Professional Documents
Culture Documents
Coek - Info - RBSR Ages and ND Isotopic Compositions of Melt Inc
Coek - Info - RBSR Ages and ND Isotopic Compositions of Melt Inc
Abstract
Rb-Sr ages of melt inclusions in quartz have the potential to provide a unique and illuminating record of the
differentiation and accumulation history of silicic magmas. Here we report the first Sr and Nd isotopic measurements of melt
inclusion bearing quartz (MIBQ), extracted from the Bishop Tuff (BT), representing what is arguably the world’s most
controversial yet classic zoned silicic magmatic system. Early erupted BT airfall pumice, representing material from the top
of the magma chamber, yields individual quartz crystals with Rb/Sr > 70 and variable apparent differentiation ages of
1,420 f 80 ka, 2,100 f 100 ka, 2,150 f 150 ka and 2,500) 200 ka. A bulk quartz separate has an apparent age of
1,900 f 300 ka, while a quartz separate from an individual airfall pumice has an apparent age of 1,330 f 80 ka. These ages
are similar to the Rb-Sr differentiation ages (2,047-1,894 ka) and k-Ar eruption ages (2,100-1,300 ka) of the adjacent
early pre-caldera rhyolites at Glass Mountain (GM), indicating that the upper portions of the BT magma chamber included
components that episodically differentiated over the same time interval as early Glass Mountain volcanism. However, at this
stage the magmatic systems were separate, as indicated by lNd of - 1 for airfall melt inclusions, similar to the BT and late
GM rhyolites, compared to lNd of - 3 for early GM rhyolites. In contrast to BT airfall, melt inclusions from intermediate
and later erupted BT ignimbrite pumice, from deeper levels of the magma chamber, give younger apparent differentiation
ages: 1,040 f 140 ka (late erupted), 1,180 +_80 ka (intermediate erupted clast l), and 1,060 * 110 ka, 1080 f 110 ka
(intermediate erupted clast 2, two quartz size fractions). These ages are indistinguishable from the Rb-Sr differentiation age
of the late pre-caldera GM rhyolites, 1,140 f 80 ka. The eNd of intermediate and late erupted melt inclusions is - 1, the
same as BT airfall melt inclusions.
These data are consistent with an early episodic development of the evolved upper portions of the BT magma chamber.
However, we find no evidence that most of the differentiation took place at around 2 Ma, as recently suggested on the basis
of 40Ar-39Ar dating. Most of the Bishop Tuff magma, particularly the later erupted portions, differentiated in major events at
around 1.0-1.2 Ma. Early differentiated melts repeatedly accumulated in the upper portions of the chamber concomitantly
with differentiation, accumulation and eruption in the nearby Glass Mountain magma reservoir, which possibly represented a
separate cupola to a larger complex system. By _ 1.2 Ma the Bishop Tuff magmatic system expanded to include the Glass
Mountain region. We find no evidence for differentiation younger than _ 1.0 Ma in any samples from the Bishop Tuff (or
Glass Mountain), confirming the view that over the > 300,000 years that elapsed before eruption of the Bishop Tuff, the
stratified magmatic system remained largely stable.
Tableland
Volcanism in the immediate Long Valley region l-
BT-38
extends back to 4 Ma [20],[21]. Between 4.0 and 3.5 Adobe
Ma, basaltic flows were erupted which are composi- r I$mic
tionally and isotopically very similar to much younger Mono
(60 ka) flows erupted within the post-Bishop Tuff
caldera. The early basaltic activity was followed by c
Early Late
the building of dacite/quartz latite domes between
Fig. 1. Schematic representation of the eruption sequence of the
3.0 and 2.7 Ma. The earliest volcanism which may
Bishop Tuff based on apparent Fe-Ti oxide temperature. figure
be directly associated with the Bishop Tuff produced modified from [2] to show sample positions in the eruptive
the rhyolites of Glass Mountain [ 12,131. These rhyo- sequence. The position of sample BT-20 within the Adohe Lobe
lites were erupted between 2.1 and 0.8 Ma and can stratigraphy is uncertain. Dashed line for airfall represents the
be divided, on the basis of chemical and isotopic results of recent stratigraphic observations, indicating that eruption
of Plinian airfall was synchronous and continuous with interca-
composition and time of eruption, into early (2.1- 1.3
lated early ignimbrite flows [23].
Ma) and late (1.2-0.8 Ma) suites. The two suites can
also be distinguished on the basis of their age of
differentiation, which in some cases was 2 lo5 yr vidual pumice clasts and from multiple-pumice sam-
before eruption [5,7]. The Bishop Tuff consists of ples. For the airfall sample (LV/g/l), quartz sepa-
1750 km3 [22] of high silica (75 wt%) rhyolite rates from multiple pumice clasts as well from a
erupted to form a series of airfall tuffs and large single clast were analyzed. In addition, four individ-
ignimbrite outflow sheets, leading to caldera collapse ual inclusion-bearing quartz grains (Qtz A, Qtz B,
[2]. The ignimbrite flows can be divided into early, Qtz C and Qtz D) were selected from the multiple-
intermediate and late Bishop Tuff (BT), representing clast quartz separate for individual analysis. These
approximately an inverted stratigraphy of the magma quartz phenocrysts contained large (- 200 Frn di-
chamber [2]. Early ignimbrite flows are underlain ameter) glass inclusions. The quartz separates from
and intercalated with a co-erupted Plinian airfall the intermediate ignimbrite sample (BT-38) are from
pumice sequence [23]. a multiple pumice clast sample, and two discrete
Samples discussed in this paper are from: early pumice clasts (clast 1 and clast 2). The quartz sepa-
erupted BT: airfall, Gorges Lobe, and San Joaquin rate from pumice clast 2 was divided into two size
Lobe (lower cooling unit); intermediate erupted BT: fractions (1.4- 1.7 mm and 0.85- 1.40 mm) for anal-
Tableland Lobe; and late erupted BT: Adobe Lobe ysis. For the late ignimbrite sample (BT-20), a quartz
and Mono Lobe (Fig. 1, locations in Appendix A). separate from a multiple pumice clast sample was
MIBQ was separated from pumice samples repre- analyzed. Individual, large (3-4 mm) crystals of
senting early erupted Plinian airfall (LV/8/ l), inter- sanidine and plagioclase were also analyzed from
mediate erupted ignimbrite (pyroxene free, Tableland samples BT-38 pumice clast 2 and LV/8/ 1 pumice
Lobe) 0.5 m above airfall (BT-38) and from late clast 1.
erupted ignimbrite (pyroxene bearing) of the Adobe Quartz phenocrysts from the samples are bi-
Lobe (BT-20). Quartz was separated from both indi- pyramidal in form and range in size from about 0.5
550 J.N. Christensen, A.N. Halliday/ Earth and Planetary Science Letters I44 (1996) 547-561
initial *‘Sr/ “Sr and the uncertainties in the mea- MIBQ separates from individual intermediate ign-
sured “Sr/ 86Sr and “Rb/*‘Sr (* 1%). All age imbrite pumice lumps (BT-38) give similar results.
uncertainties are stated at the 2a level. The bulk MIBQ separate from pumice clast 1 has a
The bulk MIBQ separate from the multiple pumice model age of 1,180 + 80 ka. Two size fractions of
airfall (LV/8/1) sample yields a model age of MIBQ crystals from pumice clast 2 give identical
1,900 +_300 ka (Table 1). The four individual MIBQ model ages (and similar Rb/Sr) of 1,080 + 110 ka
crystals reserved from this population have model (1.4-1.7 mm) and 1,060 & 110 ka (0.85-1.4 mm).
ages of 2,100 + 100 ka, 2,150 + 150 ka, 2,500 ? 200 On a Rb-Sr isotopic diagram (Fig. 2a). three of
ka and 1,420 + 80 ka, with a weighted average of the MIBQ samples (BT-20, BT-38 pumice clast 2
2,000 + 700 ka, indistinguishable from the apparent MIBQ separates) define a linear array indicating an
age of the bulk MIBQ separate of the multiple age of 1,140 ka. These samples fall along the same
pumice clast sample (Table 1). The variation in line defined by late BT fiamme (Mono Lobe) glass
apparent age of the individual crystals indicates that and feldspar (BT-27 and BT-22, data from [6]).
the bulk quartz separate derived from multiple airfall Together they form an isochron indicating an age of
pumice clasts consisted of a mixed population of 1,100 k 30 ka (MSWD = 0.86) and an initial
crystals. A single airfall pumice clast yielded a bulk “Sr/ 86Sr of 0.70594 _t 5 (Fig. 2a). The MIBQ sam-
MIBQ separate with a model age of 1,330 _t 80 ka, ple from pumice clast 1 of BT-38 falls outside the
supporting the conclusion that, for the airfall, there error envelope of this isochron. Instead, it falls on
may be clast to clast variation in melt inclusion age. the isochron defined by late GM rhyolites which
This age is within error of the model age, 1,420 _t 80 gives an age of 1,140 k 80 ka and a initial *‘Sr/ 86Sr
ka, of one of the individual MIBQ crystals. of 0.7061 + 3 [5].
In contrast to the MIBQ separates from the airfall None of the individual MIBQ crystals from the
sample, samples of intermediate and late erupted airfall sample (LV/8/1) plot near a late GM
ignimbrite yield MIBQ with younger model ages isochron nor near the ignimbrite isochron (Fig. 2b).
(Table 1). Again, model ages are calculated assum- However, two of the four individual MIBQ analyses
ing an initial 87Sr/ “Sr of 0.7060 -+ 1. The bulk (Qtz A 2,100 ka and Qtz C 2.150 ka) fall near the
MIBQ separate from the late ignimbrite pumice sam- isochron for early GM inner lavas (2,047 _t 13 ka).
ple (BT-20) has a model age of 1.040 + 140 ka. Analyses of bulk airfall glass and sanidine (sample
Table I
Rb-Sr isotopic data
Sample Rb L Sr L: R7Rb/ 86Sr s7Sr/shSr Apparent age
(ppm) (PPm) (f2u) (Ma k2a)
LV/8/ 1 MIBQ multi-clast _ 32.62 0.706857 f 17 1.9 f 0.3
LV/8/ 1 qtz A 256.9 0.71354 + 33 2.07 f 0.10
LV/8,‘1 qtz B 214.1 0.71356 f 32 2.5 + 0.2
LV/8/1 qtz C 280.0 0.71453 + 34 2.15 + 0.15
LV/S/ I qtz D 298.3 0.7 1199 + 29 I .42 f 0.08
LV/8/1 I MIBQ single clast _ 211.7 0.709992 & 85 I .33 & 0.08
LV/S/ 1 1 plagioclase A 2.432 22.42 0.3140 0.706398 + 20 1.20+ 0.05
LV/8/ 1 1 plagioclase B 2.402 28.79 0.24 I5 0.706334 +_ 10 1.22 & 0.04
BT-38- I MIBQ clast 1 143.2 0.708405 + 52 1.18 kO.08
BT-38-2A MIBQ d clast 2 134.7 0.708069 f 91 1.08 f 0. I I
BT-38-2B MIBQ e clast 2 129.5 0.707955 + 82 1.06 + 0.1 I
BT-38-2 sanidine clast 2 109.7 24.44 12.99 0.706564 k 20 0.87 + 0.08 ‘, 0.84 & 0.07 b
BT-38-2 plagioclase clast 2 2.926 36.98 0.2290 0.706342 f 11 0.90 f 0.06 d. 0.88 + 0.06 b
BT-20 MIBQ 89.16 0.707318 * 64 1.04&0.14
’ Age calculated using data for BT-38-2A MIBQ. b Age calculated using data for BT-38.2B MIBQ. ’ No absolute concentrations are
reported for MIBQ because the weights of analyzed inclusions are uncertain. d 1.4-1.7 mm size fraction. ’ 0.85-1.4 mm size fraction.
552 J.N. Christensen, A.N. Halliday/Earth and Planetary Science Letters 144 (19961 547-561
0.706 L
From an 1”‘Ar-“9Ar study of MIBQ from BT
airfall and early ignimbrite, Bogaard and Schirnick
[14] concluded that quartz had resided in the BT 0.0 0.5 rrupBo” 1 .o tdiff. I .5 2.0
magma for - 1.1 m.y. before eruption and, there- time, Ma
fore. that the entire BT magma volume had existed Fig. 3. Sr evolution diagrams for individual feldspars and inclu-
for that period. The Rb-Sr isotopic data reported sion glass (MIBQ) from BT-38 clast 2 (a) and LV/8/1 clast 1
here for MIBQ requires some modification of this (b). The symbol t dlff stands for the apparent age of the MIBQ for
those samples assuming an initial “Sr/ XhSr of 0.706, shown by
conclusion. The 4”Ar-39Ar isochron age for MIBQ
the horizontal black lines. The eruption of the Bishop Tuff was at
from airfall is 1,930 5 12 (2 a) ka, while that for 0.760 Ma [14,27.28]. In (a) the difference in apparent age between
MIBQ from early erupted Gorges Lobe ignimbrite is the sanidine and plagioclase crystals is calculated based on the
2,300 + 800 ka [14]. These ages are supported by the uncertainties in the isotopic compositions of the feldspars alone
Rb-Sr age of the airfall bulk MIBQ sample, 1,900 + 171.
30 ka, and by two of the individual MIBQ single-
crystal Rb-Sr ages (Qtz A, 2,100 + 100 ka and Qtz
C, 2,150 f 150 ka) (Fig. 4). Total fusion ‘OAr-‘YAr bulk MIBQ from LV/8/ 1 pumice clast 1 ( 1,330 _+
ages of individual airfall MIBQ crystals range from 80 ka). Thus it is possible that there are components
1,590 k 400 ka to 2,300 _t 340 ka [14] with a in the early erupted BT both older and younger than
weighted average of 1,890 5- 60 (2 a) ka. Although -2Ma.
the J”Ar-‘9Ar data yield an isochron with a MSWD In contrast to the early erupted BT, the Rb-Sr
of 2.2, the range of individual 10Ar-3YAr apparent data for MIBQ from both intermediate and late
ages is sufficiently broad that it encompasses the ignimbrite indicate ages around 1,100 ka (Fig. 4).
Rb-Sr ages of individual MIBQ phenocrysts Qtz B Clearly, not all melt inclusions in the Bishop Tuff
(2,500 + 200 ka), Qtz D (1,420 * 80 ka) and the are N 2,000 ka. Melt inclusions of that age appear to
554 J.N. Christensen, A.N. Halliday/ Earth and Planetaw Science Letters 144 (19%) 547-561
be confined to the early erupted BT. The greater + 13 ka (“Sr/ 86Sr initial of 0.7063 k 2) while a set
portion of the BT, represented by the intermediate further from the rim (outer Iavas) define a 1,894 _t 13
and late ignimbrite pumice samples, is characterized ka isochron (“Sr/ “Sr initial of 0.7062 + 2). These
by MIBQ of N 1,100 ka age (340 kyr residence ages correspond with the Rb-Sr model ages of bulk
time). MIBQ from the airfall (1900 k 300 ka), individual
There is a correspondence between the BT melt MIBQ phenocrysts Qtz A and Qtz C (2,100 & 100 ka
inclusion apparent ages and Rb-Sr isochron ages for and 2,150 + 150 ka).
early and late Glass Mountain (GM) rhyolites. Rhyo- Late GM lavas can also be geographically divided
lites of GM were erupted prior to the Bishop Tuff into inner and outer lavas. They give younger Rb-Sr
between 2,100 and 800 ka and can be divided chemi- isochron ages of 1,090 f 50 ka (“Sr/ 86Sr initial of
cally into an early (eruption ages 2,100-1,300 ka) 0.7057 + 2) for inner lavas (G.R. Davies and A.N.
and late (eruption ages 1,200-800 ka) series of flows Halliday, unpubl. data, [33]) and 1,140 & 80 ka (ini-
[ 12,131. The later GM rhyolites are more differenti- tial “Sr/ *‘Sr of 0.7061 + 3) for outer lavas [5].
ated than the Bishop Tuff but show a chemical trend These ages correspond to the model ages of MIBQ
toward early BT compositions [ 12,131 and have Nd from the ignimbrite pumice samples thought to rep-
and Pb isotopic compositions indistinguishable from resent a large potion of the BT (BT-20 bulk quartz
the BT, suggesting that the BT magma and late GM 1,040 f. 140 ka, BT-38-l bulk quartz 1,180 + 80 ka,
rhyolites were tapped from the same magma body BT-38-2 1,080 f. 110 ka and 1,060 + 110 ka). The
[5]. Rb-Sr isotopic studies have shown [5,7] that GM model ages of LV/8/1 pumice clast 1 bulk MIBQ
lavas erupted at various times fall along distinct (1,330 _t 80 ka) and airfall individual MIBQ Qtz D
isochrons, indicating ages corresponding to the ages (1,420 F 80 ka) do not correspond to ages recorded
of the oldest lavas erupted in the early and late by Glass Mountain rhyolites, and may represent
series, respectively. These authors interpret these pre-late GM magmatic events in the Long Valley
isochrons as representing two major episodes (at ca. system. The temporal relationships between the BT
2 Ma and ca. 1.1 Ma) of differentiation and produc- and the early and late GM rhyolites are summarized
tion of high Rb/Sr rhyolitic magma in the Long in Fig. 4.
Valley system. Early GM lavas close to the present Some early erupted ignimbrite may be directly
caldera rim (inner lavas) define an isochron of 2,047 related to magmas erupted as late GM inner rhyo-
Individualmeltindwion bear@
quark grainsfrom&fall
Bulk meltindusion bearing
quark separatesfmm &fall
BulkmeltInelusionbearing
quark separatesfromignimbrik
“Arl’gAr meltinclusion
bearingquark,
Bogaardand Schirnick, 1995
Fig. 4. A schematic presentation of the Rb-Sr ages of MIBQ. 40Ar-79Ar ages for MIBQ from airfall and the Gorges Lobe ignimbrite from
[4] are represented by grey triangles. Rb-Sr isochron ages (0) for early Glass Mountain and late Glass Mountain rhyolites are from [7] and
[I], respectively. The age of the Bishop Tuff eruption is from [14.27,28].
J.N. Christensen, A.N. Halliday/Earth and Planeran Science Letters 144 (19961547-561 555
resorption episodes in response to variations in ther- glass (BT-24) appears to have been the most affected
mal input to the system. by the proposed magma mixing event, with perhaps
The Rb-Sr isochrons and model ages of MIBQ as much 30% less differentiated pyroxene-bearing
apply only to the production of high Rb/Sr liquid in rhyolite mixed in 16,261. Late and intermediate
the BT magma chamber. Because of their low Rb/Sr, erupted melt inclusions do not appear to record this
it is less certain how long the less evolved (magma) mixing event [34]. Therefore, the majority of melt
portions of the BT existed in the Long Valley magma inclusions in quartz formed prior to the putative
chamber. The feldspar/glass ages for late BT fi- mixing event, and so the inferred MIBQ ages would
amme (relatively low Rb/Sr) range from 1.400 F not be affected. The observation that late erupted
400 ka (sanidine BT-27) to 1,130 f 240 ka (sanidine fiamme glass and feldspar lie on an isochron with
BT-22) and 1,150 + 240 ka (plagioclase BT-22) (data late and intermediate erupted MIBQ suggests that
from [6]). These ages are consistent with the average these samples have Sr isotopic compositions little
age of feldspar growth in late BT fiamme magma, affected by magma mixing.
correlating with the differentiation episode that pro-
duced the high Rb/Sr magmas represented by inter- 4.8. Lack of evidence for assimilation
mediate BT ignimbrite and late Glass Mountain lavas.
The agreement between “Ar-“‘Ar and Rb-Sr
4.7. Anomalies in the late erupted Bishop Tuff techniques for a N 2000 ka BT airfall MIBQ negates
the need for assimilation [26,37] to explain the Sr
Melt inclusions from the late erupted BT have isotopic contrast at eruption between the early BT
higher Rb/Sr than glass from late erupted BT (31 airfall and the rest of the BT. The magma erupted as
(Table 1) vs. 2.2-7.2 [6]), contrary to expectation. early BT airfall is not necessarily contaminated, it
This is consistent with the observation that melt was simply older at 760 ka than later erupted BT
inclusions from the late erupted BT have trace ele- magma. Other considerations indicate that assimila-
ment compositions that are more evolved than the tion was insignificant in the BT magma chamber.
host glass [34]. Other anomalous features of the late Low Sr concentrations in the GM rhyolites and early
erupted BT include: sanidine phenocrysts with Sr- BT suggest assimilation of Sr-rich wall rock could
and Ba-rich rims [ 16,341; chemical disequilibrium not have been important [6,19,22,26]. Consideration
between pyroxenes and Fe-Ti oxides [35]; elevated of the Sr evolution plots for GM rhyolites (Fig. 4a,b)
87Sr/ “Sr of latest erupted (Fig. 1) BT pumiceous also suggests minimal assimilation, a conclusion also
glass (BT-24) [6,26]; and Fe-Ti oxide inclusions in reached by others [5,7] based on the robustness of
quartz that are similar to those in early erupted BT the GM Rb-Sr isochrons. Therefore, it appears that
rather than those in the host glass [36]. These fea- there was no roofward increase in assimilation in the
tures have been attributed to the mixing of two BT magma chamber, as has been suggested by some
different rhyolitic magmas [36,35.6,26]. However, ]371.
the compositional range of late erupted quartz-hosted
melt inclusions is consistent with fractional crystal-
lization, not magma mixing [34]. In addition, from 5. Conclusions
the observation that, for late erupted quartz with
multiple inclusions, higher concentrations of COZ , The Sr isotopic data provide evidence that the
which are correlated with higher compatible element Bishop Tuff (BT) consists of several components
concentration, are found in inclusions closer to the with different differentiation ages. The BT airfall has
crystal rims, Lu et al. [34] inferred that late-erupted melt inclusion bearing quartz (MIBQ) with a range
quartz settled into less differentiated melt during of Rb-Sr ages with an average of N 2000 ka, in
their growth. agreement with the ““Ar-39Ar results of Bogaard and
The proportion of the BT that has been effected Schimick [ 141. MIBQ from intermediate and late
by the above processes is unclear. From its Sr iso- ignimbrite pumice have Rb-Sr ages of N 1,100 ka,
topic composition, the latest erupted BT pumice suggesting that other portions of the BT magma were
J.N. Christensen. A.N. Halliduy/Eurth and Planetarv Science Letters 144 (19961 547-561 559
Long Valley Magmatic History evidence that the magma erupted as intennediate and
A B
late BT ignimbrite had resided for 340 * 30 kyr.
Thus, the results for the Bishop Tuff air-fall sample
greatly extend the possible time period for pre-erup-
tion magma storage (cf. [1411.
The history and geometry for the Long Valley
I ? Magma chamber from IV 2,000 ka to eruption of the
1.1Ma Late BT at 760 ka are summarized in Fig. 8. This history
,,,d&Glass
Mountin
I II provides a framework for considering the chemical
evolution of the Long Valley magmatic system. Pre-
viously, it has been assumed that the observed chem-
ical variation within the Bishop Tuff (and so by
inference the variation that existed vertically within
the magma chamber) was developed by a process
operating over a single period of time. However, the
observation that a portion of the early Bishop Tuff
was differentiated nearly 1 m.y. earlier than later
erupted BT precludes the necessity of assimilation to
explain its Sr isotopic composition and observed
reversals in trace element content. In order to make
sense of chemical and isotopic patterns within a
single eruption or series of eruptions requires a
knowledge and consideration of the time transgres-
sive (or episodic) nature of magma production as
exemplified by the Long Valley system.
Fig. 8. Schematic cross sections along line A-B in Fig. 7,
The isotopic analysis of melt inclusion bearing
showing the magmatic history of the Long Valley magma cham- phenocrysts, as demonstrated by this paper, provides
ber (cf. [5,14]). The last two panels depict the eruption of the BT, a means to see through effects of post-eruption alter-
first from a single vent then from the ring-fracture during caldera ation, magma mixing and crystal settling. Given the
collapse [22]. During the later phase of eruption in response to
low concentration of Sr in the BT melt inclusions,
caldera collapse, hotter, less evolved magma may have been
forced (shown by arrow in the last panel) from deeper levels and
similar techniques should be readily applicable as
erupted [22]. well to other melt inclusion bearing phenocrysts,
such olivine from MORB or OIB, which have far
higher Sr concentrations.
differentiated in the same event that produced the
late Glass Mountain (GM) high Rb/Sr rhyolites.
The eNd of MIBQ from air-fall pumice is - 1, in
contrast to early GM rhyolites which have eNd = - 3. Acknowledgements
This, together with the age of MIBQ from the BT
airfall, suggests the existence at N 2 Ma of two This research was supported by NSF grants EAR
spatially separate and isotopically distinct magma 911238, 9205435, 9304159, 9405503, 9406248 to
bodies: eruption from one produced the early GM ANH. Reviews by C. Bacon, G. Mahood and R.
rhyolites, while the other was preserved for a period Carlson improved the manuscript and are gratefully
of N 1.2 m.y. until it was erupted as the early acknowledged. Thanks are due to Wes Hildreth for
Plinian airfall of the Bishop Tuff. Previous observa- discussions in the field of the eruption sequence of
tions [7] showed that the early GM rhyolites had the Bishop Tuff. The authors would also like to
been stored for periods ranging from 60 to 360 kyr thank Gail Mahood and Diane Nielsen for their
before eruption. The data presented here provide assistance in the collection of sample LV/8/1, [CL]
560 f.N. Christensen, A.N. ~allida~/~art~ and Planetar?, Science Lettem 144 /I9961 547-561
Nd, Sr and 0 isotopic investigation into the causes of sion in K-feldspar, Na-K feldspar and anorthite using
chemical and isotopic zonation in the Bishop Tuff, Califor- Rutherford backscattering spectroscopy. Earth Planet. Sci.
nia, Earth Planet. Sci. Lett. 68, 370-290, 1984. Lett. 113, 411-425, 1992.
[20] R.A. Bailey, G.B. Dalrymple and M.A. Lanphere, Volcan- [30] D.J. Cherniak and E.B. Watson. A study of strontium diffu-
ism, structure and geochronology of Long Valley caldera, sion in plagioclase using Rutherford backscattering spec-
Mono County, California, J. Geophys. Res. 81, 725-744. troscopy, Geochim. Cosmochim. Acta 58, 5 179-S 190, 1994.
1976. [3 I] B.J. Giletti, Rb and Sr diffusion in feldspars, with implica-
[21] EA, Mankinen. C.S. Gromme, G.B. Dalrymple, M.A. Lam- tions for cooling histories of rocks. Geochim. Cosmochim.
phere and R.A. Bailey, Paleomagnetism and K-Ar ages of Acta 55. 1331-1343. 1991.
volcanic rocks from Long Valley caldera. California. J. ]32] B.J. Giletti and J.E.D. Casserly. Strontium diffusion kinetics
Geophys. Res. 9 1, 633-652. 1986. in plagioclase feldspars. Geochim. Cosmochim. Acta 58.
[22] W. Hildreth and G.A. Mahood, Ring-fracture eruption of the 3785-3793. 1994.
Bishop Tuff, Geol. Sot. Am. Bull. 97, 396-403. 1986. [33] G.R. Davies and A.N. Halliday, unpubl. data.
[23] C.J.N. Wilson and W. Hildreth, Bishop Tuff revisited: new [34] F. Lu. A.T. Anderson and A.M. Davis, The Bishop Tuff:
insights on eruption timing, Geol. Sot. Am. Abstr. Prog. origins of a high-silica rhyolite and its compositional zona-
23(2). 109. 1991. tion, J. Geol.. submitted.
[24] A.T. Anderson Jr., Hourglass inclusions: theory and applica- [35] B.R. Frost and D.H. Lindsley. Equilibria among Fe-Ti ox-
tion to the Bishop rhyolitic tuff, Am. Mineral. 76. 530-547. ides, pyroxenes, olivine. and quartz: Part II. Application.
1990. Am. Mineral. 77. 1004-1020, 1992.
[25] K. Ludwig. Isoplot, version 2.71. Rev. of USGS Open-File [36] F. Lu and A.T. Anderson, Mixing origins of volatile and
Rep 91-445, 1994. thermal gradients in Bishop magma. EOS Trans. Am. Geo-
[26] J.N. Christensen. A strontium isotopic study of processes in a phys. Union 72, 312, 1991.
silicic magma chamber: the Bishop Tuff, Long Valley, Cali- [37] W.A. Duffield, J. Ruiz and J.D. Webster, Roof-rock contami-
fornia. M.S. Thesis. Univ. California, Los Angeles, CA, 81 nation of magma along the top of the reservoir for the Bishop
pp., 1987. Tuff, J. Volcanol. Geotherm. Res. 69. 187-195, 1995.
[27] G.A. Izett and J.D. Obradovich. Dating of the Matuyama- [38] S. Nakada, CR. Bacon and A.E. Garmer, Origin of phe-
Brunhes boundary based on “Ar/s9Ar ages of the Bishop nocrysts and compositional diversity in pre-Mazama rhyo-
Tuff and Cerro San Luis rhyolite. Geol. Sot. Am. Abstr. 23, dacite lavas, Crater Lake, Oregon. J. Petrol. 35, 127-162,
A106, 1991. 1994.
[28] A.M. Sarna-Wojcicki and M.S. Pringle Jr., Laser-fusion [39] F. Lu. A.T. Anderson and A.M. Davis, New and larger
“‘Ar/ “Ar ages of the Tuff of Taylor Canyon and Bishop sanidine/melt partition coefficients for Ba and Sr as deter-
Tuff, E. California-W. Nevada, EOS Trans. Am. Geophys. mined by ion microprobe analyses of melt inclusions and
Union 73, 633, 1992. sanidine host crystals, Geol. Sot. Am. Abstr. 24, A44. 1992.
[29] D.J. Cherniak and E.B. Watson, A study of strontium diffu-