rich domain presumed below rarely or never is
Schmitt et al., 2003); (2) the interbedding of crystal-rich tapped. This observation, in concert with the fact that
dacites and rhyolites in the Toba Tuffs, Sumatra these rhyolites are commonly small and have very high
(Chesner, 1998); (3) the Whakamaru group ignimbrites, Rb/Sr ratios, may point out areas where the mush has
in the Taupo Volcanic Zone, New Zealand, where a reached a fairly high crystallinity ( 70–80 vol. % crys-
stack of cogenetic, low- and high-SiO2 rhyolitic ignim- tals, on the basis of Rayleigh fractional crystallization
brites show large variations in crystallinity (Brown et al., with typical partition coefficients for Rb and Sr in these
1998); (4) two canonical examples of compositional gaps systems; Fig. 1b), reducing the segregation rate and
1572
BACHMANN AND BERGANTZ ORIGIN OF CRYSTAL-POOR RHYOLITES
0.7
(a) Deer Mountain rhyolite (b) Whakamaru Group ignimbrite
0.4
Long Valley Caldera, USA Taupo volcanic zone, NZ
Age (Ma) Reid et al., 1997 0.6 Brown & Fletcher, 1999
0.3
0.5
0.2 0.4
1σ error bar
1σ error bar 0.3
Eruption age = 0.32±0.02 Ma
0.1 Eruption age = 0.115±0.003 Ma 40Ar/39Ar; Houghton et al., 1995
K/Ar; Mankinen et al., 1986 0.2
0 5 10 15 0 10 20 30
Analysis number
( 10 vol. %) of interstitial liquid is necessary to form the near-liquidus magmas has often been invoked, following
largest erupted rhyolites. Limited melt loss from a mush is the influential works of Grout (1918), Bowen (1928) and
capable of reconciling cumulus textures and melt escape Wager & Brown (1967). However, the recognition that
features with these in situ crystallization trends in plutons sizeable volumes of near-liquidus magma stored in a
(see also Langmuir, 1989, fig. 3). colder environment are destined to convect at low crystal
content led to the realization that settling velocities using
Stokes’ Law [equation (4)] were too fast. Settling velo-
Extended range of crystal ages cities in convecting systems were, therefore, assessed. In
One of the most appealing aspects of melt expulsion from particular, Martin & Nokes (1988) proposed that a char-
a long-lived mush is the fact that it predicts the extended acteristic time scale for crystal settling in convecting
age range recorded by zircons in silicic magmas. Long magmas is
crystallization intervals (>200 kyr) are implied by several
studies (e.g. Reid et al., 1997; Brown & Fletcher, 1999; 9H m
Fig. 7), which show large mean standard weighted devia- tsettle ¼ ð10Þ
2gDrr 2
tions (MSWD) in the datasets, indicating a scatter in
excess of analytical precision. Melt expulsion from the where tsettle is the settling time scale, H is the chamber
mush is likely to entrain some phenocrysts, in particular height, m is the magma viscosity, g is the acceleration due
small ones such as zircon, as settling velocities of zircon to gravity, Dr is the density contrast between crystals and
crystals with radii of 10–50 mm in a rhyolitic melt (m ¼ magma, and r is the radius of crystals. For m ¼ 105 Pa s,
105 Pa s) are of the order of 1010–1012 m/s, slightly Dr ¼ 500 kg/m3 and r ¼ 103 m, tsettle ¼ 6600 years,
slower than the melt expulsion velocities in a compacting indicating that some settling should occur over time
mush calculated here ( 109–1010 m/s). Therefore, scales >104 years. However, the experiments of Martin
the crystal-poor cap inherits crystals that have started & Nokes (1988) were performed at particle concentra-
growing in the mush, prior to extraction. This model tions of <001 wt %, which are unrealistically low for
also predicts that smaller crystals should record ages magmas. Koyaguchi et al. (1990) have looked at settling
close to the eruption age, whereas the larger ones should in slightly more dense suspensions (03 wt % particle;
have a longer crystallization interval. Such grain size 01 vol. %), which is still low compared with magmas,
dependent zircon ages were observed by Brown & as volcanic rocks have 1–50 vol. % crystals, with an
Fletcher (1999) in the Whakamaru group ignimbrites. average of 25%; Ewart, 1982), but noticed a signifi-
cant change in the behavior of the system. Although
settling removes a fraction of the particles, cycles of
Alternative mechanisms of two-phase flow overturn occur, remixing most of the layers periodically
in magmas and increasing clarification time. Settling rates in
Several models of two-phase flow leading to differentia- convecting fluids at geologically realistic crystal frac-
tion of magmas have been proposed. Crystal settling from tions (>2–5 vol. % particles) are poorly known, but
1573
JOURNAL OF PETROLOGY VOLUME 45 NUMBER 8 AUGUST 2004
it appears likely that equation (10) still provides an difficult to explain the presence of old, and complexly
overestimation. zoned phenocrysts in silicic magmas, which requires that
Another argument in support of low settling rates in early crystallized material does not become isolated in a
convecting magmas involves the Stokes number of the boundary layer, but remains in the dominant volume or
crystals. The Stokes number (ST) relates the response is periodically recycled.
time scale of the crystal to the response time scale of the (2) Crystalline boundary layers (or solidification fronts)
fluid (e.g. Burgisser & Bergantz, 2002) and allows an rarely appear to be preserved in the rock record. Con-
assessment of the behavior of a particle in a fluid. At centric zoning, recognized in a large number of plutons
ST 1, particle motion is not dictated by the motion of (Bateman & Chappell, 1979; Halliday et al., 1984;
the surrounding liquid, and will follow its own trajectory Stephens, 1992, 2001), has been used as evidence for
(settling in this case). In contrast, at ST 1, a particle will sidewall crystallization (e.g. Sawka et al., 1990; Verplanck
closely follow the fluid. As the ST is vanishingly small for et al., 1999). However, at least two observations disagree
nearly any conditions in silicic magmatic systems, as a with the interpretation that concentric zoning results
result of the small size of the crystals and the high viscos- from inward crystallization of a large magma body.
1574
BACHMANN AND BERGANTZ ORIGIN OF CRYSTAL-POOR RHYOLITES
advocated by Philpotts and co-workers for large mafic only account for dripping crystal–melt instabilities,
bodies (Philpotts et al., 1996; Meurer & Boudreau, 1996; producing the trace element and crystallinity gradients
Boudreau & Philpotts, 2002), although we suggest that recorded in high-SiO2 rhyolites, but also shed nearly
compaction is not the only two-phase flow mechanism solidified lumps as it becomes locally unstable, providing
occurring in silicic mushes. As the rheological transition an explanation to the presence of cogenetic xenoliths in
from melt to solid is a complex process, which varies as some ash-flow tuffs (e.g. Bacon, 1992; Lowenstern et al.,
a function of system properties, strain rate, and shapes of 1997).
particles (Lejeune & Richet, 1995; Vigneresse et al., 1996; Another mechanism potentially able to induce differ-
Barboza & Bergantz, 1998; Petford, 2003), it is likely that ential motion between crystals and melts in porous media
a large crystal mush encompasses regions slightly above is gas-driven filter pressing, as first proposed by Anderson
the RPT, and regions just below. Therefore, we propose et al. (1984) and expanded by Sisson & Bacon (1999). As
that segregation occurs as a combination of multiple many silicic mushes have water-rich interstitial melts
processes (hindered settling, micro-settling, compaction), (4–6 wt % H2O on average, but up to 8 wt %), and are
once convective currents become negligible. As discussed commonly stored in the upper crust (5–15 km), a signific-
by Philpotts et al. (1996), the crystal-poor layer forming ant fraction of them will reach volatile saturation and
above the mushy region is probably overlain by a crystal- exsolve a low-density, low-viscosity fluid phase (e.g.
line boundary layer at the roof. This hanging crystalline Wallace et al., 1995; Wallace, 2001). Sisson & Bacon
ceiling, denser than the underlying magma, could not (1999) suggested that overpressure following exsolution
1575
JOURNAL OF PETROLOGY VOLUME 45 NUMBER 8 AUGUST 2004
of this gas phase could drive some interstitial melt out of by dripping of solid–liquid mixtures as crystallization
the porous matrix. This process might participate in proceeds at the roof (Bergantz & Ni, 1999). Because of
segregating melt from crystalline residue in the water- the high viscosity of the magma and the conduction-
rich and shallow systems, but some of our recent work controlled heat loss, convective currents are likely to
(Bachmann & Bergantz, 2003) suggests that it may be remain slow, leading to a density stratification, such as
rather inefficient in expelling large quantities of rhyolite roofward decrease in crystals and increase in volatile
melt, and it could certainly not explain the generation of content (e.g. Jellinek et al., 1999). The trace element
rhyolite in relatively dry settings, such as those in the gradient recorded in the interstitial melt and melt inclu-
Yellowstone system. We therefore propose that it facil- sions (early erupted material has a stronger crystal frac-
itates segregation in certain situations, but is not the tionation signature; Hildreth, 1979; Knesel & Davidson,
dominant mechanism. 1997; Anderson et al., 2000) may be explained by slow
‘crystal settling’ induced by dripping solid–liquid
mixtures from the roof (Bergantz & Ni, 1999), depleting
Compositional gradients in magma the upper and early extracted rhyolite of its compatible
1576
BACHMANN AND BERGANTZ ORIGIN OF CRYSTAL-POOR RHYOLITES
stronger crystal fractionation signature displayed by Barboza, S. A. & Bergantz, G. W. (2000). Metamorphism and anatexis
silicic volcanic units (Fig. 1b), the models proposed for in the mafic complex contact aureole, Ivrea Zone, Northern Italy.
Journal of Petrology 41, 1307–1327.
the generation of granitoids have been inclined towards
Barnea, E. & Mizrahi, J. (1973). A generalized approach to the fluid
crustal melting, whereas crystal fractionation was favored dynamics of particulate systems. Part I. General correlation for
for volcanic rocks. Our model of melt expulsion from fluidization and sedimentation in solid multiparticulate systems.
crystal mushes reconciles these views by coupling their Chemical Engineering Journal 5, 171–189.
evolution until shallow storage, where melt extraction Barnes, C. G., Burton, B. R., Burling, T. C., Wright, J. E. &
from crystal mush occurs, imparting a more pronounced Karlsson, H. R. (2001). Petrology and geochemistry of the Late
crystal fractionation signature in the erupted portions of Eocene Harrison Pass Pluton, Ruby Mountains core complex,
northeastern Nevada. Journal of Petrology 42, 901–929.
the systems.
Barth, S., Oberli, F., Meier, M., Blattner, P., Bargossi, G. M. & Di
Battistini, G. (1993). The evolution of a calc-alkaline basic to silicic
magma system; geochemical and Rb–Sr, Sm–Nd, and 18O/16O
ACKNOWLEDGEMENTS isotopic evidence from the late Hercynian Atesina–Cima d’Asta
We thank Wes Hildreth, Nick Petford and Mark Jellinek volcano-plutonic complex, northern Italy. Geochimica et Cosmochimica
1577
JOURNAL OF PETROLOGY VOLUME 45 NUMBER 8 AUGUST 2004
Cashman, K. & Blundy, J. (2000). Degassing and crystallization of Druitt, T. H. & Bacon, C. R. (1989). Petrology of the zoned
ascending andesite and dacite. Philosophical Transactions of the Royal calcalkaline magma chamber of Mount Mazama, Crater Lake,
Society of London 358, 1487–1513. Oregon. Contributions to Mineralogy and Petrology 101, 245–259.
Castro, A., Pati~no Douce, A. E., Corretge, L. G., de la Rosa, J. D., El- Duffield, W. A. & Dalrymple, G. B. (1990). The Taylor Creek Rhyolite
Biad, M. & El-Hmidi, H. (1999). Origin of peraluminous granites of New Mexico; a rapidly emplaced field of lava domes and flows.
and granodiorites, Iberian massif, Spain: an experimental test Bulletin of Volcanology 52, 475–487.
of granite petrogenesis. Contributions to Mineralogy and Petrology 135, Dullien, F. A. L. (1979). Porous Media Fluid Transport and Pore Structure.
255–276. New York: Academic Press.
Chen, C. F. & Turner, J. S. (1980). Crystallization in double-diffusive Eichelberger, J. C., Carrigan, C. R., Westrich, H. R. & Price, R. H.
system. Journal of Geophysical Research 85, 2573–2593. (1986). Non-explosive silicic volcanism. Nature 323, 598–602.
Chesner, C. A. (1998). Petrogenesis of the Toba Tuffs, Sumatra, Evans, O. C. & Hanson, G. N. (1997). Late- to post-kinematic Archean
Indonesia. Journal of Petrology 39, 397–438. granitoids of the S.W. Superior Province: derivation through direct
Clemens, J. D. (2003). S-type granite—petrogenetic issues, models and mantle melting. In: De Wit, M. J. & Ahwal, L. D. (eds) Greenstone
evidence. Earth-Science Reviews 61, 1–18. Belts. Oxford: Clarendon Press, pp. 280–295.
Clemens, J. D. & Wall, V. J. (1984). Origin and evolution of a Ewart, A. (1982). The mineralogy and petrology of Tertiary–Recent
peraluminous ignimbrite suite: the Violet Town Volcanics. Contribu- orogenic volcanic rocks: with special reference to the andesitic–
1578
BACHMANN AND BERGANTZ ORIGIN OF CRYSTAL-POOR RHYOLITES
Heumann, A., Davies, G. R. & Elliott, T. (2002). Crystallization history Koyaguchi, T. & Kaneko, K. (1999). A two-stage thermal evolution
of rhyolites at Long Valley, California, inferred from combined model of magmas in continental crust. Journal of Petrology 40,
U-series and Rb–Sr isotope systematics. Geochimica et Cosmochimica 241–254.
Acta 66, 1821–1837. Koyagushi, T. & Kaneko, K. (2000). Thermal evolution of silicic
Hildreth, W. (1979). The Bishop Tuff: evidence for the origin of the magma chambers after basalt replenishment. Transactions of the Royal
compositional zonation in silicic magma chambers. Geological Society of Society of Edinburgh 91, 47–60.
America, Special Paper 180, 43–76. Koyaguchi, T., Hallworth, M. A., Huppert, H. E. & Sparks, R. S. J.
Hildreth, W. (1981). Gradients in silicic magma chambers: implica- (1990). Sedimentation of particles from a convecting fluid. Nature
tions for lithospheric magmatism. Journal of Geophysical Research 86, 343, 447–450.
10153–10192. Langmuir, C. H. (1989). Geochemical consequences of in situ
Hildreth, W. & Fierstein, J. (2000). Katmai volcanic cluster and the crystallization. Nature 340, 199–205.
great eruption of 1912. Geological Society of America Bulletin 112, Lejeune, A.-M. & Richet, P. (1995). Rheology of crystal-bearing silicate
1594–1620. melts; an experimental study at high viscosities. Journal of Geophysical
Hildreth, W., Halliday, A. N. & Christiansen, R. L. (1991). Isotopic Research 100, 4215–4229.
and chemical evidence concerning the genesis and contamination of Lindsay, J. M., Schmitt, A. K., Trumbull, R. B., De Silva, S. L., Siebel,
basaltic and rhyolitic magma beneath the Yellowstone Plateau W. & Emmermann, R. (2001). Magmatic evolution of the La Pacana
1579
JOURNAL OF PETROLOGY VOLUME 45 NUMBER 8 AUGUST 2004
Martin, D., Griffiths, R. W. & Campbell, I. H. (1987). Compositional the Ardara pluton, Ireland. Journal of Structural Geology 112,
and thermal convection in magma chambers. Contributions to 1543–1558.
Mineralogy and Petrology 96, 465–475. Munksgaard, N. C. (1984). High d 18O and possibly pre-eruptional
Martin, H., Peucat, J. J., Sabate, P. & Cunha, J. C. (1997). Crustal Rb–Sr isochrons in cordierite-bearing Neogene volcanics from SE
evolution in the early Archean of South America: example of the Spain. Contributions to Mineralogy and Petrology 87, 351–358.
Sete Voltas Massif, Bahia State, Brazil. Precambrian Research 82, Murphy, M. D., Sparks, R. S. J., Barclay, J., Carroll, M. R. & Brewer, T. S.
35–62. (2000). Remobilization of andesitic magma by intrusion of mafic
Masturyono, McCaffrey, R., Wark, D. A. & Roecker, S. W. (2001). magma at the Soufriere Hills Volcano, Montserrat, West Indies.
Distribution of magma beneath the Toba caldera complex, north Journal of Petrology 41, 21–42.
Sumatra, Indonesia, constrained by three-dimensional P wave Niemi, A. N. & Courtney, T. H. (1983). Settling in solid–liquid systems
velocities, seismicity, and gravity data. Geochemistry, Geophysics, with specific application to liquid phase sintering. Acta Metallurgica 31,
Geosystems 2, paper number 2000GC000096. 1393–1401.
Matthews, S. J., Sparks, R. S. J. & Gardeweg, M. C. (1999). The Nozama, T. (1983). Felsic plutonism in Japan. Geological Society of
Piedras Grandes–Soncor eruptions, Lascar volcano, Chile; evolution America, Memoirs 159, 105–122.
of a zoned magma chamber in the central Andean upper crust. Oldenburg, C. M., Spera, F. J., Yuen, D. A. & Sewell, G. (1989).
Journal of Petrology 40, 1891–1919. Dynamic mixing in magma bodies: theory, simulations, and
1580
BACHMANN AND BERGANTZ ORIGIN OF CRYSTAL-POOR RHYOLITES
Rice, A. (1981). Convective fractionation: a mechanism to provide KAlSi2O6–CaMgSi2O6: implications for compositionally zoned
cryptic zoning (macrosegregation), layering crescumulates, banded magma bodies. American Mineralogist 40, 1188–1207.
tuffs and explosive volcanism in igneous processes. Journal of Steck, L. K., Thurber, C. H., Fehler, M. C., et al. (1998). Crustal
Geophysical Research 86, 405–417. and mantle P wave velocity structure beneath Valles caldera,
Riley, T. R., Leat, P. T., Pankhurst, R. J. & Harris, C. (2001). Origins New Mexico: results from the Jemez teleseismic tomography
of large volume rhyolitic volcanism in the Antarctic Peninsula and experiment. Journal of Geophysical Research 103, 24301–24320.
Patagonia by crustal melting. Journal of Petrology 42, 1043–1065. Stephens, W. E. (1992). Spatial, compositional and rheological
Robinson, D. M. & Miller, C. F. (1999). Record of magma chamber constraints on the origin of zoning in the Criffel pluton, Scotland.
processes preserved in accessory mineral assemblages, Aztec Wash Transactions of the Royal Society of Edinburgh 83, 191–199.
pluton, Nevada. American Mineralogist 84, 1346–1353. Stephens, W. E. (2001). Polycrystalline amphibole aggregates (clots) in
Saar, M. O., Manga, M., Cashman, K. V. & Fremouw, S. (2001). granites as potential I-type restite: an ion microprobe study of rare-
Numerical models of the onset of yield strength in crystal–melt earth distributions. Australian Journal of Earth Sciences 48, 591–601.
suspensions. Earth and Planetary Science Letters 187, 367–379. Steven, T. A. & Lipman, P. W. (1976). Calderas of the San Juan
Sawka, W. N., Chappell, B. W. & Kistler, R. W. (1990). Granitoid volcanic field, southwestern Colorado. US Geological Survey, Professional
compositional zoning by side-wall boundary layer differentiation: Paper 958, 1–35.
evidence from the Palisade Crest Intrusive Suite, central Sierra Taylor, S. R. & McClennan, S. M. (1985). The Continental Crust: its
1581
JOURNAL OF PETROLOGY VOLUME 45 NUMBER 8 AUGUST 2004
the Pyramid Peak granite, California. Geological Society of America Wolff, J. A., Balsley, S. D. & Gregory, R. T. (2002). Oxygen isotope
Bulletin 114, 909–920. disequilibrium between quartz and sanidine from the Bandelier Tuff,
Wilson, C. J. N. & Hildreth, W. (1997). The Bishop Tuff: new insights New Mexico, consistent with a short residence time of phenocryst in
from eruptive stratigraphy. Journal of Geology 105, 407–439. rhyolitic magma. Journal of Volcanology and Geothermal Research 116,
Wolff, J. A., Ramos, F. C. & Davidson, J. P. (1999). Sr isotope 119–135.
disequilibrium during differentiation of the Bandelier Tuff: con- Zandt, G., Leidig, M., Chmielowski, J., Baumont, D. & Yuan, X.
straints on the crystallization of a large rhyolitic magma chamber. (2003). Seismic detection and characterization of the Altiplano–Puna
Geology 27, 495–498. magma body, central Andes. Pure and Applied Geophysics 160, 789–807.
1582